EEVblog #1364 - Compaq Portable PSU REPAIR

hi when we last left our intrepid adventurer he was uh caught up with this ibm compatible the world's first ibm compatible compact portable machine from the early 80s and managed to get the motherboard up and working and determined that the power supply was a shot we weren't getting any negative rails whatsoever which actually stopped the main processor from booting up because there is a logic output from this uh comparator circuit around here that obviously uh checks the rails on all the voltages there's plus 12 there's five there's minus 12 and minus five at least i think and they're all obviously sum together and gives a signal good output and it was given a signal bad output and that was going to the cpu holding in reset stopping the machine from starting him up but once we powered the motherboard up with an external uh power supply and bodged in that power good signal the motherboard came to life and were able to get video out of it it still worked so let's get back to this a power supply here this pesky little power supply reason we didn't look at it last time is because well look at it it it look it's like you can't even get in there to read or measure half of the parts it's all higgledy piggledy it's all over the shop it's just a really horrible layout to work on and of course murphy uh ensures that any trace that you try to uh you know trace on here it's going to go through to the top layer where you can't damn well see anything so yeah real pain in the butt anyway we're going to solve our minus 5 volt and minus 12 volt issue on here so let's take a closer look by the way we're going to be using the new eevblog bm786 multimeter available exclusively on the eev blog available on the store as of today i just got stocking so uh yeah anyway check that out um so what we've got here is well obviously a switch in power supply because it tells you it's made in the united states of america smps switch mode power supply this is obviously the primary side here we've got 240 volt mains or one of the 110 yankee rubbish uh coming in here and uh then we've got oh yeah a couple of little bridge rectifier two bridge rectifiers anyway we've got bridge rectifier down there then it just rectifies the input mains we've got our high voltage caps here and then um this contraption which looks like a giant inductor it's not this is actually your switch mode uh transformer you can probably see deep down in there there's some wires running off there so this is like it's really ugly they're just like hand wired over to points down in there there there and all over the shop um anyway so yeah essentially our primary side here and uh we've got a made in britain made in great britain thank you very much a 2n65 45 for those playing along at home anyway um obviously primary side and obviously the rest of this is secondary side with this is our isolation transformer and the taps for this thing are right down in there some like secondary uh taps so these are all of our secondary uh regulation capacitors now we're basically just looking at the uh topology basically because we've already gone over this ball we've done the smell test we've done the visual test to look for you know a blown stuff or dry joints you know things like that and we did actually detect a few dodgy joints on the lm338k but that's actually working because this is actually this is an adjustable voltage regulator here's the data sheet so this is going to be for either the plus 5 volt or the plus 12 volt rail so obviously right off the bat and because look at the size of the heatsink on this thing um it's well it's basically the largest one on the board here so that's telling us that well we've got a linear voltage regulator so even though this is switch mode uh power supply we're looking at possibly um not any active switch mode regulation on the secondary side over here we've got a linear regulator so what that points to is just a simple uh bridge rectifier or you know a diode halfway from the secondary side and then just going into the caps and then we've just got linear regulators now there's another device down in here you know you can't get in here to read any of this stuff it's ridiculous anyway now these two here on small heatsinks these are diodes not sure if they're primary or secondary but they've got cr on them and i can see that's an actual diode in there trust me about that here's our mains coming in over here we've got our bridge rectifiers going to our main filter caps here and then obviously yeah this is this is all the connections for the uh primary side of the transformer there and then we're going to have some opto coupler feedback somewhere there they are optocouplers um they look weird but uh yep old school package opto couplers there so i've got some feedback but i don't necessarily think it's doing much in terms of secondary regulation as i said we've got that linear regulator there so anyway uh primary side and then secondary side here we've got some extra stuff in here so yep this diode over here is on the secondary side and the other diode on the heatsink here also on the secondary side so what we're looking for probably is like a classic culprit in these is our blow-in rectifier diodes um yeah i have to link in the uh it was the hp oscilloscope uh repair where we had a really remarkable and uh really elusive uh thermal fault inside oh no spoiler alert