EEVblog #804 - HP1740A Oscilloscope Repair - Part 2

hi just a follow-up video to this HB 1740 800 megahertz dual channel analog oscilloscope and the repair of this thing and where we got to last time was that we found that once we took the case off this thing it basically would not fail except for one very small CAPTCHA that I got by coincidence as I was tweeting a photo where all of the power rails 5 different power rails on the output of this secondary linear transformer this is in a switch mode converter supply so five different linear rails all dropped at once and well I left the thing for like at least four hours it might have been closer to 5 hours or something like that still would not fail so it's either a thermal problem where the unit where because the the case is now off this thing the heat can escape it can't build up inside and the thing is not is upside down now so if it is something to do with the power supply for example or are something on the bottom like this board here then you know the heat doesn't sort of build up inside that case but it may not be a thermal problem it may actually be a mechanical issue so let me actually explain on the schematic here what I'm talking about but before we do that if you're trying to capture intermittent faults which I wasn't here by the way because I you know I thought I'd just sit there for an hour and then just fail I wanted to see what the rails did but I didn't know it would have that like intermittent you know a dropout like I I captured by accident so if you want to do that with your meters you'll hook them up and you use your min Max mode so that's what I've done on these 4 meters here and I've left the thing once again for another like hour or so and I haven't been able to capture everything anything at all before this thing was easily failing within the hour and it was fairly repeatable at that but I haven't put the case back on everything anyway I've set them to min max mode so it'll capture any transients which go low so I've got it display in the minimum at the moment but it's always it's in min Max record mode all of these meters even this old flute 27 has it and I'm actually displaying the minimum so maybe I can actually how do I simulate a mains Brown out or something just by wiggling the power cord or something like that perhaps here we go I'll give it a go and they'll all jump down just a trap for young players if you've got a negative rail like this you don't want to set it to minimum like I've said this one if you're looking to get Brown outs and drops in the voltage you've got to set this one to maximum because it's a negative voltage so it'll go up towards zero effectively so you know so it's actually maximum so I don't know let me see if I can wiggle this mains lead on the input I could switch it off an arm but that's it you know I want to get like a way there we go there we go I did something there we go drop down there we go so that was just by wiggling the mains lead on the input so we're able to work capture that now as I said in the previous video looking at the schematic for the main power supply board you can see it's it's basically a standard linear transformer it's got multiple isolated taps two four six different tap see all going to standard bridge rectifier a full wave bridge rectifiers and then the big filter caps the big old ones from us Prague and Mallory I think they are anyway but yeah we've got a linear regulator chip you know pass transistor mounted on the back all kind of you know it's standard stuff but basically I'm measuring the output of most of these are measuring five I think we've got six taps here anyway I'm measuring five main rails here and they all dropped according to that photo even though I have not been able to reproduce it yet but we did actually capture that drop so you've got to assume what can drag all five rails after the linear regulators down as I said they have each one's got individual current-limiting hidden here a current limit output resistor and it's got current limiting inside the chip whatever chip that is I'm not sure but it doesn't matter um how would you drag all the output of all those linear regulators low well the most obvious think like because if you just shorted out one here for example even if they say this diode bridge rectifier shorted out it did something weird something you know you load that down didn't have any protection etc I don't know the cap was doing something weird you know it's unlikely to drag down any of the others because of the low impedance source through the transformer it's just not going to affect it these other channels so odds on it's got to be something on the primary side which is causing a drop out because then anything that happens on the primary side if it can't deliver enough power then the outputs here because these are all drawing I don't know how much power this scope draws you know I don't know 10 20 watts or something right you know it's a reasonable amount of power okay so if this primary side for some reason cannot deliver that power all of these outputs are likely to drop while the outputs of the outputs of the linear regulator here but you know obviously the outputs of the transform will drop and then bingo it'll go through to the linear outputs so the first thing I would suspect is something on the primary side of this transformer and just forget what's going on down here with these couple of transistors this is just some interface stuff for the two B and C's the gating outputs the main and delay gating outputs on the rear panel there just happen to use those as a convenient jumping from one board