Home Automation Hangout 2020-09-27: EtherUNO prototype assembly

please [Music] good morning superhumans okay so very first thing i'm going to do a little audio test well my audio test for now is i acknowledge the traditional owners of this land i pay my respects to their elders past and present and to aboriginal elders and peoples from other communities who may be taking part in this live stream today now we need an audio test because this is the first time i've tried with this particular radio microphone going into obs and i'm seeing the audio jumping around a whole lot oh peter hey good to see you um and sorry about the music giving you a shock and it looks like it might be really noisy but i don't know i think your shirt is making noise yes it could be um we'll see what happens the problem is that today i've got to be moving around a whole lot which means that uh the microphone that is just at the computer is probably not going to be the best thing a bit flat muted does not reverb the room okay um part of the slight buzz might be the all right i know this is kind of a bit silly to be doing this messing around on camera but i am going to try switching to a different microphone plugged into this one so uh we'll see if this makes it better i assume you can still hear me at the moment yes this is all just making it up as i go so oh come on get that out and get that out of the way okay so at the moment i'm wearing a body pack microphone and it's probably not the best way to be doing it all right so i'm going to hit mute for a second because i'm about to plug in a little lapel mic into the body pack and it's probably going to make a bunch of noise while i'm doing it but first i've got to find my cursor there it is okay here goes silent mode actually yes james's suggestion let's try that i'm going to take that i'll just clip it on here and see how that goes maybe i'll clip it on lower down we'll try there all right that probably made that probably made a whole bunch of noise and annoyed you while i was doing it oh well no more scratchy when you move yes all right so i think next time i'm trying to use this microphone i'm going to use the lapel mic plugged into the microphone pack so ah okay what i'm going to do today is assemble an updated ether uno prototype hello from reading pennsylvania um oh unexpected see on so what's the mic it looks cool it is a road what is it road go i think it's um in fact in a second i can show you the receiver so right now you are looking through my my main camera i'm going to switch to the poker phone so now we've got a portable camera as well and the receiver now i'm making you all c6 so there you can see my main camera and you can see the little receiver that is stuck on the top of the camera that's the rode receiver and the transmitter basically looks the same so it's like size of a box of matches or something it's got a built-in microphone and it's also got a thing where you can plug in an external lapel microphone um the really nice thing about this is that it's got usbc on both the transmitter and the receiver and it's got a lipo battery in it which in theory should give me about five or six hours of life which is a huge improvement on the other radio mic which had two double a's in it and it would go really flat uh it would go flat really fast and because what i really don't like about using radio mics with regular batteries is if you can't begin an event with a battery that isn't full because you don't know how long it's going to last so typically what happens is you end up pulling out the batteries and replacing them right before you start every single thing all right yes fix the camera review okay so today is going to be a little bit strange because a lot of this i'm going to have to be focusing on and it's going to be a little bit hard for me to follow what you're talking about i've got oh pancake you're here awesome i've got a tablet over here so that i can hopefully see hang on can i unlock it yes so i can see the feedback coming in from andy on the back channel all right so let's start with applying a bit of solder paste and i was hoping to have a whole lot of things organized in advance at the live stream so just before this live stream started what i did was pull out some old boards you can see some old donor boards here these are old ones that have failed for whatever reason oh dead psu and ethernet so you know scrap boards whatever and i went through and pulled off a bunch of parts things like resistor networks and voltage regulators and a couple of other odds and ends there are some inductors here these are parts that i don't have ready to go so like as new so what i've done is just stripped a whole bunch of them off old boards and then i can populate them onto the new one so some of the parts we're going to populate using the pick and place machine and some of the parts i will just migrate over so things like i even grabbed a an atm mega 16 u2 so this one's already got the uh the usb to serial thing pre-loaded into it so it saves flashing it there's a cpu atmega328 voltage regulators grabbed a card cage these things are really hard to get off without damaging them they're very annoying because they've got this big plastic base on them when you heat them up the base just tends to melt so you've got to be very careful with it so i used my pcb preheater i wonder if i can get this across into the field of view well if i take off the power cable i can so i used this which is my pcb pre-heater it's very very cool it's basically a heater in the box it blows hot air out the top here and these are magnetic board mounts so what you can do is take a donor board like this one and you just pop it in between those two mounts so that it's over the area where it says hot surface and you can see there that it's suspended over the heater outlet oops it slid and then if this was actually plugged in what i could do is turn it on you can adjust the temperature of the heat output and you can have it in hot or cool mode so what you do is put the board on top of this and you just turn on the heat let it sit there for a couple of minutes and what happens is the whole board heats up and it doesn't heat up enough to melt the solder this is not for reflow this is just for bringing the temperature of the board up to the point where it's close to um to melting the solder and then i grab the hot air bring it down and then you just warm up the particular area of the board where you want to get the parts off and then the solder melts and you can lift the parts off so the idea is that doing this is better for a few reasons one is that it brings the whole board up to a certain temperature instead of having very intense heat in a certain area if you start with a room temperature board there are parts of the board that are going to be at you know 20 or 25 degrees or whatever your room temperature is and parts of the board that are going to be at 300 degrees so it can put mechanical stress across the board if you bring the entire board up to up to something that's below melting like say you bring it up to 120 or 140 degrees then when you heat up specific parts of it to remove or rework those areas there is less thermal differential across the board so you get less mechanical stress because of expansion and contraction and less thermal shock because the temperature is not changing as rapidly as you heat those parts up oh look we're chasing squirrels already i thought i was going to be assembling boards today but let's see oh i just saw there was a question from mike causer removed with hot air gun uh yeah i'm kind of already answering that question so yeah i went a long long time without a board pre-heater i've had this one for a couple of years now but it's one of those things that i'd heard about as being cool and worthwhile and i um but you know there are some tools that until you've used them you don't understand the value of them because you just don't have any reference or yeah it's that thing if you've used it then you understand how good it can be so where is that pcb over here i stuck it on the pick and place machine which we will get back to in a while but first we need some paste now i'm going to try to keep track of questions and keep talking while i'm applying paste but this is not exactly a spectator sport and also i don't have um i don't have a stencil for this so in production of course you would use a solder paste stencil to apply all of the solder paste now one thing that's a little bit strange is that the field of view of what you can see through this camera is a lot smaller than the field of view that i have so i've got to i'm going to have to work to keep the the pcb visible under the camera by trying to keep everything in the center so i'm just gonna sit here and squirt on some solder now question is what can i talk about while i'm doing this that is not going to be too much of a problem and stop me from uh from being able to do this properly i don't know let's see um what is that part i don't know oh that must be the fuse is it it seems to be on the wrong side oh no that's right the fuse ended up moving i'm holding the board upside down at the moment so the my perspective on it is all twisted but the the poly fuse now this is the bit on the the part on the board that i hate the most it's a qa fan now doing this with a stencil is pretty easy doing this with a little syringe like this is kind of annoying getting the right amount of solder paste out onto it and trying to get it so that it will reflow without bridges now you'll see that many of these pads don't have any connections to them and so you might be wondering why i'm bothering to put solder paste on those pads the reason is that if you don't surface tension can end up pulling the chip out of alignment so [Music] if you have pads that are pasted on one side of the chip and not on the other the surface tension will end up floating the board in that particular direction and it can end up skewing so you can see for example down this side of the chip there are many pads here that just don't have any connections it's only the pads on the end that are connected but i've put a bit of solder paste on them anyway and i also do the same thing on this so this is the usbc connector and the other thing is that i could just run a bead of solder straight down across all of these pads that's what i used to do and recently i've started putting paste directly on each pad like this and it's all going to smush together when it reflows anyway so i'll just squirt a bit in here that'll give it a bit of mechanical mount i'm going to use a soldering iron and that's interesting that slot is shorter i wonder if that's going to fit oh well we'll see when we get to it so when i put the socket on what i'll do is put the socket on and it's going to be reflowed in the oven but then i'm going to use the hand soldering iron and i'll make sure that it's properly soldered anyway for mechanical strength so i've just put a bit of the solder paste into the mechanical holes for the usb socket so that when it just does go through the oven it'll have a bit to hold it in place now one of the changes for this version of the board see that one there that's a diode is it in the field of view yes it is and somewhere around there's another there it is there's another diode so there was some discussion on discord uh about two weeks ago now it was after i showed a previous version of this and there was a discussion about isolating the power source from you know the power of ethernet power source so that it couldn't be back powered and it's something that i didn't do on the and in fact the ether mega the ether deway none of them have an isolated back power protection on the poe header so what i mean by that is power can come into these boards from several different sources here i'm just mushing it around here rubber squirting some paste uh so power to the ether 10 can come in from usb obviously if you just plug it into your usb port it'll get 5 volts from the host computer and that goes in basically it ducts around the back of the voltage regulator the five volt regulator and feeds the five volt rail directly on the board so if you power from usb the on-board hoops smearing it around a bit there the on-board regulator is not going to be used at all it just takes the raw five volts from usb and it feeds it to the rest of the circuit including the input to the 3.3 volt regulator there is also the dc jack and if you power from the dc jack it goes into the input side of the regulator but there is the complication is that with these boards because they've got power over ethernet on them if the i should explain what these particular parts are too because it's okay d2 are from power you'll see across here there are a couple of resistor networks and i'm really screwing up the solder paste application here but it doesn't actually matter all of that much these pads here so this chip is the w5500 which is the wiznet ethernet chip in fact i'm being lazy i'm going to smoosh straight across the pads here and do it here and do it here now this probably means i'll end up with solder bridges but that can be fixed fixed in post-production uh so these just up here these resistor networks they are they are bias resistors on mode pins for the w5500 now there are how many from memory i can't remember the circuit is up on the screen but i can't see it right now i think there are about six oops i don't need to be doing that one uh there are about six mode pins well there are eight connected here but there are a couple for other purposes and what they do is set the operational mode of the w5500 so it can be you can fix it so that it will only run in 10 base team mode or only 100 base team mode this one i don't think i'm actually going to be placing sorry chasing squirrels again uh so you can fix different modes or