SDG #224 UniSolder 5.2 with a JBC C245 Handpiece

hi today we've got a pcb from jlc pcb and this one was actually sent to me by one of my subscribers anton and i do apologize anton it's a long time since she sent me these boards in fact it was back when i turned 10 000 subscribers and i gave away some jlc pcb vouchers he ordered some of these uni solder boards with those vouchers and sent a couple to me so really quite a long time ago now and i've decided to take a look at this project again now that i'm looking at all these soldering stations i thought we'd have a look to see how much the controller has to do with it and how much is just the actual soldering iron and the cartridges and this is part of looking at that so what this is is the unisolder pcb which is a project started by sparkybg and i'll put a link to the github repository in the description down below but this is designed to be quite a unique soldier 9 controller that works with a wide range of soldering iron hand pieces now it doesn't work with induction hand pieces it does need to be a standard resistive heating element but it works with those with thermocouples or resistive sensors those with the cartridges or with just the tips that you swap over onto the end of the soldering iron so quite a wide range of soldering iron compatibility you can see the compatibility list that has been tested so far so heiko stations the pace jbc hand pieces uh some of the weller ones and the ursa rt80 um and it probably worked with quite a lot more of these but what we're going to try and do is get this working with one of the jbc c245 handpieces so that we can compare it to an actual genuine jbc station and just like the pcb the schematic is also split into three pages to match those boards so we can feed it with 24 volts ac direct from a transformer or 24 volts dc it goes through a bridge rectifier it goes through a fuse and a tvs for transient overvoltage protection and then it goes into the various parts of the power supply circuitry here so this is the dc to dc converter that provides most of the power for the board so we've got a 3.3 volt rail coming off it and they've also tagged in a little bit of extra electronics here to give us a minus 0.6 volt rail for some of the analog electronics on the board all of this stuff up here is over voltage protection to clamp the voltage at i think it's 36 volt so that this chip doesn't get damaged if the voltage goes higher than that it ends up being dissipated in the transistor up here then we've got some mosfets here that are doing some of the switching for the actual heating element then we've got the schematic for the little oled board which is for a one to eight by 64 oled display it's got a little dc to dc converter on there for the oled power supply and then we've got the front panel pcb which basically has all of the analog electronics the 80ds the voltage reference all that kind of stuff in the thermocouple interface as well as the pic32 which is driving this whole thing also a usb interface which is isolated and also the seven segment displays if you choose to fit those so i do have a few problems with this project which is partly why i shelved it after soldering some of the resistors on so first of all the author chose not to create a schematic description which i think would have gone a long way to making this a little bit more accessible for people because there is a very long forum thread on the dangerous prototypes forum but it's spread over so many pages and the information probably is all there but if someone just collated all of that together and created a description or if the author had done that it would go a long way to trying to work out and explain how the circuit works and if you're trying to modify some of the firmware for example or if you're trying to troubleshoot your design just trying to understand where you're looking and what the problems are would really improve the accessibility of this project and the other problem i've got is the design of several things with the design so first thing is the bill of materials has not been optimized in any kind of way there's several different types of the same component and so you go through the billing materials and you order them and then you realize that actually you've got duplicates because there's three different types of 100 nfl capacitor for example it doesn't make sense um from that point of view and also the schematic has not translated well to the pcb and the reference designators are basically just all over the place on this board so when you are assembling this board up you're looking at the various part numbers in the build materials but you just can't find it on the pcb now helpfully someone has actually created an interactive pcb tool but really this shouldn't have needed to be done so this user and i will put a link again to this repository in the description down below has collated loads of information about this project to try and avoid you having to search the entire internet and work out exactly what's going on so some of this they've included in here or rewritten there's some details about the design and how to get it working there's this really useful interactive pcb map which if we go to that basically if you're trying to find where that part is on the board you can type in the designator at the top and it will say all of the other components that are the same so c4 is up here but for example you know you've got seven 10 microfarad capacitors and it highlights the location and more importantly the orientation because there are quite a few components on here where the silk screen is misleading and it makes it look like you should apply components in one direction whereas in fact they're in another direction and there's just pads that are too close together and not really any identification to show which way around those resistors go for example also a really useful part is equivalent parts in the biller materials there's a lot of components either that aren't available or the part