no no no follow along i'll link it in anyway interesting now my first suspect would be a uh diode in there somewhere so um we'll just check uh these two biggies and then well i can see a couple of diodes really deep down in there i don't know how i'm going to get down and measure a couple of those we have to flip it over and measure from the bottom man cross our fingers we don't get the wrong pin um they're measuring diodes in circuit not always the best thing the one over here is here so 0.07 of volts 0.072 aha well that could be low in circuit impedance 0.03 no okay that may be uh low in circuit impedance we can go over and measure our ohms but i suspect so yeah 67 one way yep so they're not short it but that doesn't indicate that they're a dead ski so um what we're looking for of course is not a uh shorted die because we're measuring like nothing i believe there was like nothing on those rails so it wasn't popping any caps it wasn't doing anything so the diet was shorted we might be in a spot of bother um you know could blow something up or you know it's going to cause issues so i would suspect that probably like more than likely like a diode's gone open somewhere so a bit of light to get down in there and what do we got aha seven nine one two there you go negative that's our negative 12 volt regulator that's not much of a heat sink is it for your negative minus 12 so yeah but probably not like for the minus 12 volts it'd be like for rs 232 or something you wouldn't really need it for much hence the tiny heatsink but there's another device on the other side in there which i can ah i can't yeah so some people are wondering why we can't measure those diodes aha are they shorter are they open or whatever well look at these big ass resistors down in here there probably are bleed resistors directly across these caps here these are very common they discharge the caps very quickly if you turn the supply off so it's a regular half wave rectifier and then you've got a resistor in there you've got the coupler ohms in the uh secondary side turn in here plus you know however you know they might be you know a couple of hundred ohms down there so effectively um either way you're measuring the diode you've effectively got like maybe you know 100 or a couple hundred ohms in parallel with your diode and yeah that doesn't really let you measure the diode in circuit you'd probably have to take these out to measure them but that would be a red herring because i know they're not associated with this and i can't really film down there but i was able to uh get in with a magnifying loop i get in there tongue at the right angle with the light at the right angle and that's a 79 so bingo there's our minus five and minus 12 volt regulators on the one heatsink so hopefully you can see the cramped area i've got to work with anyway i can see two diodes down there one is horizontal and the other is tucked under that white gunk between the white silicon between the capacitors there and that one's a vertical joby that looks bigger so i'd say the bigger one is probably for the like the minus 12. okay so what we've got three pins there three pins there and the center pin on both of those is going to be the input well any uh 79 series uh voltage rig and the one on this side then one sorry this is the 12 volt this is the five volts so this one ground over here should be ground and certainly it is and then of course the trace goes around there to match this one that's ground okay so i can't remember which pin is our negative 5 volts but we can find that there it is there's our negative 5 volts and our negative 12 volts will be there you go there's our negative 12. 12 because the 5 volts is almost certainly fed from the 12 volts so that would be my guess any but we can check that so this is our 5 volt input here is that coming from our 12 volt output there and yes it is and no workers so that explains why both rails are dead because the if the when we don't get our minus 12 volts out then we're not going to get our minus 5 volts out so both of those are cactus so either we're not getting voltage to the 12 volt airport or the 12 or the 7912 is dead it could be either of those things so our dave cad reverse engineering drawing is going to look something like this we've got our transformer we've got our secondary uh tap i think it's probably only going to be a uh like half wave rectifier diode in there and yes you'll notice that i drew my capacitors backwards because i'm in an absolute twit anyway yeah positive on the on the ground side there because it's negative so we're getting negative whatever voltage out of here you know a couple of volts above uh 12 at least um and then you've got to take into account ripple and all the rest of it so anyway it's probably significantly higher so i've got a one two and then that just powers the 7905 and then we'll have some uh output uh capacitance as well also drawn in backwards but i got that one right got that one right and what you'd do as well is you'd go in there and you'd have a squeeze around the magnifying loop to which i use my macro lens for my camera works great times 10 macro lens and just look for any cracked joints or anything like that because that could be the problem may not actually be a component thing and with this power good here with this lmr 339 quad comparator what's going on here is a rough sketch um we've got all the power supplies