if you leave the pins floating it goes into an auto negotiation mode so you can do things like enable or disable whether it does mdix and which is negotiating the polarity of the connection which is for avoiding well not really polarity it's negotiating the tx and rx so if you've got two devices plugged together you don't need to use a crossover cable that feature you can turn on and off oh my syringe is getting really mucky there using those bias resistors so that's these resistor networks here now if you leave them all floating it goes into an auto negotiate mode on everything basically it enables all options but the internal pull-ups and pull-downs some of them go up some of them go down on those particular pins are fairly weak i think according to the data sheet they're equivalent to like 86 kilo ohms and so the recommendation is that you can either just leave them all floating and it will go into order negotiate everything mode or you can bias them explicitly externally which improves its uh emc rating so it makes it less susceptible to external noise that might cause it to have i finished this already i've just been talking and not paying attention to what i'm doing so yeah it puts it into a mode where it will go sorry i'm pausing while i'm checking over this board to see if i've missed any parts i don't think so uh i'm gonna have to switch to the other camera in a second it biases the pins internally i've lost my train of thought entirely and the because the pull-ups and pull-downs are weak if it experiences a lot of external electrical noise it can possibly boot up in an undefined state or an unknown state something that it might see one of those pins is being asserted when you don't want it to be so what i did was explicitly add resistor networks to bias those pins that was a very long-winded explanation all right so what do we have here let i'm going to switch this camera around just temporarily i think so that you can see this other screen i don't this is on a totally separate computer i don't have any screen sharing or anything set up so i don't have a direct way of getting access to this screen and this is going to be really quite awkward to use because i need to be aware the camera is all right so what i'm going to do is start by homing the machine well the machine's already homed i was just wondering if i should show you that process but i don't think i will i think i am just going to check a couple of the feeder positions so the crystal feeder if i click this little button on here that will move the camera to the pickup position for that crystal and i can see on here that the position looks accurate so that's okay and 10k resistor move the camera to that pickup position for that looks okay it's slightly offset to the left but oh well close enough i'm not that fast so just park the machine and oh i better check that the board is in the right place so if i go and have a look at let's see we'll pick a resist oh pick a diode led and move the camera to that position and here i can see that the position looks all right let's just pick another part pick a crystal where is that that looks centered enough let's find a fiducial uh oops there's one let's go and have a look at the fiducial it's very slightly off center that one's spot on that one's very slightly low so anyway the board position looks alright now when i'm actually running this job unfortunately there is going to be quite a bit of noise when i'm doing this i'm normally just moving you around to another perspective yeah when i'm running the machine i normally wear noise cancelling headphones because the compressors the compressor is annoyingly loud well i don't actually have a proper compressor on this i've got little 12-volt air pumps so [Music] this is kind of going to suck but first i'm going to enable the auto feeders so feeders are on i've got to turn on the air pump that does the blow off so that is the pump that provides positive pressure and i'll turn on the main back pump i'll just turn it back off again so i can talk to you again for a second before i run the machine now was there anything that i needed to see in the chat because for the next few minutes this is going to be making noise and i'm not really going to be able to talk to you um okay oh mike asked a really interesting question is it worth having a generic stencil with only a qfn so you could use it once on any board with a qafn that would be really useful but you'd have to put it on first so that's actually a really clever idea you could just align it with the particular qfn part and put the stent put the paste onto that and then manually put the the paste onto the rest of it all right so [Music] now this is another one of those situations where i have failures on this pick and place machine so often that this could all end up being a disaster but let's see how we go all right i'm going to enable the air pumps so we're going to have a whole lot of noise and then a few minutes of action while it's putting pieces into place and while it's doing it i'll probably pick the camera up and see if we can get a slightly better view of it so here we go vacuum pump on and then the noise will begin [Music] me [Music] me [Music] so me me green green [Music] me um [Music] [Applause] [Applause] um um [Applause] [Applause] [Music] [Applause] three now you can hear me again all right so i'll just move the camera back down here and then i'll chuck this under the microscope and we'll see how badly we stuffed it up well we i i'm not putting any of the lame for this on you so there are a whole bunch of parts on here that haven't been placed and i'm sure that most of the parts that are on there are only approximately in the right position but i'm not really paying putting a whole lot of care into this uh so let's see oh yeah oh look these ones are all offset sideways i wonder why that is that was weird [Music] oh look it's only barely on the pads that would have probably reflowed okay oh these ones are really off it's all off to the right i wonder why uh i think there were i know that's right these were odd value parts that i'm going to have to pull out of my mixed parts kit i didn't have those in feeders i don't know what that part is we'll find out and those ones lots and lots of parts missing so unfortunately i'm going to have to manually place quite a few parts here what's going on with that yeah it's all right it's all right not great but it would reflow oops i think my finger was on that one okay so that leaves a very sadly large number of parts that still need to be put on by hand so let's start with just chucking on the ones that i've lifted from this other board uh where's the oh yeah there's pin zero so i'll chuck on the atmega and i also need to get a w5500 from somewhere that could be kind of important oh and i need uh one of these chips i might even leave that off for now i can always do more of this later hang on that machine is buzzing and making horrible noises in the background so let's turn this off uh machine is parked and i'll disconnect from the motion controller and yes yes yes timed out waiting for connection do i want to save yeah okay so quitting out of open pnp and you probably can't even really see me right now turn that off turn that off turn that positive pressure pump off and now we have some peace hooray much more relaxing oh except that the feeders are still active i'm going shut down that computer and then the feeders will shut down as well yes now we have some peace so where was i that's right i was grabbing pieces off here so i'll chuck that one on that's one that i lifted off some other part and these parts are really close to the edge right where my finger goes so that's a little annoying all right so a w5500 i've got to grab from somewhere i should have a few of those in fact i probably have them just up here i do indeed now what's happening in the chat um for hot air sand in a pan and oven sand for hobbyists on the cheap which i wasn't expecting that's interesting so in the chat there was some question about using hot air for hobbyist level stuff and a suggestion for a video uh by great scott about using sand in an oven i've never heard of that technique i assume that the sand is just a way of providing some thermal stability or something so that it maybe so that it can [Music] um yeah give some thermal mass and bring everything up to the same temperature i don't really know what that's about uh what else the uh okay so some questions about the pick and place machine so ian said is that the sound of the servo motors or stepper when aligning the fiducials so that uh there were two different things you're hearing one is the stepper motors and they kind of whine all the time at the moment the motion controller is set up so that the motors are engaged constantly which means that even when the machine's not in motion the motors are powered and that stops anything from moving so that it maintains its position because it's an open loop control system it's not closed loop it's got no feedback on its actual position other than counting steps but the the main annoying buzzing noise is that all of the feeders have servos in them and they're just little hobby servos total really cheap little tower pro digital servos so when all the um the feeders are powered up those servos just sit there going constantly which is really annoying uh um inc i asked if i've tried sticking some foam around them no i haven't so the servos are just inside the 3d printed feeders um and oh paul asks it's picking and placing all the components no so i just did some of the basic jelly bean parts all of the unusual ones like the the ics voltage regulators and that sort of stuff that aren't loaded in the machine i'm just sticking on by hand but what i find is that using the machine for hybrid assembly is a real help because i've got things like all the really common stuff like 1k resistors 10k resistors 100 nanofarad capacitors you know all of the things that you find that you have in pretty much every single project are loaded on the machine leds you know common stuff which means that it's actually a really big time saver even for a single board like this to put it on the machine and let it put all of those basic parts in place and then all you have to place by hand are the exceptions so just the unusual parts that aren't already loaded so what else do i have i've got a um an fdn here at some point i'm going to have to look up the the overlay for this board and see what parts are actually meant to go on it these ones i just put on because i happen to know what they are they're obvious it's a bit hard to miss the cpu but many of these other parts i don't know what they are not without checking the positions so while i'm here and to get them out of the way i'm just going to stick the big fat voltage regulators on there [Music] and there are some electrolytic caps that go on here but i'm going to leave those for now because they are very easy to bump and end up getting out of position so now i'm going to do some uh diodes so just standard uh 1 in 404 so you know the classic rectifier diodes i actually have these on a in a feeder in the machine and i'm i didn't place them because there is a something weird going on with that particular feeder i did a board a couple of days ago and when the machine moved across to that feeder it lost its position and then started putting all the subsequent parts down about 30 millimeters too far to one side not just slightly out it was like it had missed a bunch of steps or something so i need to figure out what that is but in the meantime i just put those parts on and i skipped them in the setup on the pick and place machine now uh that's a one meg resistor i happen to remember that one so i've got a resistor kit here which i use for all of these parts so a little one meg resistor that goes across the crystal hmm interesting why didn't i put a one meg resistor across the 16 megahertz crystals i don't remember it's pretty normal to have something like a high value resistor across the uh the crystal so you got the two loading caps on the side of it the crystal and then in this case a one meg resistor across it but i didn't do that on the 16 megahertz crystals i don't think they are on the ether10 either ah okay so seon's comment about the foam was about the vacuum pumps they sound like they are vibrating yeah they're noisy because it's they're not a proper if i had a proper compressor and someone else who was it oh dodgy suggested having a big vacuum canister like some cars do to provide a vacuum reservoir yeah that's pretty common as well and on diy pick and place machines i've seen a few people that have used these little vacuum pumps the little 12 volt vacuum pumps and then they just use something like a can as a vacuum reservoir and attach some fittings to it and hoses that seems to help as well so that's something that i might do all right i think it's at the point oh not quite i was going to say i think it's a point where i need to look at the overlay but there are some more parts that i can put on here without needing to look anything up there are some little tantalums they're going to different places here where are there i think there were a few tantalums on this board there was that one surely that's not the only one maybe it is i may have ended up substituting other parts in place of the tantalums just scanning around the board now trying to find it hopefully this isn't making you seasick no i can't find any other oh there's one there is one other tandem that goes on there that must be on the 3.3 volt rail yes it is so this is the 3.3 volt regulator these are the status leds for ethernet and you can see that there are some resistor networks that go in here for current limiting on those now what value did i have i've just got to have a look over here so i used 1k resistor