number was incorrect for example the trimmer potentiometer had the wrong part number um but some of those components aren't available although they were in 2015 probably so this user has also collated lots of alternative parts that will do the job so it's if you are thinking about building this project it's well worth reading this page so let's get some of these parts onto the ptb so that's the majority of the components soldered on next we need to get the oled onto the little oled board so i think we're going to separate all these pcbs now you can see where we've got to cut at these little adjoining strips between the three boards so i'll just do that with some side cutters and then let's try and solder this oled up and what we need to do now is solder this oled so pin one on the oled goes the opposite end to the oled marking and we need to solder it so it faces up and then basically the display gets glued or double-sided taped onto the other side of the board so it's going to fold round like this now i am aware that some people have modified their uni solder to use a much bigger display for now we're just going to try and get it working with the original parts but this is quite a small display it's one of those 0.93 inch displays it's quite small so i think once we've got it all working we'll probably modify it with a bigger one and then we'll work out how we're going to mount this into a 3d printed case so i think the way we're going to attack this is we're going to put a little bit of flux across the pads uh we use this uh re-life halogen free flux because this one was quite easy to clean up and so we'll use that and then hopefully we'll just be able to drag solder a little bit across the pads to solder this oled in place so we'll just try touching a little bit of solder onto one of the pads just to try and get a mechanical fixing at the moment it's just held in place with a bit of tape probably hold it somewhere in the middle might be a good start so i've just quickly cleaned that up with some flux cleaner and then i've also stuck the old down with a couple of pieces of double-sided sticky foam and that's that module complete now unfortunately there's no keying of any kind on these pin headers so you do have to remember which way around to insert it into the pcb it's with the flat flex at the top so we'll insert this into here and i think that's the top pcb ready to go so we're getting pretty close to be able to powering this thing up i think what i'm going to do first of all is apply 24 volts uh maybe a lower voltage from the bench power supply and just check that the voltage regulator is doing the right thing and then we get the right voltages on the power going to the front pcb and then we'll be able to connect them up and then we need to try and get the firmware onto the unisolder board right so i've hooked up the bench power supply just to the dc input on the pcb so if you look you can either choose the dc input which just passes through one diode for polarity protection the other side goes straight to ground or you can use the ac input where it uses the full wave rectifier in total and i'm going to set the current limit to 50 milliamps i'm also going to drop the voltage down to 12 volts just to limit the amount of energy on the board because first of all i want to check that we haven't got any gross errors there's no components backwards which will draw a lot of current we don't want to try and blow everything up but also we just want to check the switch is working and this one's using an lm2675 you can see that the on off input which is to enable or disable the switcher that can also be used for under voltage cut off and that's not connected at all so this should work quite happily at 12 volts and give us the 3.3 volts on the output now the minus 0.6 volt output might not work properly as it is because we've got no load and it's tagged off the switching output of this regulator and there's not really much capacitance here so without any load on the 3.3 we might not get any switching action so that minus 0.6 volts might not really do anything but anyway let's have a look at what happens so 50 milliamps uh it's drawing 8 milliamps which is quite reasonable for what we've got in here so that's a good start and you can see we've got our 3.3 volts let's see if that negative output is doing anything so it was negative but it's creeping back up so let's put a little bit of load on the 3.3 volt output right so we've got a little led here it's a little bit fragile let's try and get this in here and then let's try to measure the voltage on the negative output and there we go so basically minus 0.6 volt so that all appears to be working okay so in the github repository there are some pre-built hex files that you can use i'm going to try and build this in mp lab because i do want to be able to modify the firmware and so you do have to install quite an old version of the xc32 compiler you can see it's version 1.33 but it appears to have all uh built properly so we're going to try and load this onto the pcb i've got my pickit 4 which is what we're going to use today and let's see if we can get this the right way around so the square pin here is pin one um it's probably easiest to try and poke it in from this side there we go so that's in there let's turn on the power to the pcb again and we're drawing 22 milliamps which seems fairly reasonable now we've got this additional pcb connected so let's go back to here i'm going to put the bootloader on as well just to make things easier so we can update the firmware from the usb port on the front so let's see if this works take kit 4 that's probably going to have to load some firmware onto the pick it is quite a long time since i've used it and programming and verify complete and that is doing what it should do so this is what it says it should do with the bootloader just so you can check that the oled and everything's working so that is a good start that also means that mplab is working properly we're able to compile the code