coming in there might be more than this will be plus five plus twelve minus twelve minus five should have drawn an extra one there but they all basically i go into uh comparators here uh there'll be a voltage reference on these pins so it's basically you know are they uh like below a certain spec say five volts you might put you know four point seven five is it above that your typical five percent then it gives okay i'm good and how they're doing this diet arrangement here you could do it like as an and array or an or array it depends how you actually configure it and how you configure the positive negative inputs here and how you actually uh compare them but anyway basically the idea is that yeah this will give a output happy if all of the inputs here are above their thresholds that's it okay so let's follow the input pin of that regulator and the there's one diode here but it's got a little signal trace going out of it not sure if you should be able to see that um so i'm not sure it's that and the other big one here so let's try this so it should be a note and nobody dope and nobody dope oh that one's there we go we've got a charge discharge thing and so therefore yes it must be and of course it's the last one we check so yeah so that large diode that's the vertical one in there that looks like a big uh one watt uh jobby let's go to diode mode that's all right and the other way around is open okay i'm just going to uh power this sucker up here i managed to when i disconnected the fan i'll manage to uh get it out of the chassis just no touchy all of this section here on the primary side here we go it shouldn't need a load it should be okay but yeah there's our five volt rail 5.1 no worries and that drains pretty quick with no load they must have a output drain resistor on it and we'll just verify plus 12. yep no worries and that one drains yep quick too okay i've got a clip in there go into [Music] don't our output of our transformer tap there so let's have a look so we should um well it's it'll give us ac won't it nothing wow really nothing out of the transformer and vfd mode on there 0.3 out of the tap of the transformer so i yeah i've got circuit ground and that doesn't make sense at all and i can't actually um show you it's too deep down in there but i can actually see that the uh point of the diet i'm probing actually goes over to um the pin on the transformer so oh what the ah you're not going to believe it please and the answers in the comments down below i found it i found the problem hang on i'll get that come out your bastard show yourself there's the ugly turd there it is can you see that orange wire down in there that is the transformer tap for the negative um the 12 volts and negative 5 volt rail it was it's like it's come off so like i don't know is this just like because this is just flapping around in the breeze i think this has just vibrated loose over the years um you know the fan it's got a you know 240 volt fan on there i'm not sure how much this thing thing's going to vibrate but that's what it was so tracing down all that well there could be something wrong with the circuitry i don't know but um yeah the bloody wire unbelievable wow could have really chased a red herring down a rabbit hole there um just for the sake of a wire because like you know like you see it looked like for all the world that it was connected down to that point but it damn well wasn't unbelievable no wonder we'll get naf all out of it there it is there can't quite see it is that a single strand yeah that's a single strand jobby i think it's just broken off from the board unbelievable there's not much length left on that gotta attempt to get in there and strip that this is a real dog the hell kind of insulation is on there unbelievable that was a real dog get that out let me tell you so yeah i got to uh get in there with the needle nose pliers and feed that back down yeah it's not easy to get these damn connectors out either um so i might take the whole chassis over or bring the soldering iron over to here so here's the tap here these two pins so this one goes to ground obviously and this one goes over to our diode over here like this there'll be a filter caps in there and then or somewhere and then that'll go directly into over to there um and that's the input to our negative 12 volt regulator so i've cleaned out that hole so i should be able to feed in the wire from the top um and hold my tongue at the right angle and hopefully it'll get through and then hold it with your finger on the other side and then solder from this side because you don't want to be soldering from the top this is just ridiculous seriously got no idea it's right back under the transformer sorry i can't shut oy did i get it i think i might have got it hang on stay in there you bastard i think it's going to stay in there because it's a single core so it's stiff as oh yeah yeah good enough for australia yep there we go you can see something there so let's solder that back it's got enough length on there she'll be right flow that through and we're good to go but always give it a tug afterwards just to make sure yep yep tug test complete i think we're good to go all right let's power this thing up uh so this probably hasn't been powered up for quite some time look i don't know when it uh broke or how under what circumstances but obviously our negative 12 and negative five have well i think have never