networks on there as current limiting on those leds they oops ah i messed that up they don't need to be bright so in fact you could even use a higher value than [Music] 1k but this will do for now and while i'm doing this i might as well put on the other resistor networks as well so what have we got we've got one up here we've got two across here there's another one up there another one here so this one is the so this is the back of the ethernet connector and then these are the data lines from the ethernet connection so it's tx positive and negative rx positive and negative and you can see just here this is where the static protection things go on so i need to add those as well i can't remember what that part is but these ones are 49.9 ohm resistors i think let's see did i pick up the right one yes i did 49 uh 9. they go into the data lines like there's um some other yeah it's so long since i've looked at this circuit now it's been weeks and weeks so my memory is not that great uh other things somewhere in here there is an inductor in fact that might no where is the inductor i haven't fitted it i need to put it on manually somewhere around there i need to put an inductor in okay so first though i'll just chuck a couple of resistor networks on here and these ones are bias i've got a bunch of 1k networks so i'm going to put those on these could just be 10k could be 22k or whatever it doesn't really matter this is just to provide a stronger bias to those mode pins 1k is a lower value than i would normally use on on bias resistors particularly if the bias resistors need to be overcome by something else like if you've got an input that you're pulling in one direction or the other and then you want to what is that one 10 micro farad one and then you want to be able to overcome that oh one micro henry there it is so there is the inductor so normally if you are having some kind of an input like um a digital input connected to a switch now where is it there it is that's where the one micro henry inductor goes uh so the purpose of that inductor is to provide some isolation from or for the digital supply for the ethernet chip from the rest of the circuit so it keeps the power supply for the uh for that chip nice and clean now i also need what else is there so a resistor network that goes right up in this corner for these other leds and different things is probably 1k yes it is so that's a 1k resistor network so i'll chuck that one in and what's this one 10k all right so i've got to stick a 10k resistor network on there and a 1k up in the corner where are we up in the corner there and while i'm here i may as well stick in the atmega 16 u2 so pin one down in the bottom right corner and i've got one here that i floated off one of the other boards so basically this chip is totally upside down alrighty uh where were those resistant networks over here so this one is the 10k resistor network is that a 10k yes it is i've just got a little pile of resistor networks that i pulled off these other boards so i've got to check each one just to make sure i'm putting the right one into place this is the mac address rom and unless i pull one off another board i don't think i have one handy oh a couple of more resistor networks there's that one there and this one here these are the voltage dividers for the sd card and i can't remember which way around they go so let's cheat and look at the circuit what do we have here so the one down the bottom is a 2k 2 and the one above is a 1k so so you can see that's a 1k resistor network and down here is a 2k2 resistor network so 1k goes up the top 102 and then this one is a 2.2 k resistor network just trying to read this with without using the microscope yes i picked up the right one i'm sure that assembling boards like this is not a spectator sport this seems like a particularly boring way to be spending your time sitting here watching someone else assemble a board particularly when that person is distracting themselves talking about other things and not working very efficiently i'm just going to grab a surface mount button out of my my parts cabinet stuck on the wall over there and stick the reset switch on there okay so this here that you can see there's no crystal on there that's because i don't actually have any 25 megahertz crystals in that footprint so i'm going to bodge an hc49 us format crystal into that position now i am actually really tempted to find one of these uh mac address roms on another board and float it off and put it onto here so that we can test that feature i think i might do that even though it's going to be noisy i'll use the it'll be a chance to show off using the um the preheater which is cool for those who haven't seen that in use before but first what's going on up here just saying but first reminds me i can't remember if i've mentioned this before it is um a friend of mine uh paul fenwick he's got a bit of a funny sense of humor many many years ago he implemented a uh a feature in pearl because he's one of the core pearl developers i'm not sure if he's doing that anymore but um he used to hand just into richardson used to run pell training australia and as a bit of a joke at open source developers conference many years ago he wrote a module for pearl which implemented a feature called but first and the uh the logic for it was that you're writing your code and you've written all this stuff and you're going down down the page and then you realize that you want something else to happen before the thing that you've already written now the logical thing to do of course is to go back up and edit your previous code but in paul's bizarre logic he decided wouldn't it be funny to have a language construct where you would type that out and then you would type but first and then do something totally different and then when it was executed it would go through and execute the thing that you'd said but first first even though it was not first in the program so you could write programs backwards and as if pearl is not enough of a brain twister that made it even more interesting so uh where what am i looking for i'm looking for those two pads that you can see there that are unpopulated which are directly behind the usb connector and what are they oh there's a 5k one that's in place my overlay position here is not very good oh pgb so these are static protection tvcs and what else are we missing around here nothing okay so we need two static protection um things which are once again i keep those you know draw over here so i'm going to go and cut a couple of bits off the tape i've got little bits of cut tape stuck in the drawer over there where you couldn't see out of camera view so i've just cut off a little bit of tape drop it out on the bench yeah this um whatever i'm talking about during this process is even more of a random train of consciousness than usual because i keep being distracted by what i'm trying to do which is assemble this board i haven't even got back to the whole issue of the power source and the extra diet that i started talking about right at the beginning of this process all right so these are the little xena protection things that limit the voltage that can be applied to the data lines on the usb connection and that is to provide some protection both for the host computer although they're usually pretty well protected oops i'm out of field of view there and for this device now what is that part that wasn't placed i've got to scroll over here one microfarad 603. that's interesting i thought i had one microfarad 603s loaded in the machine but apparently i oh i do why wasn't that one placed i don't know maybe i just didn't set it up when i was loading the job into the machine so i'll have to go for the generic resist generic capacitor pack and there we go so one microfarad o603 oh look you can see i've got just a dab of solder paste ah there must have been some solder paste on the nozzle which can happen after something like um where things go wrong like you get a mispick and then it tries to place it because it hasn't detected the part isn't there and you can end up shoving the nozzle of the pick and place machine into the solder paste which is never good you end up with solder paste stuck on the nozzle and then it attaches to whatever part you pick up next and sometimes it can do things like make it difficult to put the part down and that's one of the reasons that i set up positive pressure on my machine so i've got a little aquarium pump oh look there's a part that isn't placed and a couple above it what is that one and what are the two above it so yeah i've got a little aquarium pump providing some a small amount of positive pressure now where on the board is this it's between the voltage regulators which is uh down here so that one is placed this one is not oh it's another one microfarad 603 i think yes it is i shouldn't have put away my cap kit ah yeah so i've i um originally had the vacuum nozzle set up so that when it was placing a part it would just isolate the nozzle from vacuum so it was no longer positively pulling it onto the nozzle and i then changed it so that i've got to blow off positive pressure but you need to make sure it's only a small amount of positive pressure because the problem is that otherwise it can basically be like a little high pressure air jet what have we got here 4.7 microfarad and 10 nanofarad okay so what can happen is it'll blow other bits of other parts around off the board i've had that happen before it's not fun now this one this is going to be a real bodge this is a 10 nano farad it has to go in between and i'm putting an 0805 part here onto an o603 pad that'll work just fine these are slightly oversized pads anyway you can see that the pad still overlaps the end of the part and then i need a 4.7 micro farad that is a weird value to have what have i got 1 10 hmm how can i do that i don't think i have a 4.7 microfarad i don't even know how critical that one is i have a feeling that that is just a bias no what is it i think it might be for the internal regulator of the there we go switching to schematic mode for the internal regulator of the w5500 and where is that part there it is 4.7 microfarad and it is 2 cap so 2 cap yeah that can be 10 micro farad i remember reading about that on the uh on the wiznet forum someone was questioning that so that is just part of the internal voltage regulator internally the wiznit this the w5500 i think runs at 1.8 volts i can't remember it'll be there somewhere shown on the nets or 1.2 volts so internally so what you do is you feed it 3.3 volts and it generates its own internal 1.2 volt reference and that 4.7 microfarad capacitor which is where the one is missing right now that you can see in the center of the microscope view that is essentially a like a bulk cap for the internal voltage regulator and it is perfectly happy if you give it a bigger value not ridiculously bigger but the data sheet suggests 4.7 microfarads but you can put 10 microfarads on there and it's not a problem well when we get to testing this we'll see if it's a problem now we've got two more parts here what are these ones ah okay interesting so well there's another 10 micro farad up my 10 nano farad up there but there is also a 12.4 k resistor and this is the external voltage reference this was discussed on a previous live stream so first i'm going to stick on the 10 nano farad which i also don't have in i-603 i do have it in o402 so i could just stick an o42 on there but that'll do so this part that is missing now is a sorry i'm looking at my resistor kit it is an external reference for the internal voltage regulator once again so that one needs to be 12.4k i've actually just put on a 12.1 k and yes i know that works this is just to check that i haven't screwed up the pcb fixes the new layout so what have we got here 10 nano farad 22 nano farad and 10r okay so these are changes 10 ohm resistor i don't have one of those in the pick and place machine so i'll grab one of those manually and what i really should do is have a look at the chat in a minute i'm probably talking to myself because you've all left because i'm ignoring you uh 10 nano farad and 22 now farad [Music] let's see where is another 10 nano farad this one another r805 just like a jumbo part you know those novelty pencils that are oversized and it takes two hands to pick one up that's what it reminds me of using oh 805 parts they're so big 22 nano farad all right where is a 22 nano farad oh look i don't have those you know 603 either only in o402 and oh 805 so i'll stick an 0805 on there uh what's going on here we're getting dangerously close to having all the parts placed now we still need the sd card i'm leaving that towards the end because it's easy to pump that crystal doesn't go on oh that's right we don't have the mac address rom yet and i'll leave that till last then we've got the two capacitors amazing i think we're nearly done thanks for sticking with me through this all right so to get this mac address rom i am going to pull it off one of the other boards what will i pull it from okay there's this one yeah i think i'll pull it from this one this is the second most recent of the prototypes so to do this let's see if i can do this in a way that you can see uh i'm gonna might even do it under the microscope which normally i wouldn't do but that way it'll be easy easier for you to see what's going on so can i get this into the right position and also this is going to be noisy really noisy oh i've got to lift the microscope way way up because oh look you can see some of the hacks that i did on this board oh look little chops and things so this is on a previous prototype now that part that says hsr7 that is the mac address rom i'll just move it across there a bit and it's going to become a bit loud now unfortunately so on and heat is on all right so while that's warming up i'm