properly so now let's try and build the main firmware and update it over the usb port and then what i'm going to do is i'm going to plug in the little usb connected to the front and it has detected that the pic32 is connected to the pcb so in the repository there's a piece of software called unisolder.exe which i've loaded up just here and there's a button to click which says update firmware so we need to go in and since we've loaded the bootloader onto the pcb we need to use pic32 with bootloader production and then here is the hex file that i've literally just built then it slows down briefly and then it starts doing stuff so that is actually the main program now running now keep saying 21 and then power lost i'm assuming that is because it's now not got the ribbon cable connected so we'll connect that up next and see if that makes it operate normally so we've got the ribbon cable connected up and let's power it up and there was a tiny beep from the beeper certainly not an annoying beep like we see on some of the other instruments and there we go so we've got our main program working properly which is good news that doesn't mean that all the electronics is correct that just means that our software is running on the pick and we've got our oled connected up um so the next thing we need to do is connect up a soldier line so i've made up a little lead so on the github repository it tells you how to connect everything up and you can see there's a pair of resistors which tell the unisolder pcb which handpiece is connected however what's a bit bizarre to me is there is a connector on the pcb that's called sensor and there's a connector called iron so you'd think for example on the jbc soldering iron there's only three connections the heater and the thermocouple and then there's the ground connection on the cartridge so you'd think all of those connections would be on here and then this sensor which is a three pin thing would just be for the two sense resistors so it knows which hand piece is connected but it's not it's connected across here which means that in this example where i want to use a jbc handpiece and i don't want to change the connector i've had to make this uh weird wiring loom with the resistors across the two connectors which just makes it all a little bit messy i'm not really sure why it's been done in that way however let's see if this works it should detect that the soldering iron is there with these sense resistors but it'll be confused because i haven't got this soldering iron actually plugged in so that's all connected up but no soldiering lying at the moment let's turn it on and yeah it's picked up the c245 because of these resistors and heater open because it's not connected up so next we'll connect the soldier line right so the soldering iron is now plugged in i've set the input voltage only to nine volts just so it slows the speed at which the cartridge actually heats up just in case it all goes wrong so let's see what happens now current limit set to 3 amps and it's drawing about 2.3 amps and the temperature is going up so [Music] let's see if it actually manages to regulate the temperature set to 270 for now oh yeah we can see the power dropping [Music] yep 270. so the calibration is probably off first of all we've got to calibrate the system there's a little trim pot here so let's do that next that's a good sign though that all seems to be working properly so to calibrate the station you need to put a 10 ohm resistor between these pins on the power pcb this should be a 0.1 percent resistor or you need to measure it with a high resolution meter so measure this one with my bench multimeter it's reading 10.0001 so we go through into calibrate and then this is the value that we want to tweak and we want to get this to 1000 in the case of a 10 ohm resistor so basically you multiply your resistance by 100 and that's the value trying to get on here and because mine's reading exactly 100 ohms we need to tweak this to a thousand not the easiest thing to do with a metal screwdriver there we go that's pretty much a thousand on the dot so that should be this station now calibrated okay so let's test out the calibration it says 370 degrees c so and it's uh definitely melting solder let's try on the tip thermometer three three four so we're out by quite a bit there let's try a different temperature so that's 300c and we're about 25 degrees out there so it's not the same offset if i increase the temperature further so 400 degrees c let's see what that does we're about 37 degrees out there so yeah i'm not sure if there's further calibration to be done next we'll try using the toroidal transformer that i bought for this project this one's a little bit oversized 160 va and probably 120 is suitable you could go less because you're not going to be drawing 120 watts continuously with one of these soldering irons and these toroidal transformers handle overload quite nicely so you probably get away with something a bit more compact but i've gone for this one this one was cheaper than all the others so i figured i'd just go with the slightly larger one make sure you get the phasing correct if you have twin secondaries like this one this is twin 24 volt secondaries but obviously if you get the phasing incorrect then you'll just short out all of the windings together and then we've just got two outputs here that's the ac that goes into the other terminal on the power pcb so this top one which does say ac next to it and let's see if that powers up okay so yeah that powers up just fine it should heat up the iron much quicker so we'll try that next oh yeah that's much quicker heating up much more like what we saw on the genuine jbc station now i didn't find any solution from my quick search about that calibration error i don't think connecting it up to the transformer because i did have it on quite a current limited supply before i don't think that will have changed anything but we'll quickly check it now so no it's about 23 degrees out maybe i'll try it with a genuine jbc tip