to the other because the board happened to be there so they're mounted on there and doesn't matter so that's got nothing to do with this main site and the main site is incredibly simple we've got a an i AC mains input connector here one amp fuse on the input and then we've got our switch okay that's our front panel line switch on the front panel and then we've got a the voltage selection switches which is on the base of the unit down in here and and then then you've got some socket wire in contacts you've got your PCB connector going off to here so I am suspecting possibly first point of call would ever be the front panel main switch or the main selection switch here because these are you've got to remember these are like what 30 this is 1980 okay so this is a 35 year old scope so these are 35 year old switches 35 year old contacts on there and contacts can pit and corrode over time with use for example so if the contacts are pitted and they can and because they got power reasonably you know not super high power flowing through it right we're only talking 10 or 20 watts or whatever this scope takes but it's enough current to cause a potential issue so if the contacts either in here or maybe the front panel main switch are pitted then it could potentially have no change well like under our time maybe it heats up a little bit maybe it's got a little bit of high resistance and then slowly heats up inside until some you know something happens on the surface contact of the switches in here which then it causes not to be able to deliver enough power to the transformer so all the outputs are gonna drop that's the first thing I want to check so yeah you could go right down the rabbit hole trying to look at all the other boards in the thing and what Evers hooked on to the output of all of these channels here see if one of them's load and down and you can chase that rabbit hole but I think that's pretty foolish because they all seem to drop so this I think definitely worth looking at so I think if you didn't have a look around here first you were you know gone for a wander down the garden path now the problem with this is is I cannot reproduce the fault so what on I'll probably have to do first maybe off cameras to disconnect all the multimeters put the covers back on run the thing for an hour and see if I can actually reproduce the problem but anyway there is our mains input voltage selection so there's some like PCB contact switches under there it looks like we might be able to take that off an inspector but they're probably like you know PCB solder connectors on there and yeah I can go and you know it's probably almost impossible to expect inspect because they're probably sealed switches in there we'll have a look in a second but yeah I could go and spray some you know contact cleaner in them in in those but the problem is if I can't reproduce the fault how do I know that I've fixed it so yeah we have to try and reproduce the fault first before we go spray the contacts it's uh it's great to have this theory that it's probably something to do with the contacts somewhere it's not likely to be the board to board but you might take those out have a look at the contacts on there just as matter of course though we got one I got lucky I just switched I just disconnected the mains power cord out the back and then reconnected at am bingo look I on the screen down here now it's ad why bother setting up another camera shot but have a look it's like yeah like there is nothing I can't adjust that we get in you know just that fixed line and all of the rails all the rails are low look at that why hmm okay so these aren't in min/max anymore okay so I'm gonna wiggle that mains cord and see what the problem is nope so I'm wiggling the mains input connector okay so it's not that let me show you that I'll show you my wiggling let's wiggle the front panel line switch give that a little bit of a can see that I'm given that a bit of a bit of a jiggle there we go nothing doing down in there I'll get my isolated Trotter well these are all secondary so no drama there give those a bit of a bit of a wiggle these are the main mains connectors give those a wiggle and main switches those faulty selection switches although I'm only hitting the top maybe I can hit the board no nothing it's permanently low so here we go I will my power let me repay with the back here there's all the rails droppin hey look has it permanently failed now excellent oh that's what we want that's what we want permanent failure you bloody rippa now we're getting somewhere Murphy's on his lunch break I think all right so let's test our primary side Theory here the way we can do that is take for example just one of the rails we're doing one at a time otherwise I don't need like ten multimeters let's take the five volt output here okay so the five volt output goes through this pass transistor and it goes into this side two plus nine point five volts there to this fifth s c11 this 53 hundred microfarad cap on the output of this full wave bridge rectifier so if this doesn't measure nine volts orth actually says nine volts here and then nine point five volts there a little discrepancy anyway if it doesn't measure around about nine volts or you know significantly higher then the five volts because the dropout voltage of the pass transistor here so let's assume in say it's a two volt maximum dropout voltage it's gonna need at least seven volts to regulate this thing so if it's not at least you know it really should be around about that nine volt figure if it's not then we know bingo the primary side is not being able to supply enough power on