going to have a quick look at whether there are other things i need to respond to in the chat which i'm sure there are by now okay all right so oh my cause i asked our resistor networks hang on i'm going to move away from this uh noisy little thing so that i can talk to you without being overwhelmed without you being overwhelmed by the noise of it so are resistor networks just 4x in the same package in parallel yes exactly that that same job could be done by putting four single resistors in place resistor networks are they're convenient because it means that you can put them on with a single movement of the pick and place machine so instead of having to go to the feet or four times and put four things down you can pick up one thing and put it down once which means effectively it speeds your assembly process up and makes it four times faster and it also saves it can save a little bit of space i'm going to get back to chat i'm going to get this piece off so that i can then turn off the hot air because it is loud and annoying so i don't know if you can see what i'm doing in the side camera so i've got the machine i've got the pre-heater running the board is preheated and i'm going to use the hot air to warm up that particular part well you can see down below i know it's off your field of view there is flux on the board that is reflowing already so you can see that i only had to touch the hot air on that for a moment and that part lifted right off so now i'll switch it into cooldown mode and shove that back out of the way i normally leave it running in cooldown mode for a minute not so much for the benefit of the board but for the benefit of the reef of the pre-flow heater because if i turn the pre-flow heater off when it's at peak heat the airflow stops and then there is potentially a problem with heat soak and overheating the mechanism of the heater but that's been running for a minute now so i'll turn that off which will be a great relief to everybody all right so now i've got to get this microscope back to its normal operating height down about there somewhere looks okay look there's a hair and getting into the the works all right i can't pick it up okay there it is okay so back up to the top of the board up here so this is where the mac address rom goes pin one is in the bottom left now that question a second ago was about resistor networks and you can see here for example i've got these two resistor networks that could have been just eight resistors in parallel which would have meant eight placements but instead we've got two networks and it's two placements so you can see mechanically it's quite obvious what it is it's basically it's almost like you've got four separate resistors bound together mechanically so that it's into a single thing and typically if you're using something like one of these k16 networks physically that will be a little bit less board space than if you have four o 603 resistors side by side now let's get this into place so pin one is down here that's where the dot is i don't know if you can see it through the camera but there is a dot on this part as well which is on this so that's aligned okay i think we are just about at the end of our assembly process i'm going to stick the sd card on there now the thing is that this particular revision of the board has it does have a couple of important changes on it like related to the power supply but much of it is exactly the same as the previous prototype including the sd card area so i'm as confident as i can be that the sd card is going to work but it's always a good idea to do a complete end-to-end test if you can so i'll populate that and also this part of the circuit hasn't changed at all in multiple revisions of this and it's always worked perfectly so that part is probably okay now we are down to the last couple of things let's stick on these electrolytics oh look at that you can see under the microscope that this electrolytic has a melted base let's see what's the field of view like there it is look down around here it's all melted up here it's melted that's because i've pulled that off with the hot air gun off another board just before the live stream began i just got a couple of those boards up and just started stripping parts off them and now we've got the usb-c socket so i'll turn the whole board around because i'm right-handed i want to come in from this angle and put a socket on so first i've got to get a socket i've got them on a strip of tape over here i'll cut one off and put this one on so usb c sockets are a total pain i hate them well i love them and i hate them i love them from the point of view of the cables and the mechanical connection and convenience of them i hate them from the point of view of board design and compatibility now let's have a before i stick it on the board i'm going to turn it over and show you the bottom of it so this is the bottom of the usbc connector and this is actually a simplified connector a full usbc connector has uh 24 pins plus a couple plus like shield and that sort of thing this has only 16 pins and that's because this particular connector is designed specifically for use in systems that only require backward compatibility with usb 3 and they don't require all of the extra feature pins like you're not going to be running video or other things through it it's just a an old style usb connection but with a new usbc connector so you can see here these are the power pins the ones that are doubled up to increase their current carrying capacity because usbc can take a ridiculous amount of power and then we've got a small number of pins through here which are the ones that provide data and things like the mode pins the mode that need to be biased to say what mode you're operating in because usbc does everything okay so i'll just chuck that one on the board uh the first few usbc devices that i made had um uh yeah had full usbc connectors on them and they were horrible to work with they were a combination of surface mount so you can see along the here we've got the surface mount connections that are going to the usbc socket but they also had through-hole pins because i got so many connections that they needed to use a combination of both and the pins were tiny and the holes had to be like 0.3 millimeter and there was no clearance around them just generally horrible things to work with now i'm going to take this out of the microscope have a quick overall look at the board and just make sure that i'm not missing anything i think i have all of the parts on it so let's go back to this view and unfortunately there's going to be a bit of noise with this as well now can we see there okay the reflow oven tends to make a little bit of noise and we've got this huge 3000 watt reflow oven that's going to have a single board in it which is kind of annoying so i've just got to select the temperature profile and start and if you have a device in your home she who shall not be named one of those home automation devices i've got to use a command now so maybe hit the mute button on top of yours so that i don't make yours do silly things so i'm going to say alexa turn on the exhaust fan okay all right and now here is the little control for it and this is the um the control panel for the reflow oven and i've modified it to put it inside this wooden box so basically built a box that goes over the top of the oven and then i cut a hole in that to expose the control panel and down here you can see that there is an exhaust vent there is a white t piece down the back it comes down to this and that blue exhaust goes down under the bench and it runs all the way around the room and there is a oops i'll just move this back to here and there is a an exhaust fan down that way that pulls the air out and it vents it outside the room so while that is doing its thing let's have a quick look at comments that i've missed uh come on open up all right yeah that reflow oven is going to take a few minutes to run i think it's about six minutes give or take for a cycle so let's see all right lots and lots of comments that i've got to catch up on right so ian said could it pipeline the component feeder rather than feed before pick um could it pipeline the component feeder rather than feed before pick so does that mean that you that it would run the feeder to have it in the pick position in advance so yes that is just a setting in if i'm understanding the question correctly i may not be there is a setting in open pnp but it really comes down to how the feeders are configured so what you can do is set it so that the feeders advance immediately so that there is always a part ready and available and that mean basically it just makes your job run faster because the machine doesn't have to stop and wait for the feeder to run before it picks it up with the head i don't have that set up on my machine because i use it in this sort of way a lot i use it for a very small number of boards often just one board at a time and i don't want it to be sitting in between jobs with all the parts exposed so i take the small performance hit and make sure that all of the tapes are sealed and closed until they're actually needed otherwise there'd be like a row of parts exposed sitting on it and no dust gets in and all sorts of things um all right so yeah there are some things i was talking about earlier that i need to get back to but i'll just work through these questions uh all right so number 11 was all the caps i'm not sure what that question is so gerald42 said number 11 was all the caps i don't get that um all right so steve said can you speak about how you get things lined up the initial setting and any drift of components once placed okay cool maybe once all of this is done i'll go back into open pnp and i'll show you some of the setup of that the way there are a couple of different ways that can be done open pnp has import scripts for for a variety of formats so in your cad program doesn't matter what it is if you're going to be sending files off to a factory you need to be able to generate files that they can load into their pick and place machines and the standard way of doing that is um use centroid files so it's a file it's basically a text file and it just lists the position of all the parts so the parts name its value its rotation in degrees and its position x and y and typically they have a dot mnt extension like a mount extension and you can generate those placement files from ltm or keycad or eagle or whatever it doesn't matter what you use it's a pretty standard sort of thing and in open pnp it has import routines for a variety of formats including native eagle files so for so i've got a basically there is a dedicated ubuntu machine running the pick and place machine there's a pc motherboard stuck in the bottom of it running ubuntu and i've got that synced on dropbox so what happens is that as i'm editing files in eagle they all get synced onto dropbox and updated pretty much live on the pick and place machine so inside open pnp i can go to i can just select new board import eagle and it just imports the dot brd file and populates everything in the in the pick and place job and then i go through and assign which feeders are for which parts uh which ones are fiducials and a few other things there's a little bit of setup but for this particular board i did that setup right before the live stream started that was um that was what i was doing up until a few minutes before and it took me maybe 10 minutes to set up the job including so that was from importing the file assigning the feeders for the parts etc it doesn't take that long all right so uh oh so jax tech uh clarified that thing earlier about reflow if you don't have a reflow oven sand in a skillet okay so i guess in that case the sand is uh yeah providing a bit of a thermal mass in the skillet so it's not just the metal itself and um peter asked if i can get back to talking about the power regulator diode yeah okay so that part of the circuit in fact while we're waiting for the reflow oven to do its thing i'm going to switch back to desktop here we go we can look at it in eagle oops that's not what i want this is what i want all right where is that particular part of the circuit i have here we go this is the big change this thing here is a new diode that i just added in this version of the pcb so power input to an arduino can come from the uh the usb connection or the dc jack so with the dc jack where is that um [Music] which shoot am i looking at okay here we go so we've got the usb vcc auto switch we've got v in which is okay i'm going to ignore most of this second for now i've talked about this in the past this particular part of the circuit is a switch that prioritizes power coming from the dc jack over usb so if you have both usb and the dc jack plugged in at the same time it will disable usb power and it will draw power from the dc jack and that's because there are often situations where you might want to use vn and pull it out for some other purpose maybe you're putting 12 volts in on the dc jack and you need 12 volts to power your peripheral devices or whatever you've got connected and also you can pull more current through the dc jack than you can through usb so the priority is to take power from the dc jack and only use usb if there's no other power available that's what this particular circuit achieves so what we've got is the output here goes to the 5 volt rail coming from usb vcc that's directly to the 5 volt rail which is behind the regulator so if we look at the regulator up here and you can see that that net is highlighted that is because five volts coming from usb v6 comes to the back side of the regulator but if this power was not available you don't have usb and you do have power coming in on the dc uh jack and what will happen is that the power comes in here from the dc jack so this is the 2.1 