EEVblog #861 - Rigol DP832 PSU FAIL & REPAIR

hi you've no doubt seen my Roy Goldy p83 to power supply in many videos before and there's been a few issues with this supply and I have to link in those other videos down below if you haven't seen them but something funny happened with this the other day when I was shooting some of that damn high speed video footage I didn't show this because it wasn't really there I didn't really get an interesting shot out of it but I tried to shoot some because somebody asked for it try to shoot some high speed video a thousand frames per second of a relay actually contact a relay in in there actually closing and opening and here's a shot of it by the way it's nothing exciting at all but what I did is I thought oh ok I'll try and put some current through this thing like the actual contacts on the air the two contacts and actually I'll put a decent amount of power through it and then actually see what happens when I break that power for example or shorted out and I didn't get anything out of that but what happened is I think the magic smoke has escaped from my gold EP 8:32 I was using channel 1 here and I had it set for 30 volts at 3 amps it's maximum capability for this channel ie 90 watts so I was basically putting this across the contacts and then I was using another channel to turn on the coil and then I'd activate or deactivate the coil after I press the camera record function to see if I could you know get anything interesting on the relay but look what's happened on this thing I first of all let's have a look at channel 2 here I've set channel 2 to the same 30 volts 3 amp current limit so the maximum it's capable of and if I switch on channel 2 it does exactly what you would expect BAM the output jumps Britain straight up to 30 volts like that I've got no load on there at all so you know at least significant digit there and everything's hunky-dory right turn it off and on but look what happens to channel 1 I've got to set to 30 volts 3 amps I switch on channel 1 yeah and even if I leave it for a while it's it's not gonna get up to 30 volts look it's that barely gets not even going to make 5 volts I don't think it there is something blowing in this thing like the output series pass transistor is blown or something like that if we have a very quick look at a day of CAD drawing of a typical output on a power supply like this and I've done this in that various videos I won't go through it again but basically there's what's called a series pass output transistor here it passes because it's in series hence the name it's in series between the internal supply from the transformer and the output that's your regulation element there so it's called a series pass transistor because it passes the current passes the power through and there's an error amplifier here which then taps off a divider on the output and then there's a reference voltage which will be your adjustable control coming from your DAC or from a pot or whatever on your power supply let's set your output voltage and then this error amplifier just drives your transistor bit a MOSFET or BJT like this it doesn't matter there's various configurations and that error amplifier is just a loop that just keeps this output voltage at a constant level because due to op-amp action here these two voltages will the op-amp will do anything it needs to on the output here to keep these two voltages the same so if you set one volt here then you're going to get one volt here it'll drive this transistor and do whatever and that's how it creates regulation and regulates your output and this series pass transistor is a really quite a low impedance you can think of it that way so really that's how it gets all the current through okay it wouldn't do that if it was a really low impedance so when you switch this output on here even if you've got no load you're expected to instantly switch to the 30 volts there but we're not getting it's just slow this output voltage here is just starting out at 0 for example and it's just I don't know slowly rising up I don't know if it's linear or but it seemed to add tape it off there and as you saw it was only get into like five volts or something like that even though we'd set it to 30 volts it just wasn't getting there was taken forever where is the good Channel just went switch on BAM right up to 30 volts here so what's going on my best educated guess would be that this series pass transistor is blowin because it's the most likely thing take it and blow when you're shorting the output because that's what I was doing I was basically shorting the output with the contacts but hey this is an adjustable current you know led but possibly it's supposed to do this I had the current set to 3 amps so if it you know it should have current limited to 3 amps and I you don't expect lab power supposed to blow when you show them you expect them to go into current limit everything's fine but it blew I'm not sure whether or not it blew when I was turning the thing on or whether I was switching it off and getting back EMF or something I don't know but yeah I reckon you know educated guesses that we're blowing the output past transistor because there's no other thing