that particular winding and most likely on all the other windings as well so here we go I've got another meter here set up across that fifty three hundred microfarad I cap down in there it's still got some charge on it hasn't been able to bleed it off because they haven't got a bleeder resistor on there so the all that you know the regulator I see is the pass transistors switched off and there's still some charge there anyway so let's pair it on hopefully it still fails so let's pair it still oh oh it's working nine volts you bastard oh no no hang on no seven volts you saw it here we go there we go it's droppin so there's our five volt rail there's a five volt rail 4.2 volts no wonder you know it's it well actually point two volts regulation I'm pretty good actually so uh yeah but look so it's dropped and all the others rails have dropped as well but you can see that the output of the full wave bridge rectifier at its failed of course and so the output of the full wave bridge rectifier is it cannot you know is is dropped so that means our price is most likely our primary side of the transformer cannot supply enough power let me check one of the other rails okay I've now hooked it up to our six thousand micro farad cab that's through our plus fifteen volt regulated output the output of the bridge rectifier on that cap as we should read on here it should be about 21 volts or thereabout so it's paradigm yep it's failed again highs as repeatable hey no we're still kidding look at that that's interesting Wow there you go and lost that bit that's enough that is more than enough to give out regulated fifteen volts output but we're not getting our regulated fifteen volts output so that's really interesting hmm and there we go that's interesting this is the negative RA oh it shows positive the hair I've hooked it up back to front Murphy got me anyway that's the negative rail so we're looking at negative 22 volts so that's correct as well oh wow I really lost that bet and so but our negative rail is minus 12 but it's got more than enough voltage to regulate it it's got the regular output voltage expected from that full wave bridge rectifier and let's try the 42 volt rail as well I've got it across the 500 Mike 75 volt cap should be about 55 volts according to the rail and yet but we're all good again bloody hell come on fail fail come on you can do it you can fail come on and there we go it's failed but it's at 60 volts so it's actually jumped up which seems to and you know like there's less load on there so yeah that's interesting but we lose regulation so maybe there's something that's tie maybe an overload on the five volt rail or perhaps into well that causes dropout of regulation of the other channels hmm but the bridge rectifier outputs the all the other ones are fine it's only the five volt rail I can actually whack that one back what was it this big one here there we go no hang on no twas this one here was it can never remember yeah there we go four volts no good whatsoever and I'm just having to look at the ripple on the five volt rail here and we're on two volts per division and as you can see it's just over that tight yeah that four volts that we can see over there I haven't got the multimeter on the rail but yeah it's you know that 4.2 volts or whatever we were seeing before and the ripple is basically bugger oh look at that the main output voltage is dropped and by the way if you're gonna our scope probe these things I've done a whole video on light mains ground earth referencing and you're probably better off for using an isolation night transformer when you're testing something like this or just make sure your ground reference point for your probe is actually shezzy ground otherwise you can blow the arse out of you scope and all your product okay I'm trying to make it come good now but it's no it's not going to come good but you know it's not like the capacitor has failed and then otherwise would see huge amount of ripple on here cuz it should be what normally aren't nine volts or whatever so you know two four six eight you know it should be like up here and you know if the cap was troublesome we'd see you know a large amount of ripple but I was seeing hardly any ripple on there at all so it's it must be drawing excess current and it's it's just dragging that down the output winding it can't provide enough power so what we need to do now is go back to our schematic and have a closer look at what's happening here because what we've been doing up until now we didn't like sit down there and analyzed all how the power supply work made a few assumptions and it was actually quite reasonable to suspect the primary side because all of our outputs dropped and so we did the right thing we said we came up with a quite plausible theory about the primary side here we went about testing it by and we actually found that our five volt the output of our bridge rectifier here on our five volt rail there's nine volts actually dropped right down okay so that seemed to confirm that theory but then when I went to double check always double check this okay don't assume anything so I went and measured the other rails here and these other rails weren't being dragged down so that fact basically ruled out our primary side Theory high impedance primary side dragging everything down on these secondaries so we have to go check the rest of the circuit and see what's what okay so what we've got here is our three let let's just look at this like ignore all these complicated looking ones with the transistors up top let's just like concentrate on what's happening to the five volt and the plus minus