mil dc jack power comes in here there is a reverse polarity protection diode here and that goes on to the input side of the voltage regulator and then you get five volts out but this is the tricky thing normally arduino boards only have those two sources of power if you don't talk about the headers just ignore the headers it's usb or dc in and because this one has power over ethernet support this v in net here is also used for another purpose so you can see here we've got the protection diode on the dc jack and if i go back to where is it so that one no which shape do i need to be on this sheet okay so this is the network socket where power over ethernet can come in and what happens is that if you feed power over ethernet down through here oh hang on i've got to turn off the oven and i'll turn that off and also alexa turn off the exhaust fan okay okay so all the background noises has gone now alexa stop so okay back to this so power comes in from the ethernet cable down this net goes to the poe regulator or like the module or just through the jumper and then it comes in and this diode now feeds into v in so what do you need to keep in mind it's unfortunate that these are on different sheets of the schematic because it's not so easy to see but we have power coming in this is the power from power over ethernet through this diode going into vn if we go back to this sheet here we have power from the dc jack coming through this diode also going to v in so essentially we have an it's a diode based or gate you can take power from either of those sources but because they're both feeding through diodes neither of them can back power the other and that is the primary change that i made with this version of the pcb so the if you have a dc power source plugged in here and you might be exposing i don't know 12 volts or something on here on vn with previous versions of this board so the ether10 in particular what would happen is that v in is powered which is this net here that diode didn't exist and v in could be fed back through here and if you had this jumper in place and you plugged this into an ethernet switch that didn't have an injector what you would do is end up applying your 12 volts back down this line here and to here and you would potentially back power to ethernet now there is something weird going on you see this line here v plus comes in here and this is a little schematic of the internals of the rj45 socket because this socket supports power over ethernet it's got these little rectifier bridges built into it and it's got these little transformers as well in theory if you apply plus 12 volts on this it should be stopped by all of these diodes you can see that there are four diodes here that are facing in towards this point and it shouldn't be possible to back power the ethernet by bridging this across but there is definitely something weird going on i've seen this myself i've had other people report it to me that this is a potentially dangerous situation because you can end up with power going back up your your ethernet connection and possibly damaging your ethernet switch and i've actually seen it i did a test where i put jumpers onto this header i put a 12 volt power supply onto the board and connected it to my little desktop ethernet switch and the ethernet switch got stinking hot and stopped working luckily it started working again a few minutes later after it cooled down when i disconnected everything but even though the internal schematic of this rj45 socket says that power should not possibly be able to come in here and go anywhere else there is still something going on inside the rj45 socket that allows that to happen so that is why i added this diode even though it should be redundant anyway now we have multiple ways of feeding power and it should be a little bit safer all right let's have a look at the board and see how it came out so there is our little tiny board in our oversized reflow oven now i did see there was a question earlier about hacking toaster ovens and yes that's what i did for many many years before i had that oven in fact i still have a toaster oven sitting out in the garage with a pid controller attached to it so i've got a thermocouple there's an arduino which is running and it's got a screen on it so it does the proper reflow curve it brings the temperature up to the soak temperature holds it for a while goes up to it does the spike brings the temperature up to reflow and then the cool down and i've even attached it to a servo so once it gets to the appropriate point in the cooldown cycle it uses the servo and pulls the door of the toaster oven slightly open so that the airflow can get through when it cools down more quickly but i stopped using that when i got this one but if you're thinking of doing that check out the reflow master reflow master pro i think it is which is seon's project to make a controller for reflow ovens so if you're thinking about doing a retrofit to like a cheap toaster oven then have a look at seon's reflow master oh philip uh pointed out i keep saying rj45 and he said ap8c not rj45 okay so i was using the um the lazy terminology so the name of that socket technically i think is not rj45 all right let's have a look at this under the microscope and see what has happened see how much of a disaster we have uh okay first thing i'm gonna look for is solder bridges and weird things going on now if we look at these you can see the the resistor networks here actually fairly clean they've ended up reflowing nicely i think i can see a tiny little solder bridge just in here and i might have to do a little bit of spot cleaning up yep there's a solder bridge just there i don't know if you can see that hang on i'm gonna try zooming right in unfortunately the view with the camera is not as good as the view through the optical eyepieces so i can actually see this a whole lot better than you can and i'll try turning i'll adjust the brightness a little bit which way that way yeah so to me it's very clear that there is a little solder bridge between these two pins and you could probably just barely see it right there okay now that is um a little bit too zoomed in to get context so i'm going to zoom it back out again so what i should be doing is fixing these as i find them not just moving on and then possibly forgetting about fixing it so i've just got a little bit of solder wick here i'll just pull that one out and pull out that solder bridge go on uh oh look there's another one up here just up in the top corner i'll lift that one out gone and the rest of that chip looks okay so coming up here i don't see any major problems uh the atm mega is such a coarse pitch that there's um there are no problems with that nice older bridges there that looks okay there are some little spots of solder you can see like just on the top of this led there was a spot i think that was put there by the nozzle button seems okay crystal looks alright coming across here that all looks okay the other crystal looks all right the 16 u2 looks okay like sometimes you can see if there are little bridges you can see the bulges on the outside of the chip are joined i probably put just a touch too much solder on these pads and you can see it's bulging out as the chip has squished itself down the solder has bulged out but it looks acceptable uh looking across the back of the usb-c connector that looks acceptable oh look this has skewed itself this was actually straighter when it went into the oven and it ended up twisting itself pulling under surface tension probably because i didn't put the solder paste on straight i just vaguely squirted it onto the pad now one thing i often do with these connectors is try to look down into them so i'll pull the microscope off the bench so i can get the right angle and look into the back of it and you can see that there are no visible bridges there sometimes there are bridges between the pins down inside it where the case is hiding it from you so if you look down the back like that you can often pick up problems that you can't see any other way but that looks fine what else coming down here caps are attached properly voltage regulators are on oh that solder joint looks a little bit iffy it hasn't wetted properly and this is because i'm using crappy parts that i pulled off another board and i didn't have i didn't use flux or anything i just i didn't clean the parts up i didn't do anything i just pulled them off the other board and stuck them on which is not ideal so all right come on get some heat into that nope better still let's get some flux onto it [Music] smoosh smoosh smoosh on the flux all right and yeah the solder is not taking to the the pin very well i don't think this really matters because there will be an electrical connection there i'm putting a lot of heat into this because this is connected to the ground plane this particular pin and it's um a lot of thermal mass there those are crappy looking joints disgusting oh well it should be enough to test the functionality of this board at least uh now what else well look i was a bit enthusiastic with the solder paste and it's squirted some little solder balls outside now next stage with this you can see that it doesn't have the the ethernet socket on here it doesn't have the dc jack and [Music] doesn't have any of the headers obviously either but we can add those if we need them don't need them for testing though so what i'm going to do is do some very basic functionality testing using uh well first thing i'm going to do how can i show you this i need to get another usb cable where's the usb cable right here i need a usbc cable so that i can power this up in view of the camera switch to overhead and can we get some zoom on it how far in can we go and still focus approximately there okay so what i'm going to do is power that up and it hasn't exploded which is good and in fact you can see oh i know what i need to do i need to put this other before i get to that though i need to put the other crystal on there but before i get to that i'm going to do some just basic functionality testing on this you can see that we've got the power led on and you can see that this is actually running blink that's because i pulled this atm mega off another board and it's already got that sketch loaded onto it so what i'm going to do now is pull up the arduino ide and just try loading i'm going to do the blink load and let's see what happens um just so that you can't see what's coming up in my ide i'm just switching to the front camera temporarily because i think the ide is going to open up with some projects that i was working on for a customer and they're yeah i can't really show that because it's under nda so um i'll have to get those off the screen so okay in the meantime while the arduino ide is opening it's still trying to open all right back to comments uh okay so andrew brown asked how long do you have before the solder paste goes off that depends on the solder paste for it's in the order of like as a very hand wavy number it's in the order of hours so it wouldn't be a great idea to put the parts on and then leave the board sitting on your bench and come back the next day and reflow it it would probably work but it wouldn't be ideal because the carrier for the solder paste would have dried out would evaporate like the carrier would evaporate out of it and so yeah a few hours is fine uh generally though what happens is you end up putting the parts on put it into the oven and it might be on the board for you know 10 minutes or maybe half an hour before it goes into the reflow oven that's pretty normal sort of time okay so just got to quit out of a few things here [Music] and that's better so now i can switch back to my desktop view all right you can't see anything that you shouldn't see link status mac address demo all right well this actually this is a um a reasonable thing to try so what this is going to do what this particular program will do is talk to the mac address rom which is at address hex 50 and it will read its value out i'm just going to change that to 1 1 5 200. it gets the mac address and it just outputs it onto the serial port so i'm going to set the come on just trying to open the tools menu uh okay i'm going to set the board to arduino uno because essentially that's what this should look like and go to pick a port the port is which one serial 14 310 that would be it so now i'm going to load the mac address demo sketch onto it hang on there's something weird going on here i think whoa i think my board is not working properly i do have some fail on it that's interesting so i was just powering it off usb and the part of the board has got warm hang on i don't know what's going on with that i think there might be a problem with a voltage regulator on it so we might have some actual debugging to do which would be interesting problem uploading to board yes come on okay yeah after the board has been running for a moment it gets around the voltage regulator it gets very hot and which v-reg is that that is the 5-volt regulator that seems to be getting hot i wonder if i've messed that up that's the one that had the really dodgy solder joints on it all right so before getting too much further into this what i'm going to do is try powering this off a bench supply so i'm going to stick i'm going to solder a socket onto it just for convenience so i've got a oh what how has the time gone it's already 11 45 what happened these things always take longer than you expect they will all right so i'm just going to stick on a 2.1 mil socket normally i'd be turning on my fume extraction for doing this but fume extraction is noisy a lot of the um a lot of the things i do in here i have noise cancelling headphones on while i'm doing it so it's not great from a live stream point of view okay lab power supply time and let's see what's going on i want a lead for dc