really I mean yeah it's really quite difficult to blow like your error amplifier or something like that that's usually got decent protection and the series pass transistor often there'll be a diode across here as well a back diode for some protection as well but you know I don't know about the configuration of this particular supply but obviously for it to slowly ramp up over like a minute and not even get to its set voltage obviously there's no low impedance here it's just leakage trickling or something like that so let's see if we can confirm that the way we're going to confirm this is to use my new BK precision 8601 DC electronic load this is a new model you've seen my BK precision 8500 electronic load before BK precision are kind enough to send me a replacement this is their new model it looks quite similar but it's got now joule line vacuum fluorescent display it's a much better functionality and the specs are improved in various ways and things like that a - to do a separate video actually teardown not comparing the two and doing some comparisons and stuff like that anyway very nice so what we're going to do okay I've just got channel 1 which is our blo and output just connected here and I haven't got any like I have not got the load switched on but it will sense the voltage and display it here on the terminals okay so this is not a load it's like open circuit high impedance so let's switch it on and see if this voltage tracks this it should let look so what we're checking for here adjust it where you could use a multimeter you don't need to use an electronic load for this but this happens to work as a nice precision 0.05% voltmeter as well excellent specs on the thing so basically we're measuring the voltage on the output just to see if it's really on the output terminals here or whether or not the display is just displaying something ridiculous and we might be getting our 30 volts there I don't know who knows let's test it so let's switch it on and not look the two are tracking there you go they're both tracking so this is genuinely the voltage on the output this power supply is definitely blonde so just for curiosity so you can we're actually going to switch this off and see how quickly it drops back down yet pretty quick that's what you'd expect a power supply to drop back down to okay so but now I'm actually going to switch on the load to see if the output transistor here can actually drive any load at all my guess is it won't I think it's blowing and or so no there's something in the past regulation element that is blowing and it can't provide output current so we'll put a constant current load so that's what CC is here constant current mode I've said it for a constant current load of naught point one amp so it's not high you know it's just a very low current 100 milliamps you know bugger-all so let's switch that on so if now had a load on the output let's switch it on what do we get why C doesn't even doesn't even charge up now it doesn't get to anything and if I change the constant current on that to like even point o 1 okay so 10 milliamps 10 milliamps right they can't even drive 10 milliamps this thing is buggered so there you have it I killed this thing by just short in the output with a relay contact that's all I did just opening and closing it it should have current limited at the three amps and I was being a bit brutal to it but hey lab supply its supplies like this as supposed to survive shorting the output that's the whole idea of having the current limit and the maximum spec to three amps if it can't survive that it shouldn't damn well let me set it as a constant current so maybe I got some you know some back EMF from the contacts opening they arced or something on it you know and it killed something in there but before I did this I just start test it I actually just had my leads like this they're coming out and I just shorted the two leads together and it worked just fine so what I'm going to do is just do that on channel two I won't try and use the killer relay I don't want to kill my other channel but yeah look I'll show you i smear 30 volts three amps okay here we go boom and it's gone down but it's gone up to three amps there it is and of course the output voltage is almost zero because you know it's just a little drug better drop across the leads and everything's hunky-dory right it can survive that just fine at current limits at the maximum three amps but I effectively did just that by hooking it across these two relay contacts turning on the coil and shorting out the contacts that's all I did I swear it's a shame I don't have footage of it unfortunate thing about this power supply is that it's a little hard to troubleshoot you can't just open up and then just probe around you've got to actually take a few things apart so yeah this may not be easy but the good news is it's not too bad once you actually swing this top board open like this you can judge you could actually just power it up and leave it open so it it all be a bit hairy scary but yeah you can actually do it so you can actually access the bottom board down here and you can access the top board just teetering on the brink of death here now we've got this is a three