fifteen volt rails here now so we've got three regulator ICS here okay and but when you actually look closer at these okay this one here is the one for the +5 volt rail sorry plus 15 volt rail okay so we've got our plus 22 volts coming in here and we measured that that it was still 22 volts yet our output was actually being dragged down now this looks a fairly typical look here's our output here our plus 15 well there we go sorry you can't see that but here there it is a plus 15 volt output here and we've got our L look we've got a voltage adjustment pot here for the fifteen volt so we've got their typical output voltage divider feeding back and actually into our regulator here using this external pass transistor and that is a very very typical you know voltage regulator but we know we're measuring 22 volts here but we're not getting 15 volts out of here why but more interestingly let's take a look at another one down here okay is the negative 15 volt one where is where are the feedback resistors from this minor from this negative output rail look aha here's our feedback resistors look it's reference to the fifteen volt rail the plus fifteen volt rail here so this is not independent it actually relies on the fact that this fifteen volt rail is set correctly and then if you go look at the five volt voltage regulator uh-huh right here's our output here is our Apple current sense resistor here's our series pass transistor where is the voltage reference coming from bingo the plus fifteen volt rail again so that plus fifteen volt rail drags down of course it'll drag down the five volt rail of course it'll drag down the negative fifteen volt rail so it looks like the all the rails there are referenced to that plus 15 volts output and of course wouldn't you know it if you actually go and read the theory of operation of this thing it tells you exactly that look at this all voltages plus five forty three hundred and twenty plus minus fifteen what would be measuring and the high voltage our reference to the plus fifteen volt supply dole so everything cause it so it must be made operational first the supplies a current limiting type of who as we've seen they've got those current limit resistors so any excessive loading on the vertical horizontal etc will cause the supply to read twenty to thirty percent low and that's what we've been seen so of course they're going to it's going to drag down all of the rails okay so what it actually told you to do in the troubleshooting procedure and what's obvious is to actually remove this board here which connects the output of the power supply here to all the other boards here the horizontal and the vertical boards and well that's a really that's really is very nice it just that slips out like that because we've got card edge connectors on here they are looking great condition there's no no corrosion at all on there everything's beautifully gold-plated and be very thick gold plating to top-notch no worries whatsoever and looking at the rails and of course they're all bang on I might just leave it for a while and see if it fails but I suspect neither there's something that's dragging down one of those rails and of course with no horizontal and no vertical what do we get we get a dot straight in the middle because it ain't driving it left-right up-down or wherever and you can actually see the high voltage output here still connected to the board down here so we still drive all there high voltage stuff we're still driving our CRT and everything else it's just that we're not connected through to our horizontal and our vertical boards here so what we're doing now is just checking to see if it's the horizontal or the vertical boards at fault here see if our problem returns but as we've as we've been seeing here these intermittent faults are a pain in the ass because you don't know whether or not you're just getting lucky and the faults not showing up it could be in the high voltage section which is still being powered from here as I said it could still be in there but you know there could be some reason why it's not showing you know Murphy you'll get you every time so you know just because we could leave it here for an hour and it might still be good but that doesn't actually prove anything as such this is why in a bit and fires are such a pain in the ass you can waste a lot of time you can go down a lot of you know chasing a lot of red herrings down the rabbit hole and well yeah so but it's not failing so far so I don't know you name the odds of the high voltage power supply section being at fault I it's not I suspect it's on either vertical all the horizontal sections and that suspicion is backed up by remember our +5 volt rail is the one that actually went down here and the output of this bridge rectifier was really loaded down and by the way it was that past transistor that was getting hot and this voltage regulator here you - I've actually checked the position on that on the component overlay and there was that one that was getting hot for that five volt rail so a five volt rail over here we've got assembly a 14 and it looks like a five volt rail doesn't go anywhere else so I am suspecting a 14 over here I don't know what that is we'll have to have a look well there you go that doesn't help here's our power supply up here this is the interconnecting board the a14 interconnect we just start physically removed and that +5 ox comes out of the power supply and goes off to both the horizontal sweep assembly and also to the vertical preamp assembly so it could be either one of those horizontal or vertical well thanks for that and if we have a look at this