jack let's switch cameras so that you can see something a bit more useful than that uh overhead okay so that is the board currently plugged into the bench supply which is currently turned off and what am i going to set it to i'll set it to 10 volts and i'll set it to 300 milliamps and turn it on and it's whoa that is i set 0.3 enter that is not happy it's hitting the current limit which suggests that there is a serious problem on that board interesting there are many possible reasons for this so i think what i would do next is try floating this back off take the um the 5 volt regulator back off it and see what happens when i'm powering with regulated 5 volts into the 5 volt rail at that point all right any other comments oh wow yes i didn't realize that i needed to scroll in here i'm way behind on comments okay okay so adam said a few weeks ago i said i was considering revising the timing of the live stream if i had any further thoughts on this my only thought on that is that i'm putting that decision off and i'm just sticking with the current timing for now so james said what is the bomb cost to add the micro sd card slot the card holder itself is not particularly expensive i think it's in the order of 50 cents u.s give or take and then there are a couple of other supporting parts because an sd card is an spi interface and it needs the it's 3.3 volt spi which means we need voltage dividers so we've got two resistor networks to do the voltage division for that um so basically the cost of adding the sd card is the cost of the carrier which is 50 cents plus a couple of supporting parts so it's not all that much yeah so station 240 said you can get 24 hours older paced be many to get some myself for a board with 100 parts um yeah oh steve pointed out do you have a thermal camera good for finding hot spots i do it's right here that's a great idea what i could do is power this up and see where the board gets hot and then that is probably the problem that is such an interesting thing to try that i think i'm going to do it right now but uh let's see so philip said have i had a look at the grid connector taiwa 2s line of products from bunnings yes i have i have some right here which i bought for the purpose of pulling apart and showing doing the um the reflashing to tasmota i just haven't done that yet uh yeah so oh and in relation to station 240 thing about 24 hours older paste yeah there are a few different types of solder paste you can get solder paste that is called thermally stable which means you don't need to keep it refrigerated i have a little bar fridge in so that door there the white door that goes through to a little bathroom and you would have seen that if you saw the lab tour a couple of live streams ago and there is a bar fridge in there and i have a plastic container where i keep solder paste so this solder paste that i used is the this is not thermally stable solder paste so the thing is that thermally stable solder paste doesn't that doesn't mean that you can expose it to air for a long period of time what that means is you don't need to store it in a fridge although it is still better if you do so there are a couple of different things one is storage temperature so with different types of solder paste some you refrigerate some you don't need to and the other is its lifetime once you have put it onto the board before you reflow it and that's what that 24-hour solder pace is referring to so with that it's all the pace that you can put onto a board and you can you can leave it up 24 hours before it goes into a reflow oven and it should still perform properly with most solder paste it needs to be done within you know an hour a couple of hours so and if you've got a very large number of parts to place as station 2 have already said like a board with 100 parts on it it can take hours and hours to put all of those parts on by hand so having solder paste that you can leave exposed to the air for a long period of time is useful [Music] all right so what was i going to do that's right thermal camera let's see i'm going to get some more i haven't used this camera for quite a while so i'm going to let's see if this works and i'll switch to the proper view in a moment but first i've got to find the camera when i put it oh there it is all right so now i need a [Music] cable for it there we go and so camera is plugged in is it working i can't tell i have a feeling that this camera may need to be plugged directly into the computer instead of into the usb hub there was some oh no there it is there was some weird thing about it last time yeah there we go all right where's the board oh i can't even see it because it's the same color as the background all right so now i'm going to figure out a way to get this into a position where it can be seen and then i'm going to switch my view to showing my desktop which is where the software for the thermal camera is coming up uh now when i've done this in the past what i've done is used like a rubber band and attached this to oh let's see if i can do that again i'll just do the same thing grab myself a steel ruler and put a weight on it what can i use as a weight how about a roll of solder that will do and now i need a rubber band this is really high tech as you can tell in fact i'll get a couple of rubber bands i'm gonna put one on there and then attach this camera with the rubber bands come on and slide it along all right now that actually works out okay all right i'm gonna switch to your view now so at the moment i've got the thermal camera in fact let's hang on did i just see a super chat or something flash past i'm going to show you how dodgy this setup is before we get into using it camera poker phone let's see here we go hang on has that stopped working i think my pocket phone has gone to sleep can i wake it up come on come back i might have to reconnect to the i'm going to switch it to the desktop while i screw around with the settings for the no i can't i'm going to switch to that in order to select that oh i know why it'll have failed it's because it'll have switched its wi-fi network so one of the problems i have at the moment is i've got multiple wi-fi networks and [Music] if i'm not on the right network the camera doesn't work come on nope okay my droid cam is dead i was wanting to show you the view 252 deactivate activate no droid cam has failed so i'm going gonna switch back to just using okay the overhead view and desktop all right now can i make this bigger what you're looking at here is the software for the thermal camera and you might think that the purple patch at the bottom that you can see is the uh is the board that we're wanting to test but it's not that is actually just a reflection of one of the ceiling lights and if i stick my hand in here you'll see it come in looking very hot so that's my finger and this here is the uh the ethernet prototype and if i get the angle right you'll see that a couple of bright spots appear on it that's because those are metal parts like down here we've got the sd card slot which is silver and if i move it at the right angle it's acting as a mirror and the thermal camera is seeing the heat of the in that case i think it's the skylight so i'll just move it like this and i'll plug in the power jack so it's now uh i've moved my hand out of the way which means that it's auto scaling i'll just move my hand into it again oh well the auto scaling will fix as soon as it gets hot so i'm going to turn on the bench power supply and we'll see what happens see if there's a hot spot oh look there's a hot spot you can see the two voltage regulators that one oh and just beside the voltage regulator what is going on there i've got to find something to point with that hotspot there ah that is the diode that's interesting all right so this is the voltage regulator see that purple spot there on the right the big one i'm going to put my finger over it [Music] it's barely noticeably warm like with my finger i'm having trouble telling that that is actually getting hot but the thermal camera is sensitive enough that it picks it up so let's have a look at the actual temperature if i move the spot temperature over that point it is 37 38 degrees over the rest of the board it's 23 24 degrees around here 24. let's have a look on the diode so on the diode it's 40 degrees cool um and if i okay i'm going to try floating that diode off the board oh okay so pancake legend yes thank you for that super chat that's uh that's very generous so james said what thermal camera is that it's a pure thermal lepton something i can't remember i bought it a couple of years ago from group gets or somewhere like that i'll look it up in fact i'll stick it on the gear page the um i'm going to drop a link in oh electron ash say diode backwards no i don't think that would prop that would stop it uh gear all right so i've just dropped a gear link into the chat and what i'm going to do is after the live stream like sometime today i will look up the the specs of the camera that i've got and i'll stick it on that gear page because it's actually a really useful tool great for things like this now i'm really thankful to to steve for suggesting this because often i have this camera sitting here like it's sitting right here on the shelf above me and i don't even think of using it but it's great for this sort of purpose so what that's done is narrow down that there is a lot of heat being emitted around the voltage regulator and the diode so there is possibly some problem with that i've got a bench supply plugged into this and i've set the current limit to 300 milliamps which is enough that it's not it's going to show up thermal issues like this but it's not enough that it's going to make smoke come out of the board probably so when you're bringing up a board like this for the first time this is actually a really good idea a good thing to do make sure it's plugged into a bench power supply and set the current limit to a sensible value probably high enough that the board will power up if it's working properly but not high enough that it's going to [Music] it's going to go up in flames on your bench all right so that is very interesting to see well let's take that off i'm going to switch back over to the overhead where are we overhead view um overhead view and remove this out of the way what i really need to do is make a proper mount for that thermal camera so i can put it into ah hang on let's look at myself there's me on the thermal camera and you can see the glasses thermal cameras are fun so i'm going to turn off the bench supply and unplug this so what i'm going to do first off is just flow off that diode now if that diode was backwards in theory it shouldn't have any effect excuse me where's my hot air there it is for this one i'm not going to bother using the preheater i'm just going to butcher this chuck it on a piece of wood and warm it up i'll start with the air a fair distance away so that it can warm it up gradually and bring up that part of the board it looks like i've got the hot air like right on the pcb the view that you're getting from the camera but it's actually a fair distance up above the pcb warming up that whole area so what i my general technique for this is to warm up the area and then to move in and heat up the specific target part more oh there it goes it came off uh more directly and now that board is going to be stinking hot ah let's chuck this back under the thermal camera before it really cools down oh look we can see there's a hot spot that whole part of the board that i heated up is now very hot as you would expect i'll just align it there the camera is on a bit of an angle put my roll of solder back on here to hold it all in place and so now we can see how hot is that part of the board right now so around there is 90 degrees and falling you probably have a bit of trouble seeing it in the in the screen but up in the top right here there is a little temperature reading and that is the spot temperature which is where i've clicked the cursor into the image for the thermal camera so it's 76 75 degrees so if i click around different parts of the board that's 54 degrees that's 46 over here towards the corner it'll be much cooler 25 degrees so it's only you know barely above room temperature over here on the side up in the corner furthest away from where i just did the reflow so oh this is quite spooky and what i'm going to do is leave the cursor there and i'm not going to move it and you'll see the hot area slowly shrinking away from the cursor i think well actually what's happening is that the the camera or the camera software i'm not sure if it happens in the software or in the camera is adjusting its gain and you'll see that what happens is that it tries to adjust the gain so that whatever is the hottest part of the image is going to show up the brightest so even though the temperature is falling the general pattern is kind of staying the same until it all sort of evens out but if we come down here now that's only 50 degrees and that was much higher than 50 degrees now it's 40 nine and falling but you can see that what's happening is that it's evening out a little bit instead of being a very specific hot spot down here the heat is sort of spreading across the board so thermal cameras are cool all right let us try now we're not going to be able to tell much from the thermal camera because it's preheated by that reflow but what i'm going to do is chuck the lab supply back in turn it on and i'm going to see if it hits the current limit and it doesn't that is interesting so it hasn't hit the current limit uh it's in fact it's not drawing any oh i know what okay i was looking at the wrong diode i thought this was the diode that i had just added and it wasn't so this has led me on a bit of a wild goose chase what's happened is the there is something on the board that is short-circuiting