but i don't expect it to be much different so we've now got a 2.5 mil chisel on here and that is melting solder very nicely i don't expect much difference here oh well it's a bit closer only 10 degrees c error so maybe there is something in the cartridge rather than the calibration of the unit itself but that is certainly melting solder very nicely so let's try a little bit of soldering and then we'll do the 2p coin test so i'm not expecting too many issues here let's try soldering some components down so no problem on a normal pcb let's try one of the high thermal capacity boards no problems there let's try one of these at the top it says we're putting about 50 watts into the pcb which seems quite reasonable for the tip this size it's not really any problems there and so some really good performance there from the uni solder pcb with a genuine jbc 245 cartridge as you saw very similar to the genuine jbc system and also looking at the power levels we were delivering really good amounts of power into the coin it was basically limited by the thermal contact between the coin and the soldier 9 tip because when it was warming up you saw it as drawing that maximum 130 watts so this all behaves as expected incidentally the 5 millimeter tip was exactly on calibration we saw the other one was out by about 10 degrees and one of the chinese ripoff cartridges out by quite a lot so it does look like there's a bit of variation but also quite a difference between a quality jbc cartridge and one of the chinese rip-offs so i think the next thing i want to do is get hold of the larger oled and then build a custom case for this i'll probably 3d print it on the 3d printer or something like that unless i can find a decent aluminium enclosure but i do think this is actually quite a nice project worthy of having on the bench i think given the weight of these toroidal transformers this is pretty heavy it might be the most sensible idea just to find a fairly stylish aluminium enclosure and then 3d print the front panel and we can then have the custom openings for the oled and the usb port and the various buttons and that kind of thing it you can also interface a rotary encoder instead of these buttons so that might be a nice idea as well now one thing i forgot to mention is there is a port at the back here for the automatic sleep detection so there is uh i think on here you've got the 3.30 volts and then the signal line when that signal line gets pulled low then it goes into sleep mode at the moment i've just got a toggle switch on there since i've got no power switch whatsoever but this collar is basically connected to zero volts so we can use this to trigger that signal and whether that means we buy a genuine jbc cradle or whether i try to find something else i need to see what the prices are and what we can find i don't think i'm going to make something for this because i think it will look pretty crap based on my metal working skills so i'll just have to have a look on ebay or whatever and see what i can find as a cradle for this and then just a word about the project in general now this is a fairly complicated pcb to assemble uh especially with some of the problems that are highlighted previously with uh the reference designators not been in any kind of order um and you know you would need a fairly decent set of assembly skills and a decent soldier line to assemble this in the first place now given that i have found that the c245 handpieces are probably the best ones that we're going to get without going to induction heating that does beg the question whether this is even worth building when we've had that really quite low cost axion soldiering station that i reviewed a couple of weeks ago which was performing you know exactly the same basically so really it's whether you want a little challenge of building a pcb like this or whether you really do want to be able to drive a whole range of different soldering irons uh one nice thing about this is you can drive the tweezers it's got dual output so that is possible uh to drive some tweezers that have two separate heating elements uh but there is quite a lot on here and i know a lot of people have started building this project and then basically given up because it wasn't working properly um or they just found it too difficult to assemble so there is that general question one thing that i would like to do but i don't know if i've got enough time to do is to redesign this to some extent simplify the bill of materials and also do it in such a way that jlcpcb can basically assemble the majority of what's on here maybe some of the special parts you know the usb isolator and the adc and that kind of thing you'd have to solder on by hand but uh all of the passives and probably some of the diodes and that kind of thing could definitely be assembled at jlc pcb so i'd quite like to revisit this design in general um improve a few things i didn't like about it but then make it a little bit more accessible to more people to allow them to be able to build it and i think if we got it done at jlcpcb with their low-cost assembly service this would be a little bit more affordable because um all together this has cost quite a bit of money especially when you have to buy several sets of components because of minimum order quantity that kind of thing it makes the axiom solid station look a lot better value for money so we'll see if we've got time for that but you'll see this in some future videos when we try and build it all into a case so hopefully you found the video useful and enjoyable also anyone who was having trouble getting this whole thing started maybe you can use this video as a reference point a big thank you to jlcpcb for sponsoring the video also a big thank you to my patreon supporters who are really helping this channel keep running until next time thanks for watching

Channel: SDG Electronics

Views: 20,603

Keywords: unisolder, soldering, electronics, jbc, metcal

Id: xGWQ4OvsW0Q

Length: 29min 50sec (1790 seconds)

Published: Mon Aug 23 2021