rather complex looking interconnection diagram our low voltage power supply here our +5 volt output here as we saw it goes off to the horizontal assembly down here and it also goes off to the vertical assembly in here but then the +5 volts from the vertical assembly also goes over here as well there's I you know the horizontal is pretty boring there's not much doing in the horizontal so I'm more just for it sheer odds point of view I think there's more likely to be stuff happening in the vertical channel so I'm more suspecting the the vertical wire side earthing the vertical board then the horizontal board but I mean you know it's just guessing really okay so what I'm going to do now is just have a quick check of the 5 volt ray or output current and we don't have to get in and break the circuit with our multimeter to measure the current at all because we've got ourselves a current limiting resistor here this 1 ohm resistor our 30 here we can just measure the voltage across that and hopefully the voltage across it when it's not failed and then wait until it fails and then see if the voltage increases ie the current increases so let's give that a bell there we go that's our 1 ohm current shunt resistor I'm using these are parrot clips in there I love through-hole parts like this troubleshooting through-hole parts because you can get in there with your little easy hooks or your parrot clips or whatever or your even your alligator clips and we clip on to the components down at the solder on you don't have to do anything it's really easy so it will pair it on yeah we're working and we're getting 0.25 volts there so we're looking at you know 250 milliamps on the 5 volt row I'll just sit here wait for it to fail hopefully it was failing before within a few minutes so hopefully fingers crossed oh there we go there we go 4 volts it's dropping our currents actually going down yes I've got it hooked up backwards I didn't know which way it went so it's yeah but no our current hasn't increased but look now current hasn't increased but the voltage that's 2 volts per division so our AC output our rectified output tap has certainly dropped look at that there we go I've powered it back up I've put the probes around the right way so 250 milliamps and there's our normal there's our normal ripple after our full wave rectifier so 2 4 6 look look slightly up you see it drop did that drop oh is that just my imagination anyway 2 4 6 oh did we get a glitch there 2 4 6 and then oh wait no that's that's one sick puppy look at that but our current you'll notice our current did not increase so it's not like it's being overloaded bingo and just as a matter of course I'm going to check the connections on the transformer input there they look they look pretty darn good no corrosion or anything on those so based on Kirchhoff's current law what could be can be happening here where can if it is excess current where can it be going so say for example there's excess current on the output of this bridge rectifier and it is actually dragging this rail down and this diode bridge and this tap here can't can't supply the power required where is it going well it's not going through here it's not going out of here because here's our current sense resistor it can there's only two places that can be going one is somehow through the base of this transistor into now I think believe it's an LM 723 voltage regulator here or it's going around here once again into a current limit pin it doesn't seem likely at all we've measured the voltage across here there is no excess current flowing out there so I can only flow it there or there um so it can't go anywhere else so what's happening here well as I said if that capacity if that was a bad capacitor this 5300 micro farad cap if that was you know bad as well it's you know 35 years or 30 years old or whatever it is 35 is it um yeah you'd certainly suspect that but you'd see a lot a huge amount of excess ripple that's what I would have expected so I am starting to suspect either this transformer tap which is highly unlikely the interconnects in here which is a plausible just like we had on the primary side we thought maybe there's a you know some sort of interconnect issue or the diode bridge and the capacitor when we can prove I'm just whacking a 22 I think it's a 2200 microfarad 60 odd volt across the rail there and well let's have a look at our ripple as well and watch it okay everything's working hunky-dory at the moment but I suspect this puppy is going to fail and yet the scope scope still works so everything's fine I think well stow oh he did did that just drop was that my it no look it's jumping around it's jumping around remember this is the voltage across that full wave rectified capacitor there so I think if we wait I reckon it's going to drop and do exactly the same thing as before i yep there we go it's dropped BAM voltage across there four point four volts our voltage over our our five volt rail the output of the voltage regulator four point one bingo even with the extra cap on there so it ain't the cap now you wouldn't know it's really handy about having these transistors on the back and the connecters going straight on the pins because these are identical series pass transistors we can swap them this is this one here is the one for our 5 volt channel that we're looking at here but we can just swap it with this one here so that's what I've done what the-- the y's are just long enough on an angle to get over there and plug in so we can see if the fault stays with the transistors swapped over we know there's not a problem with the training in some weird way shape or