it may end up being this regulator now so the next thing is i'll float that off but what's happening is that i actually know the actual short circuit is probably downstream of the five volt regulator so what's happening is power comes in from the dc jack it goes through that diode from there it goes into the input of the 5 volt regulator and then that gets converted to 5 volts and then the output goes to the board so somewhere on this board there is a short i think on the 5 volt rail and that is what's causing the regulator to basically just draw as much power as it can and that was then causing that protection diode to heat up so right now i've got power plugged in it's a bit hard to see on the thermal camera maybe i should switch back to the normal one that we've got power coming in on the dc jack and where are we overhead okay so we've got power coming in on the dc jack and the diode which was just there was overheating because the output side of that diode so the cathode was going to the input of this which is the 5 volt regulator and presumably the 5 volt regulator was working its butt off trying to provide power into a short circuit so somewhere on this board there is a short circuit on the 5 volt rail that's my current conclusion all right so next thing i can do is try putting power onto the output of the regulator from the lab supply at 5 volts and we'll see what happens that'll bypass the lab supply and tile that'll bypass the onboard regulator and tell us if the short circuit really is downstream of the regulator so which is the um the pin for that it is okay so 5 volts comes out of the tab and it goes that is really weird where does it go oh there are a heap of fires there or vias however you want to pronounce it um so where do they go sorry you can't see what i'm looking at right now because i'm looking at how this is at the connections for oh you can see it but it's only small so i'm looking for the connections on the 5 volt regulator now if i switch back to desktop so this is the footprint right here for the 5 volt regulator and over here is the diode so we've got five volts coming in on here from that dc jack goes through that diode heading from the anode side here through to the cathode and then that goes to the v in rail v in goes onto the input of the 5 volt regulator and then the 5 volt regulator also has a ground connection and it has a 5 volt out now you can see from the part that's highlighted there is the tab of the regulator that connects through here goes to one of these capacitors and there are also a whole bunch of vias now you can't really see them very easily on this view but like if i highlight one of them you'll see it so ah get rid of that so there are there are a heap of where is it i've got to highlight one there okay so there is one there and there are there are two rows of them down here to stitch the two sides of the pcb together so the 5 volt output from this regulator goes down through all of those vias and you can see this big blue area on the bottom that is the the 5 volt rail on the bottom of the pcb and you can see that it comes through here unfortunately it goes through a veer to the top of the board comes up here back to the bottom of the board again which is a really bad practice but that's just the way it worked out on this board and then it comes over here and the 5 volt rail is used for other things so there could be a short in a few places in fact let's highlight the 5 volt rail so 5 volts is going to the atmega16u2 which is the usb interface chip up here there could be a short for example if there was a short here between this 5 volt pin and the ground pin that is right next to it that would explain this problem and given that it's a qfn and i hate qfns i want to blame that but there are a few other places that it could be as well so what i'm going to do first is just to verify that my theory is correct i am going to bypass the 5 volt regulator and feed 5 volts straight into it into that header now where can i pick that up conveniently it's on the tab it's on oh okay it's on the input side of this why didn't i just put a track across there i've got no idea that seems like a silly thing to have done i could have directly connected the input of the 3.3 volt regulator across the top of the pcb to the output of the 5 volt regulator instead of relying on passing it down to the bottom of the board and then back up again that's a silly move all right so this track here this pad is the 5 volt rail and ground is over here which means i can very easily connect the lab supply onto the 3.3 volt regulator and use that to inject 5 volts and see what happens so let's give that a shot i'm going to switch off switch away from that and what i want instead of the uh oh these ones are big horrible connections oh well on the lab supply i'm going to switch away from that lead i'm going to put in place a oh no those aren't going to work i need different leads where are my other leads these ones no i need leads for the lab supply that have alligator clips on them this one will do and i need to set the power supply to 5 volts so i'm not injecting 12 volts onto the 5 volt rail which would be really really bad not the ideal thing to do now these alligator clips are quite large and horrible i'm kind of inclined to yeah that'll do all right so what i'm going to do is just connect where will i do it i'll find somewhere that i can connect ground this one that'll do ground can go on there and then i'm just going to probe briefly anyway this doesn't have to be particularly stable because i'm just going to touch this onto the the input of the regulator all right so the lab supply is turned on as a sanity check i've got it set to 5 volts 300 milliamps ground is connected and i'm going to feed 5 volts onto the input of the 3.3 volt regulator which is the 5 volt rail so let's see what happens okay it is well you can see that the led here comes on very briefly power led as i touch it i don't know if you can see that on the camera and what's happening is that it the voltage briefly comes up like by briefly i mean it's probably milliseconds and then the lab supply hits its current limit so it's running at 300 milliamps right now all right so this is interesting because what we can do now is maybe use the thermal camera and see if we can this might help us to track down where the short circuit is i'm going to get the whole board back under the thermal camera i'm running way over time i need to go and have lunch in a minute uh what am i doing okay so switching to switching to the thermal camera view and the desktop and let's get this oh come on has the thermal camera stopped working first my mobile phone stopped working and now the thermal camera has stopped working so i'm gonna quit out of get thermal launch it again and see if it can see the camera that's better i even aligned it perfectly cool all right so now we have the board and you can see what looks like a hot spot i don't think it is there is a spot on the board just here it's only 25 degrees so this whole board looks like it's glowing hot now it's not that's because of the automatic gain control on the thermal camera and the hottest parts of this board now are 25 degrees it's like half a degree above ambient and that'll just be some residual heat so what i'm going to do now is touch the 5 volt input onto that point and we'll see if any part of the board gets hot oh look at that there that is getting hot that is the voltage regulator that's actually the 5 volt voltage regulator that got stinking hot when i applied 5 volts onto its output that should not happen that's a linear regulator which is normally perfectly happy with um with having power applied to its output like being back powered so i have a suspicion that the problem is just that that regulator is booked but then that's the one that i was messing around with earlier anyway remember i got the soldering iron onto it and held it on for ages i was just doing generally nasty things to it so i'm going to flow that off the board and then do one final test and then i think i'm going to have to end the live stream because i'm going to have to go and get lunch but at least this way hopefully we'll verify that that regulator is in fact the problem so come back to here and let's hit it with some hot air now because now this is basically the worst part on the board to try to get off because it's physically large it's got a lot of large thermal mass and it has that it has that tab which it uses as a heat sink so the way these are designed to be used is they use the pcb as part of their heatsink and that's one of the reasons that there are so many wires running down both sides of it to stitch the two sides of the board together it's so that if the voltage regulator is trying to dissipate heat that heat goes out through the wires to the bottom of the pcb and there are like 30 or 40 wires right next to it oh there it goes to come off which help to wick that heat down in fact you can probably see them if i chuck it under the microscope while the board is cooling down because i want to let the board cool down for a moment before i connect power again because if we use the thermal camera to see what's going on it's um it's no good if the board is still stinking hot from being reflowed so you can see here the wires that are used for thermal and electrical stitching this is the position where the voltage regulator is that i just pulled off and you can see there's a row of wires here there's another one just inside it and then two rows over this side and on the bottom of the pcb is a actually something i'm noticing now which is interesting uh it's like there's a um the way this board is made it's there [Music] i don't know um it looks weird to me the um the solder mask here is different to the salt oh no they've done it on all of eyes that's okay you can see that there are these wires here and these fires here it's like the solder mask is matte over the area of where the wires are and down here as well even though this is all one big piece of copper it actually looks almost like that's a separate track but it's not it's just a weird artifact of the way this board has been manufactured all right so let's see if we still have a short circuit even though the voltage regulator has been removed uh overhead yep all right so let's do that same thing again i'm going to connect ground to this particular point which is the ground point and i'm going to turn on the lab supply and just because it's a cool thing to do i'm going to stick the thermal camera back on it this is horrible i've got my little plastic container here with a ruler and a bit of a roll of solder as uh as physical ballast to try to hold it in place i really need to make a proper mount for that thermal camera oh is that running yes okay so i'll switch you back to the thermal camera view and i'll drag that out let it zoom the resolution on the thermal camera is actually really low it's i can't remember what it is but it's like 160 by 120 or something like that but it's surprising how much information you can still get from a very low resolution image like this all right so lab supply is on and i am going to touch this to the 5 volt rail oh look at this now the board runs um the interesting thing is that you can't see it on the thermal camera those leds are blinking like blink is working on this and i um yeah it's kind of funny that you can't see it the leds the power led which is up in this corner up there is on and the the d13 led just below it is blinking i can see it really super clearly with my eyes but you can't see it on the thermal camera because the leds aren't getting hot now you can see that there is a little patch warming up just there to the right you can see that little square that's starting to get warm that's the atmega328 so the microcontroller is getting very slightly above ambient temperature now that chip there that is the wiznet chip and it's getting warm now it it's actually not getting as hot as it looks because of that automatic gain control that's probably only a couple of degrees above ambient but i think what this proves is that the voltage regulator that i stuck on this board was busted so what i'm going to do is grab a proper voltage regulator a new one instead of a nasty one that i have ripped off another pcb and i'm going to put that on it okay oh electron said on the cap to the left of the v reg pad it looks like maybe a solder short to the ground plane that's an interesting observation i didn't see that so i'm going to have a look at that maybe the problem was not with the voltage regulator itself but with the capacitor to its left ah yes i see what electron ash is pointing out but that is not the problem uh yeah that's a good observation and that sort of thing can be a problem now i'll see if i can point it out so you know if if you spotted this through this view you've got good eyes like tron ash the point of interest is right here you see that little thing just next to the capacitor in fact i'm going to zoom right in and see if we can see it a bit better where is it about there might need some more illumination okay what it looks like right there is potentially a short to the ground plane with a little stray bit of solder but and it's it's hard for you to tell through the camera but looking at this with my eyes i can see that that is not actually a problem i can flick that and the solder mask is unbroken the solder mask all through here is nice and clean you can see there's a silver ring around it that's just light reflecting off it from the ring light of the microscope but there is no i'll see if i can get the angle better there we go you can see that there's nothing there that's making contact but that was a good observation that could well have ended up being the problem so thanks for pointing that out no i think the problem