channel power supply 230 volt ones and one 5 volt one so I would have remember from the previous teardown which channel was what so as always follow the money for the wiring take it from I think it's this one I think it does actually go up to here so I think we could be in luck in that this top board actually does channel one which is rather interesting in terms of arrangement I would have thought you know you'd have you two big ones on the bottom um but no um it looks like it's this one anyway yeah not sure if you can see that but you can definitely see I can definitely see the wires in this bundle here going off to channel one there so yep I think that's a winner I think it's that top board nice and as it turns out I just happen to have a second board here and you'll know the reason why if you see in the previous video it had to do with the fact that this previous board board had a design fault that used to smaller heatsink here it got way too hot over the maximum junction temperature of the voltage regulator poor design and they had to re-engineer it with this much bigger heatsink here so I've done separate video on that so I just happened to have a board and tada I should have a spare series pass transistor you can always tell the series pass transistor it's the one on the monster heatsink here and if that's not a dead giveaway then once again follow the money here's the two big Spade lugs here for the output voltage that goes oh look there's our output current sense resistor there we go it's just tapping off you can see the - look see the - trace is coming out there going off to a an amplifier in there and that will be undoing the current sensing but yet the to be traces just go off so you can't see precisely where those traces go but there it'll they'll go maybe under the heatsink and in to our output series burnt to our series pass transistor right there and we could just follow some my traces here here's negative of a huge cap there and there it is that negative point there so that's how that's going to be our ground and it's snaking its way over to there which is tada that the bottom leader that resistor there and that goes off straight to a spite like so that's obviously the negative output there's Chinese symbols there even though the silkscreen is a bit dodgy on that one that's going to say negative in Chinese I'm sure and then our resistor jumps across here and that goes to the other huge trace here so obviously they got that directly across the output rail so that's a bleeder resistor so what that bleeder resistor is doing is actually discharging the output capacitor you may have seen it in the shop before I don't know write down there's actually a couple edge related capacitor I don't know what value maybe a couple hundred Michael something like that directly on the output terminals a couple of micro farad's so that actually just discharges that capacitor and you saw that you remember when we switched the power supply off even with no load it dropped back down to zero and that's the bleeder resistor actually discharging that output capacitor there but you might be thinking well how can they get accurate current control because this is like 0.05% current adjustment and stuff if we've got a constant basically a you know this resistors going to take a reasonable amount of current on the output well you'll notice that current sense resistor is on the other side of this bleeder resistor here so the output it's sensing the current going out here not through there so this can be any value you like and that transistor is a CEP sixty in 15 and that's a hundred and fifty volt of our 30 or 60 amps depending on the package or even higher in channel MOSFET so you know something typically you might find an output power supply like this a bit this could be the culprit because MOSFETs have a downside in that they aren't nearly as robust as bjts bipolar Junction transistors they're almost bulletproof you know you put in a 2 n 3 O double v in your power supply and Bob's your uncle right you're never going to blow that thing they just robust as anything whereas MOSFETs delicate little flowers you know you've got to be careful they can be damaged by ESD or you know any you've got to be careful with them although they have you know often it's much superior our performance characteristics so there's advantages to you in them in series past applications like this but yeah I think we might have come a guts are there I can actually see also is that a couple of diodes down in there perhaps for protection and just for our completeness the series pass transistor is not the only thing in the current path there we've also got our current sense resistor that we looked at here so just for currently these are usually pretty robust it's going to be almost impossible to damage this thing as if it really gross overloads will just check that double check it yep it's still intact yeah just as a matter of course so there's not much that can go wrong here as we saw this I filled a capacitor before the output of that filter cap big fat trace there looks like they got another bleeder resistor on the main filter caps that's quite nice attention to detail big fat trace disappearing