form ok it's still working and oops sorry forgot to turn that back on it's just discharging that cap and ripple I am just is jumping around here I reckon she'll fight yep there we go failed again not the transistor not that I expected it to but hey because we can swap it very quick easy test to do ok now what I've got is I'm just measuring the transformer output tap there we can in can see that 9 volts ac there and let's wait until see if we get a fire I'll switch this back sorry I I move that a little bit there we go right there we go bingo we're still getting it look it's gone back up so that meant we're still getting the AC out of that transformer no problems whatsoever and because this looks like there's less load on it look so what's left diode bridge that diode bridge down there is our coal port because I was measuring on pin 7 and 80 pin 1 starts over here goes straight another diode bridge so I'm suspecting that puppy but I do stand corrected it could potentially still be like the solder joint on those pins or maybe the connection inside there but I can't really see any problem in there it looks really good it could be you know it could be a dry joint on the bottom of this connector or the diode bridge itself it may not be the Dyer bridge could certainly be an old fashioned dry joint now I was about to say this is actually really quite easy to get out because the the war into the past transistors on the back just pull those off couple of these disconnect the mains here disconnect the secondary take off the plate there there's a couple of wires in there for our mains input there's our main cell voltage selection switches for those fanboys but there is two screws going to there that there that there's a bottom mount that's the mains power switch right there which is no surprise because here's the mains input and there's a shaft which comes all the way on the bottom of the board and connects to there so oh and I've got a disconnect carefully disconnect this connector here through to the bottom board I've done that and I think it sort of yeah I don't know how that attaches under there cuz there's the whole high voltage supply on the back of that so jeez I don't know well I figured it out I sort of move this slight I've disconnected this main connector down here I can move it just like how like a half a centimetre towards the front panel here and then once I've got it on the front panel tada there's the line switch and the line switch she's actually square but it's protruding enough that it now lets me unscrew it tada look at that so I can unscrew the shaft from the main switch on the back this is all very very clever and by the way yes I did eventually figure out that this is actually explained in the manual dole so bingo this now with perhaps some difficulty hmm should actually come out somehow yeah well looky what we have here look at those pins can i wiggle those in the bottom not a huge amount look at those dry joints joys a dead dingos donger but the funny thing is that's not the one that I'm suspecting okay this is actually cr3 I think it is which is the flood gun the high-voltage a flood gun tap which I wasn't looking at this is the one that I'm interested in that's the that's that tap for the five volt rail but that looks that looks okay so yeah I don't know anyway so I'm not sure if I caused that problem by wiggling the connector out I don't know but geez I'm certainly going to fix it up no joy this here is the diode no sorry this one is the diode bridge we're interested in so yeah I'm going to replace that diode bridge as a matter of course that's for the five volt rail and also these ones on the other end too they might show up really well on the camera here but to the naked eye they look like good joints I had to look at those under the mantas microscope the others look good these ones also these two here I'm not sure if this will show up on camera it's hard to see on the camcorder LCD here but these ones will these ones also have that cracks in them but the one we're interested in this one here looks to be good but I'm going to re solder this whole damn connector just as matter of course and I decided that suspect diode bridge and it's interesting to note no solder has flowed through to the top side of the component there like they have for the other components look all the other components solder has fed through no problems at all but on that diode bridge it suspiciously hasn't I think I'm going to go through and resolder all the do bridges it might not be anything wrong with the diode bridge but I will replace it with a new one or a new old stock as a matter of course and that diode bridge wasn't alone IVA it looks like practically all of them are gonna have that same issue none of the solder has flown through and granted that should it be a problem but you're relying on the through-hole plating of the PCB itself and you'll notice that you know most of them have all the traces on the top half if I flip it over you'll see that very very little is actually on the bottom there so you've got to rely on all that top half connection right through thee V is there that's asking for it especially after 35 years so that might not be a dodgy diode bridge at all by the way are the pad fell off in the repair there I was using a reasonable temperature but it just came off so you know 35 year old PCB man but all of the most of the connections look all for that do bridge three for that one three for that one 3 for the 5 volt one under interest too and all of the connections for that that we saw some really dodgy dry joints on I mean like completely cracked right dry