powered up since i've had it so uh yeah it could blow something who knows release the magic uh smoke but anyway let's give it a bowl minus five yes winner winner chicken dinner so if our minus five works that means our minus twelve works as well hello minus twelve let's do that again minus five minus da twelve there we go i think i think the i just moved my ground uh probe i don't think it was making loosey goosey in the contacts but there you go fixed winner winner chicken dinner so i have absolutely no doubt that we'll be getting our uh power good signing out of here because but murphy um i might say that well something in the power good circuitry's failed as well but now look it's going to be doing going to be doing its business there you go that would have been my last guess that was a wiretap on the secondary of the switching transformer um you don't normally get this because usually they go in via a connector of course um but this one no it's hard wired into the board but not only is it hard biting in the board doesn't use any of that stranded rubbish uses single core and of course uh the thing with single core is that if it gets uh vibration and flex and everything else um then it can you know it can break off fairly easily that's why like your high quality multimeter pros for example if we cut these apart these might be like you know a couple of hundred strands of wire like the real good silicone ones with the like the real super flexible ones but basically you know regular uh stranded wire might be you know seven strands of point one millimeter or something like that uh for example and of course you know it gives you a redundancy it makes them reasonably flexible but no using the solid core wires on there causes to come a gutser wow i'm actually glad it was that and just not you know which is the most common fault might be a like an open um a diode or something like that or even the regulator really although you know usually diodes would fail before the regulators would but yeah at least we've got something interesting out of that how long has this been gone 20 minutes at least and we can actually measure our power good signal which is pin two here and there you go it's five volts yep i will get in zero before won't we so the power good so i have no doubt that this thing will now power up our motherboard over there no workers but i've got to assemble it all in the case for it to do this so that's really annoying so that's not something i'm going to do for today's video so yeah i gotta have to like reassemble the whole thing and then you know see if the crt works and all that but you've already seen that the uh motherboard works we got video out of that and we knew uh we traced it down that it actually came from uh the failed uh power good on the power supply and that failed because our negative five and negative 12 were gone but even if we had one of those rails go anyway that was simple but even simple ones can be interesting so hope you learned something and found that useful if you did please give it a big thumbs up as always discuss comment and down below catch you next time but wait hang on i just remembered that somebody uh viewer thank you very much uh sent me in the link to the schematic which i had i got this a while back i forgot i had it so that would have helped anyway let's go in and have a look i believe it's the same one it's the compact portable this is the portable plus but that's pretty much only uh it has like a hard drive and some upgraded rom or something like that so anyway how are w sams and co computer facts and um we'll see why this is a howard w samson co uh is important in a minute printing in the united states i'm no i can't see that but printing in the united states of america anyway um so yeah for those aficionados here here's all the digitally stuff but we do have the power supply in here taken while pressing spacebar there you go look nice nice annotations on the schematic absolutely brilliant and anyway let's go down to the power supply um that's the part that we want and we do have it thank you very much there you go um beautiful all on one mostly on one sheet although i don't like how it sort of like goes off here it's like you know all these like jump off like down here it's like uh it's yeah anyway um and we do have an extra one over here which just has then these letters come over here and then these are the ones that actually go to the connectors over here but anyway as you can see it mains in here we've got some chocolaty chokes we've got a full wave bridge rectifier then we've got our big mains filter caps they've got some big bleed resistors across there and basically yeah this is all just primary side switching here and then it looks like it does have a feedback coil here because that's feeding back without opto isolation into there yeah and yeah and they've got little you know if you really want to go in there and analyze it they got uh scope shots but of course we didn't need the schematic we didn't need the scope it was fairly easy but you know if we had to go in there and you know this diet here was but you know something in here or this cap was bust or something you know like you might have some issues you might have to go in there and scope out waveforms and stuff but luckily we just had a broken wire so but yeah this is handy and of course on the secondary r side and yep as i suspected here's our 12 