under the heatsink guess where it's going it's going to one side of the MOSFET and of course the other side of the MOSFET comes out so really that's the only thing in the path unless there's as I said there's some sort of you know control circuitry thing blowing but I don't have the schematic for this thing and that's the last thing that you would expect because when you're mucking around with this thing short in the output and you know that's what you were doing and then all of a sudden it failed like that you know okay we might have popped the series pass transistor that's best guess so yeah you wouldn't go mucking around chasing red herrings down a rabbit hole here with the air circuitry unless you've looked at the series pass transistor but just as a matter of course let's see if we're actually getting a voltage on our the input to the pass transistor the output of that capacitor that's it there so I'll just measure that so once again it's unlikely to be the issue but hey let's probe the voltage on the output of that cap just to be a hundred percent sure that is that there is voltage getting to the series pass transistor just as a matter of course so let's switch it on and oh whoa hello it's on the lights are on but nobody's home I'm definitely making I'm making contact Wow something else something else has blowin like yeah it's not the capacitor something maybe there's a fusible element you know a fuse in here somewhere it might be a PCB mount fuse that's actually blown instead of I hope so because the last thing I want to do is replace the series pass transistor on the heatsink what a pain in the butt so Wow there's no voltage get into that I haven't disconnected anything or everything's still coming in oh my god like even if I disconnected the control cables it wouldn't have made any difference that this is nothing to do with actual control so Wow yep Gonski I'm not an AC no one DC Wow there you go it may not be the series pass transistor my potential apologies to the series pass transistor hmm actually I don't remember any relay I don't see any realizing this thing I'm just gonna I didn't switch the output on there but there we go I was switching hey there we go I'm switching the output on that's interesting and 1.7 not to it like you know Jam no okay that's interesting but I would have expected a voltage to be there regardless of whether or not I switched on the output it shouldn't matter it should always be energized that cap should always be powered up then oh there's something definitely wrong it I don't know what's going on could be been back powered somewhere I don't know or not it's still not right bingo found the culprit there it is there's the fuse here was the input okay this is the tap coming from the transformer you can see it going off here the big thick trace is going off into the rectifier over here and it's in series with a fuse there it is PCB mount fuse 17 Meg Gonski now ordinarily you might have to take out all these screws down ahead I think there's six or seven of them to get the board out so you can access the bottom of the ball to suck out a component like that but in this case we've got easy iron access to what we just heat up the top one pin at a time lift it up so it can lift the component out no your problem I've taken it out there it is gone ski and then we just apply some solder to the back to the top there and then we get our trusty solder sucker or solder wick or whatever floats your boat and suck it out all that lovely sound so that we can put it straight back in beauty bloody stupid cheap our solder suck as it came a little this these three little knobby things here and these are supposed to go in the side there and hold the thing on it stood yeah had to be dodgy brother there we go supposed to lock forward but moving it or shuffling around on the bench pushed it back yeah bit dodgy brothers there's a culprit it's a little 5 amp job' now you know like 3 amp current limiting on the output okay it's a reasonable value a reasonable value to expect in a design like that but why it's blow it I don't know it's not like it was you know inrush current from the caps or like were why I don't know is it just a cheap us 100 that's sort of just dodgy and has failed or is there some more systemic design issue here in terms of you know actually failing I don't know yet looks like it's got all the requisite stuff on there look at that you well it's all it's all hunky-dory so of course with that board I had a direct replacement 5 amp fuse no worry so I've wacked it in there hopefully I've connected everything back up correctly and I'm going to turn it on here we go big test come on pair up 30 volts 3 apps switch on well 39 volts nope nope nope fail it's not regulating there's more than the fuse aha and also it's not just the voltage check it out his drawing point seven hours 27 watts where on earth is 27 what's going Wow that's a lot of power it's nothing on the output I swear nothing up my sleeve actually I'm kind of glad that didn't work because it's kind of a bit boring just yeah we just popped a fuse whoop-dee-doo hmm this is more interesting yeah well I guess you could say uh maybe I should have checked the series pass transistor