as a dead dingos donger every single one of those connects to the positive side so if we had no solder flow through on these dire bridges you could all assume that we had no solder flow through on those I didn't actually physically remove it and check it but I sucked them all out it wouldn't sort of budge there was some you know it's quite hard to get these sort of things out sometimes depending on the hole size but I resold them as a matter of course and all of the dire bridges so yes I did replace the one diode bridge on the top side there there it is so I replace that puppy there but I probably didn't have to I suspect those now looking at those joints I don't think there was much wrong with those diet and that diet bridge at all I suspect it's just 35 year old solder joint problems and this problem has been probably been sitting in there waiting to happen for 35 years not adequate solder flow through like they got on these parts here look all the other parts no problems at all but maybe these higher thermal mass ones or maybe they were sold as separate or I don't know what the deal is they had no flow through whatsoever and maybe the connector too so anyway resold at all let's whack it back in okay and let's power it on and see what we get I've only hooked up the 15 volt rail but up helps if I plug it in plugged in we're getting four or five volts and we're getting our 9.8 everything's hunky-dory now all we've got to do is wait but I suspect we've fixed it and I'll tell you what it seems to be the diode bridge actually I've had this thing jumping around I've just got it tested here I'm supposed to be drawing like a half amp load from this puppy and and it's it was like 5 volts before and it's not it's just dropped it's just dropped and it was actually jump back on probably Murphy means I'm probably not going to be able to get it to jump back come on damn you come on anyway I reckon there's something thermally wrong with the with the you know the diodes inside this thing and that's probably our first I don't think I've ever like I've seen diode bridges blow but not ones that sort of you know intermittent thermally fail like that so yeah I'll see if I can nab it though there we go I got it this is what it's normally like okay and hopefully we'll just see it suddenly jump bingo hey just saw it gotcha there you go it just jumped down and if we let it cool down it actually recovers and it's repeat all the bloody diode bridge unbelievable so there you have it the eevblog curse has been lifted where I either when I get repair are stuff like this it's either so incredibly simple to fix or it's BR beyond economical repair and you know too complicated and expensive to fix but this one fantastic I hope you enjoyed a look at how I traced down to the bloody diode bridge do you believe it and potentially some solder joint issues as well unbelievable I don't think I've ever seen a diode bridge fail intermittently like that usually these you know Todd bridges yeah they fail but they fail usually fail open like that and well these ones this was failing not so much like open if it failed open then it would have been fairly easy to find that we weren't getting any voltage out and stuff like that but because this thing had a quite unusual power supply arrangement in that all the voltage Rail references were actually tied to the 115 volt reference then if the 5 volt one went down and there's all the other surgery it by some mechanism I haven't gone in and you know investigated the whole thing maybe through another board or something like that can actually drag the other the 15 volt rail down and then the 15 volt rail drags down all the other rails as well and you would have actually noticed that all of them are dropped by exactly the same same percentage as well so that sort of you know clued in that they were all tied into the single 15 volt rail but that was a fascinating troubleshooting look and like a power supply fault right really really simple but because it failed in a very subtle and intermittent way you saw how I actually got a little bit lucky here in terms of that it did actually play ball in the and actually failed pretty much on cue and I could power it up wait a minute it had fail but it didn't do that the first time I played with it in my first video it was you know sitting there for four or five hours and wasn't doing a thing so if it doesn't fail and well we came up with it you know a couple of theories the primary side seemed like a reasonable theory to check and I it was lucky that I went in and double-check that of course to make sure that the Abbott voltage taps had actually dropped as well as the five volt one I saw because if I did that I might have gone off and you know I tried to look for some short end by the way if I followed the troubleshooting procedure in this thing I might have to take a capture of it and show you it basically implies I think I read it out there I did show up before it says that the horizont if they all drop by 20% or whatever exactly what we saw here right all the rails drop then you know they said either horizontal the vertical you know it's likely to be in a fault what that's off you're strictly following the troubleshooting guide you may have gone down that rabbit hole thinking ah there's some sort of overload on the horizontal or the vertical boards and it's lucky that I actually well it's not lucky I deliberately went in there and went before I do that I'll go in and check that check the current it's worth checking the current double-checking checking things before you go down that rabbit hole chasing all those red herrings where you