volt lmr 338k i suspected that was the highest power one uh and it was certainly and yep we've just got a single uh halfway rectifier there um there's no filter caps for that one so they must be on the other page yes they are there you go that's the 12 volt source um and as you can see yes it's got a 680 ohm resistor across here this one up here has got two 33k 2 watt resistors there you go so that's why the voltage was bleeding down and that's why i'll show you in a sec in fact let's go into this so if we zoom into here okay so we've got our 12 volt uh regulator here no worries but this is the uh diode that we were these are the two diodes that we were measuring here and of course look this one here look here it is 68 ohms this is your loop right here so if you're trying to probe your diode here and here regardless of which way you put your probes you're essentially got 68 ohms here in series with whatever this coil is it's going to be you know tens of ohms tops right so you got well under 100 ohms in parallel with the diode you're trying to measure this is why you can come a gutser trying to read diodes in circuit like this you might think oh it's measuring you know 50 ohms both ways or something and it must be shorter you know it must have like a or a highest impedance short out or something like that well no you've got to think about the rest of the circuit here and how there might be a bleed resistors like this in parallel so yep that's what they've got here oh sorry i was getting carried away i think this is the 5 volt c let's let's go to the next page c c c yes 5 volts here that's it um what so there's no regulation for the five volts really it looks like they've just got a big zener if we look at that one one up i think we'll find that's a 5.1 volt center yeah couldn't get that readily but yeah oh it's 5.6 volts uh 5 watt general purpose voltage reference regulator died well that's a bit how you're doing isn't it i mean they go to the effort to use a proper linear reg up here for the 12 volt roll but for the 5 volt rail they've just got a lousy zenna in there oh that's terrible muriel anyway here's our minus 12 and minus five and yep the minus five is connected just through to the minus 12 out there and yep uh we've also got a 330 ohm 2 watt in parallel here so that's why you have a hard time measuring these diodes in circuit so now if we go check out our uh power good circuit so yep there you go there's our diode and gate there and yep lm339 voltage regulator and this will be as i said it'll be coming so you've got plus 12 minus five minus 12 uh et cetera and then yeah you've got some resistory dividers and then a voltage reference will become yep it even says reference over here voltage reference yep there it is 2.5 volt voltage reference tl431 classic for those playing along at home and then we've got some opto isolators and stuff like that so there you go that is that is the schematic so that's exactly what uh we deduced it to be uh during the troubleshooting that you know it was just obvious this is what it was that it wasn't actually that it had linear regulation on the secondary side and it wasn't doing any secondary side switch mode regulation now this is the interesting thing i posted this on twitter a photo fact standard notation schematic with circuit trace copyright howard w sams and co 1987. well apparently this dates from like the 60s or something we'll briefly look at this because it's kind of interesting so i posted this on twitter and the tweeps of course came through are these scott uh def pom the sam's photo fact the service manuals published by sam they did a lot of manuals for old cb and amateur radios in the 80s but they did manuals for other types of equipment too and peter thank you very much barang you i'm butchering that pronunciation um unlocking the component pcb mystery so i still don't quite understand it but we've got an info thing but it seems to be basically uh putting annotation and stuff on uh schematics and things like that and here's a i guess their uh brochure now photo fact helps you lick printed circuit board troubles in seconds exclusive use sam's circuit trace features eliminate costly hunting for test points no more maze to trace no need to flip-flop board which means you know turn up and up bottom to top which is what we were doing trying to trace out where the traces are going here's how circe trace works for you all test points are clearly shown on the schematic and is plainly coded the test points are similarly coded on the printer circuit board so you instantly know are they um well i did like is it just like they take photos and so i don't actually know what the product actually is you know write for your free photo fact i think like september 1958 that is great right so this is like the late 50s um this circuitry stuff and like it was still still using it in the 80s apparently um so yeah it was still a thing please leave it in the comments down below if you've used uh circuit trace or not so yeah i still don't know exactly like what they're actually selling here is it some sort of documentation camera then that links how does it like i don't i have no idea anyway i hope you found that interesting and once again if you did give it a big thumbs up catch you next time you

Channel: EEVblog

Views: 56,701

Keywords: eevblog, video

Id: M6PVarUSjqg

Length: 29min 26sec (1766 seconds)

Published: Sun Jan 10 2021