as well but you know I was feeling lucky turns out I wasn't anyway but my first guess is still that series pass my MOSFET because it's gotta be to get 38 39 volts or whatever it was on the output that's probably the transformer tap here so it's like it's shorted or something like that where it's drawing those 27 watts I I don't know okay here we go can see some components heating up there there we go I just switched it on and yeah let's have a squeeze under there oh that little heatsink there there you go sixty degrees it's that little that little tiny that little tiny one in there is not the pass transistor you can see the pass transistor there it's not get in that hot at all it's that puppy interesting and if you're curious to know the rail voltage there the output of the rectifier and the transformer there we go 53 volts it's a 63 volt cap so that seems a bit high for a 30 volt rail but I don't know I can't remember if we've measured this in our previous videos or not but yeah it seems a bit high but hey it's just the transformer and a rectifier so I'm sure it is what it is can really see that little heat seep there is like 97 degrees something like that I have not got the output switched on by the way Wow so that's that little puppy in there whoa now the great thing about having a spare board like this you can test your hypotheses so let's have a look I've got it's powered on the output is not on and that same heatsink is only you know 37 degrees or there abouts in fact this little puppy next to it is a little bit hotter so it is significantly different hmm but hang on I'm not just going to jump in and take that bait I had a look at the part number on this this is a Biddy one three six and a bipolar trenny and these are pretty robust little things why is this getting SuperDuper hot yet the main regulator here it doesn't I suspect that there's nothing actually wrong with this it's just sinking the current that may be the failed pass transistor here is dumping into this thing so yeah once again you don't want to chase those um red herrings down a rabbit hole so let's have a look at the data sheet for this some n-channel MOSFET here and it's a gate drain and source let's go between gate and drain here we go so sorry if you can't see that does it matter wha two point five ohms that doesn't sound very good between gate and drain let's go between drain and source Wow look at that dead short between drain and source gate and source there we go let's swap the probes around get a different polarity gate and drain again two and a half ohms yeah let me have a smell yep culprit and if I get our good board just to show you let's go between gate and drain again here we go tada open exactly what you'd expect you'd expect the gate to be open and drain and source there we go four and a half K yep I think something's gone horribly wrong with the pass-transistor as you'd expect its MOSFET little delicate flowers they are and well I don't know actually hey look at this board you know how I showed you before that there were some diodes in there this is the old board no diodes in there look here we go it's the new one they've added some diodes I didn't notice that so they've upgraded maybe protection on this board perhaps but yeah maybe a fat lot of good it did so we're at a point that there's a pretty good bet that our pass transistor here is failed so luckily I have the spare board I have the exact type it's not like I'd have one of these puppies lining my junk bin you'd have to be really into yeah pass transistors to have one of those I might have something maybe kinda sorta would work but yet no I've got the real one so I'll just swap it well hello look at this I have not touched that I just took the board out from the metal bracket there look at the hand solder residue on that it's nothing look here's another arm here's the other than the Biddy 136 on the he think it's not it's properly um soldered at this one somebody's how'd it go with that look you can even see a scratch mark on the board from the eye and I think what and just to verify we've got it out here we go and nope that is not normal for a MOSFET the gate is supposed to be open nope Gonski so there you go replaced of course don't forget to put the heat sink compound on the back of the thing first just a little bit and make sure you screw it in nice and tight before you do the solder joints and we're back in circuit and let's do the drain and source there we go four-and-a-half k I think that's the value we'll get in before yeah and the gate here you go up Miggs yep good enough okay so what are the odds but working well I don't know with Murphy you never know so everything's plugged back in correctly let's have a ball and come on you can do it I've set it to five volts as playing around before so let's switch on five volts Weiner Weiner let's go right up to 30 bingo we fixed it it was exactly well I'm yet to do a load test but it was pretty much what I thought in the first place it was that damn it series pass transistor n channel mosfet easily damaged if you don't design your product right clearly there's something wrong with this Rygel if we can just you know blow up a series pass transistor like that