think it might be you know you might have fixed ten of these before and go ah yeah it's been the vertical board or whatever actually had a few people were email me since the first video on they said oh yeah I've seen this it's you know the vertical border it's this or it's that and stuff like that nobody nobody picked a diode bridge and potentially some dodgy dry solar small joints dry as a dead dingos dumber those things were amazing and typically you'd start out with troubleshooting something like this with a visual inspection but you can't visually inspect those joints that we saw in there until you take out the whole board so we I only had the measurement you know just take some measurements to sort of see where we led so I hope you like that troubleshooting I could have made this video shorter sorry it's been like 45 minutes or something in the second one because it's been going for like an hour I just edited the footage and it was like 40 minutes worth something yeah for not for another five sorry at least maybe ten but we really got lucky with this puppy that it was such an obscure fault that you know you're not very likely to see something like this year intermittent faults or you know happen all the time but usually something you know like that dry cracked joint it'd be you know ever a physical flex thing you might be able to you know flex the board just - I was poking around if the poker it didn't actually come and go then which indicates this probably wasn't the problem and end up being an obscure thermal issue that didn't just oh you know fail open it failed sort of high impedance which was a different thing which made it look like and it you know it could have led you down the garden path you could I could have easily wasted a lot of hours on this scope before eventually finding something like this and that's the problem with bloody intermittent fires anyway I hope you enjoyed that and there's probably people who are saying on jeez Dave that was that was pretty easy it was just a bloody power supply why didn't you find that in five minutes well I basically it did not take as long as what you see here I was waffling on and you know going through what was talking through what was in my head and stuff like that it might have been like an hour's worth of troubleshooting video but in reality if I didn't have the camera on and I was just working on this it was probably 10 15 minutes worth of work so it's you know it's probably not it took me longer I think to get the board out and then repair it clean it up and put it back in then it did to actually find the fault in the end so it was a pretty quick repair in the scheme of things so what are some quick lessons from this one will always measure your voltages thou shalt measure voltages golden rule of troubleshooting don't assume something's overloaded actually measure the thing double check if we didn't double check we may have gone down you know around don't necessarily believe any instructions you have yes they can be handy but they might also lead you down the garden path as well got to have your thinking cap on and with these intermittent faults don't just go around with you know a theory in your head I yeah it's the primary side contacts and start cleaning all the contacts and going ah yeah I fixed it because Murphy or bloody will get you I guarantee it that it'll you know it'll look fixed it'll work for a week or whatever and then the thing will come back but we got recently lucky on this one that it decided to sort of play ball but it may not have this one could have been really ugly so we made a couple of assumptions in here again came up with a couple of theories but we tried to verify them and then hey I was wrong you know it wasn't the primary side hey but it was worth a quick you know the thirty second look just to measure it and make sure but I we found it pointed somewhere else and then that pointed to another thing and bingo we found it Beauty never assumed and also this is not a bad example of where having multiple multimeters comes in handy and potentially even though we didn't get that far having a multi-channel scope a lot of people ask well what use is a 4-channel scope well you can measure for power supplies at the same time and see what they're doing and capture transiency we had some weird you know fireboat we may have gone down into that detail and maybe even because these are ground reference maybe you might have needed oh yeah we didn't get this far even if it might have needed a nicer multi-channel isolator does scope what this one two channels hopefully do a repair and tear down on this one too soon ah two isolated channels so that we can get in there and improve different points at the same time fully isolated from any reference between scopes so yeah it's handy to have more than one meter I keep saying it good example I hope you'll enjoy that that was a bloody Ripper I love a good adventure hunt like that so I hope you did too if you want to discuss it jump on over to the eevblog for links down below all that sort of stuff I've got the warranty void if not removed t-shirt I'll probably linking that down below if I remember you usually don't anyway leave YouTube comments blog comment oh let's sort of Jess catch you next time you

Channel: EEVblog

Views: 93,642

Rating: 4.9661822 out of 5

Keywords: hp, agilent, keysight, oscilloscope, fault, repair, how to, debugging, thermal, failure, power supply, electronics, capacitor, diode, diode bridge, multimeter, soldering, dry joint, solder joint, cracked, heat, pcb

Id: 7ocancoZ02Q

Length: 50min 33sec (3033 seconds)

Published: Mon Oct 05 2015