is just that the oh electron said no i was talking about the electrolytic but i think it's just where the pad was altered earlier yeah okay so uh i'm going to i'm really over time now i thought that we'd have this all up and running and then i'd have done the tests on it by this point but we didn't get that far not in this live stream anyway but it is running it's running blink now when i power it up from the external supply without the onboard voltage regulator so yeah i'm i'm fairly confident that if i stick another regulator on to it and not that super dodgy tarnished one that i pulled off a busted board it could have been busted in the first place that's the thing is these donor boards that i've had sitting in a container and i've been pulling parts off the reason they're sitting in that container is they were busted in the first place so there may well be parts on them that are no good and uh with boards like that usually pulling things i normally do things like pull resistor networks off some and you know passives like inductors and other things that aren't very likely to be a problem but sometimes things like voltage regulators or ics they could be the thing on the original board that was dead so who knows that board might have had a reverse voltage applied to it at some point or something happened that fried the regulator and that's what has now caused a problem here so i'm in two lines now about whether i put another regulator onto this and try it [Music] do i don't i i think i might and try it before i end the live stream because you probably want to know whether it's working now i'm going to need where do i get one from i don't think i have one of those regulators new and the regulators have already been stripped off all of these other donor boards oh no here's one no that one that one's got a switch mode regulator on it so uh you know i'm going to switch this camera to autofocus even though autofocus drives me nuts it's um it tends to hunt and seek on these panasonic cameras uh regulators regulators where do i get one regulator's mount up all right i might have one in fact i probably do have one in my big box of random cup tapes all right if you're all sticking around and you feel like it why not keep going just for another five or ten minutes i'm going to find a voltage regulator oh there's a 3.3 volt regulator no that's not the right one le triple one seven d pack five volt that is that's the gold that's what i want so since we got this far let's put another regulator on the board and see if it works so brand new regulator straight out of the tape and if this now has the shorting problem again then there is obviously something else weird going on but if this fixes the problem then well i can get on with testing the other parts of the pc like other functions of the board but i won't do that right now on the live stream i am just going to get past this power problem and then hopefully that is uh that's it hopefully the rest of the board will work but in any case i will stop and go and have lunch okay so i've got a five volt regulator on there now what will i do with that diode let's put the diode back on it because the diode was not the problem as it turned out the diode was well it was kind of a red herring but it was it was a symptom it wasn't the cause the diode was just doing its best to pass all the current that the voltage regulator was trying to consume which was all of the current every bit of it that it could get so it's an awkward spot to get to because these capacitors are in place i could do this with with hot air but yeah the iron is hot and it's right here come on that's it that's float and i'll just solder the other end of this diode come on get some solder in there there we go okay so hopefully we will now have great success get that out of the way switch to the dc jack connections on the lab supply i've got to switch the lab supply back from 5 volt so i'll set that to let's make it 10 volts again still set to 300 milliamps come to the overhead camera come on where is my cursor i'm going to find it on all these screens there we go okay so now plug this in and i'll turn on the lab supply and we'll see if it hits the current limit and it doesn't yes victory oops i'm getting too excited bashing the camera all right so because i have the thermal camera here set up and it's um it's kind of interesting to see let's switch to the thermal camera and just watch it in normal use so you can see now that the this part here which is the wiznet chip is starting to warm up you can see that the voltage regulator down here is warm which will still be some residual heat because i just had the soldering iron down in there now let's check the temperatures so the hottest part of it now is 37 degrees 39 degrees and it's jumping around a little bit 40 degrees it's slowly rising so that is the temperature of the voltage regulator which is now doing its job so now we have a board that is working it's running blink and okay one last thing just so that we get some success beyond blink before we end the live stream i'm going to switch back to the overhead view maybe it's out of here okay so we now have and you can see on this camera that blink is working so the little red led there is flashing so what i'm going to do next is try loading that sketch that i was trying to do before and i couldn't get it to load now i will switch you back to desktop view in a second and what have we got uno yeah we've got the serial port selected so we're back to okay so the arduino ide and what we're going to do is try uploading this sketch to it compile compile upload upload it's uploading is it going to work come on you can do it if this doesn't work it's probably a problem with that 16 u2 usb serial chip so rather than diagnose that problem what i could do is just put this in oh no i can't do that all right that is let's even see if it's showing up i'm just going to um you know why i bet you're all shouting at me right now in the chat that i didn't plug in the usb and you would be right so i'm going to unplug that and plug usb into it okay so now it is plugged into usb and it is a running blink so i don't know what board i was just trying to upload that sketch to i've got some other there is some other arduino plugged into a device somewhere here that i was just trying to upload this sketch to i think it actually might have been an esp8266 all right compiling again and uploading come on you can do it this time oh yes you can no i'm still not responding well that's annoying all right i've got another problem to diagnose i'm just checking my settings on here so i've got so there's an uno i've got the serial port selected and etc um [Music] i have a feeling that i may have pulled this processor off now what did i get it from i pulled this processor off some donor board that may not have even had a 16 u2 is it serious speeder serial converter so i may just need to reflash the bootloader onto it so that it will talk properly to the 16 u2 um but this is kind of getting into fiddly territory uh first though i'm going to verify that i do have the correct serial port selected so under tools port 143401 i'm going to unplug that and check my serial port list again and that port is gone so it is definitely port 14 for where was it no it's not back yet come on give me a uh there it is 143401 all right oh jerry l42 uh thank you very much for the the super chat um now let's see compiling let's see what happens do we get anything uploading no this is going to fail so excuse me a second okay it's definitely time for me to stop the live stream all right so we diagnosed the power supply problem and i think that that is probably okay so this problem with uploading a sketch to it is probably related to the 16 uto usb to cereal chip so um okay i'm not going to do this now in the live stream i need to finish this and go and have lunch but just to tell you what i will do from a diagnosis point of view to actually get past this next stage is i will connect an icsp programmer to the board and i will reflash the bootloader onto the atmega because it could be that this one that i pulled off the old donor board is so old it's got one of the really old boot loaders on it from back when we were using the ft-232 rl usb to serial chip so we're using that old old thing way back before switching to the 16 u2 and so that will at least verify that the processor is working properly then i will test again obviously if that doesn't work what i will do is get a usb serial converter module so i've got things like let's switch this back to here so i've got things like this usb serial converter and this has got a 16 u2 on it as well so what i'll do is connect the data pins from this onto the data pins on the target board essentially i will bypass the onboard 16 u2 and see if i can talk to it using this if i can that indicates that the problem with communication is with that part of the circuit so it's with the 16 u2 now that part of the circuit hasn't actually changed since the previous prototype it's the exact same usb to serial converter that was on the previous revision that part of the board hasn't changed so if that is not working it's probably just an assembly problem in which case if i just hit the 16 u2 with some hot air and maybe some flux and wiggle it around a little bit get it to reflow and reseat it'll probably start working and it may be necessary to replace it because once again i pulled that 16 u2 off another donor board and i'm being kind of silly just pulling these dodgy old parts off what i should be doing is using new parts that i know are okay so if that's if that continues to be a problem i'll just replace it with a new part which i do have somewhere here in a drawer somewhere and that would should get usb to serial con working i'm confident that i can do that so i'm not going to bother going through all of that right now on the live stream and keep you hanging on for another half hour while i mess around with it then after that what i will do is test that the um i'll get back to this test i was just trying to run which is that the mac address chip works once again that part of the circuit hasn't changed so that should be okay then it gets to the interesting stuff which is testing the wiznet chip and that is really where the major change has happened with this revision of the pcb because i added those bias resistors to the mode pins and so i need to make sure that's working and there were a couple of other little hacks to do with the way the data pins were connected that i changed and i i manually patched them on the previous prototype and then applied those changes to this one so at that point and there are other things like the sd card i'll need to test there are a few just general things to test but anyway you can see how the process is going once i fix the usb serial it's all just a matter of working through each of these things one step at a time and taking it carefully enough that hopefully i don't blow up the board in the process so um oh frank thank you very much for that um for that 20 super chat esp flasher where is it where is it here esp flasher is complete so that's the esp flasher and that's the esp flasher adapter not the greatest camera to be showing you this on but this gives the 1x6 0.1 inch format header plug it into this and now it's got 2 by 3 idc headers in both 1.27 millimeter and 2.54 millimeter and also down the bottom it's not populated 1x6 in 1.27 millimeter so i've been using this with the tiny little 1.27 millimeter pitch idc cables on uh on a few projects now over the last two weeks i've brought up several new boards with this connector on it and used this programmer and it works brilliantly so yeah it's really cool having it board with an idc header for the programming connection because it means that you don't have this plugged into it and then a bulky usb cable that sort of drags the board around on the bench and it's not very convenient using this combination i actually really like this so 2 by 3 idc header 1.27 millimeter pitch so the whole thing is very very small you just have this sitting on your bench plug this into your target board and then you can do reflashing so that is working out really well ah seon thanks for running for hanging around and uh thank you everyone else for hanging in so oh yeah so frank said in the shop yet uh yes i think so i kind of let me just check i did a espf i think that url will get you there um i've been kind of low-key about it because i still need to do more documentation and other things in relation to it so espf that is where to find it and i also need to still put up the flasher adapter which is a little add-on board that gives you all of the other breakouts okay um [Music] one final thing so um sharon ap said which software are you using for a pcb designing and i'm using eagle primarily at the moment that's my main thing but i'm going to bail now because i have run super overtime what it's heading towards three hours it's two hours 45 minutes so far long long time lunch time i'm getting messages in the back channel from andy saying lunch time in bold and like with arrows towards it so uh i'm gonna go now but thank you all very much for coming along to another live stream i'm going to get this board going off stream and report back in discord on the results so if you want to hear more about this make sure you join the discord and i am going to hit end thanks to everybody and i will talk to you soon okay have a great weekend bye oh the voice from beyond uh john nelson asked i may have missed it did you add that crystal the answer is no i haven't added it yet i need to do that before i can test yours before i can test ethernet bye bye you

Channel: SuperHouseTV

Views: 8,978

Rating: 4.7272725 out of 5

Keywords:

Id: C1YeK3PdmVw

Length: 166min 34sec (9994 seconds)

Published: Sat Sep 26 2020