that's just crazy and we blew the fuse of course so that we gave us a bit of early hope that or maybe it's just the fuse but nope it was the past transistor so there you go but it's it was really quite worrying that this transistor had been D soldered you saw that it had it had been hand soldered insert like iron and I'm not exactly sure why whether or not it was a production step afterwards it could be that could be a normal production step in that case I need to use some no flux our solder or clean it up or something but yeah anyway um that's it we know winner chicken dinner what the photon is going on here my beak a precision load is dead I haven't even hooked it up yet I just get like go to turn the power on like I swear this is the same power cord oh it works and I plug it into here nothing what buddy Murphy unbelievable the dear fuse is blowing would you believe it here you go working a treat now it was the correct rate in fuse it was half an amp because you have different because this is like a selectable voltage range on the back you have different fuse ratings for different mains voltages and it is the label on the back says half amp for 240 volts and that's what was in there and a blue a place with a half happened fine what the yeah these things just happen I am with monotonous regularity in the eevblog lab is it the eevblog curse I don't know all right let's check this puppy out let's switch it on 30 volts let's just go the current limits are 3 amps let's just go constant current 1 amp shall we just for simplicity turn it on yep drawing a nap no worries works a treat so that's all it was isn't who blowing n-channel series pass transistor in there why I don't know ok what we're just going to do now is just do some math shorts on the output once again set to 3 amps 30 volts 3 amps ok boom there it is ok it's recovered just fine and did it did to do okay well hey there we go got some sparks woohoo beautiful catching that but I know what everyone's thinking Dave we want to see the killer relay try it again okay this is exactly the same thing I had when I was shooting my high speed video channel one output here thirty volts at three amp maximum limit set is across the normally open dark contacts here and I've got second channel here at 12 volts hooked up to the coil so I'm just going to switch it on and let's see not it seems to be working since we're working just fine fine yep all right so your guess is as good as mine as to why it died but I swear that's all I was doing to it and I think I'd like they've did it a couple of times I think you know three or four times I got a couple of shots out of the thing and like you managed to kill channel 1 in this case nope nope since rock-solid down so there you go I hope you enjoyed that video it's a bit longer 42 minutes worth of footage well sorry Owens so I was hoping it'd be a quick but yeah in the end it was exactly what I thought it was the series pass transistor in there and that bt 139 my guess is it's are being used to drive the gate in there so because it's high capacitance or whatever these are power transistors usually are maybe you know it's used in there and that was that shunting all of that power because the gate as you measure that the gate was like shorted out to the drain and source so it was just dumping everything from the rail how it was actually measuring it on the outputs the shunt resistor is out here for the output but the year I don't know you need the circuit or to trace it out or something couldn't be bothered but there you go I hope you enjoyed that look at this troublesome Roy go TP 8:32 this is not the first issue with have with it of course there's a few and been a few notorious issues with it but I thought they were done and dusted but apparently not I don't know did I get like unlucky but it doesn't matter I would not expect a lab power supply of this price and grade to blow a pass transistor in it why when it was you know it should current limit three years handle it properly I have no idea what went wrong so maybe Rygel should look into it anyway liked it give it a big thumbs up catch you next time thanks for watching if you liked this video click here to watch similar ones you can also give the video a big thumbs up by clicking down below don't forget to subscribe to my channel by clicking little icon down in the corner here you can also sign up for email notification alert so you get notified as soon as I release a new video no worries that's the way to do it you can also comment on each video down below and all the hard core discussions happen on the EEV blog forum there's a separate forum thread for each video and yes I do read and respond it to all comments where possible and if you want to help support the channel I accept donations on patreon you can buy merchandise and other things linked somewhere here check it out

Channel: EEVblog

Views: 182,312

Rating: 4.9471993 out of 5

Keywords: eevblog, video, rigol, dp832, dp832a, power supply, psu, lab, repair, fault, faulty, fail, schematic, circuit, measurement, multimeter, regulator, how it works, tutorial, pcb, solder, relay, high speed

Id: VIZNmHznYiE

Length: 35min 42sec (2142 seconds)

Published: Wed Mar 16 2016