Hi welcome to a teardown and potentially a well maybe the first part of a repair video. Check out what we've got. We've got an Agilent sorry it has actually Angela on the front in 8760 to a system DC Power supply and this thing is a beast. Check it out.

Look, look at the specs. Oh baby! 600 volts, 8.5 amps, 50 100 watts Wow In a two rack unit high thing, this thing is an absolute beast I Got it courtesy of our chose that our trio test and measure and he got this from a customer. They blew the ice out of it, they release the magic smoke and now we're gonna toss it in the dumpster I Believe! So Joel's art saved it and he said here we can have a look at it, have a teardown and maybe have a repair of this thing if we can anyway. so let's take a look inside.

Oh it's a Bobby Dazzler look at it I mean voltage and current. Are you know, like a lot of these system power supplies are just that. They're designed to go into racks into systems which are under PC control and things like that. But this puppy has the rotary encoders on the front for the voltage and current control as well and limonene and all sorts of stuff on the front.

very handy for arc control. I Hate the system power supplies that don't have any front panel indicators at all. It's just annoying. And this particular series of that system DC supplies comes in many different models of course, as is common for these type of things anywhere from an 8 volt unit with 400 amps capability up to the like.

The highest voltage model we've got here: 600 volts at 8.5 s50, 100 watts. And yes, 50 100 watts is more than your average power point can supply. So 240 volt or 110 volt. You know, 2400 watts here in Australia is the maximum.

This one is 5100. So yes, it is three-phase and the specs are incredible. - this is 0.02 five percent basic DC Voltage class instrument 0.1% on current. Absolutely incredible for such a high voltage high current unit.

Tell us the price, son. Well, this particular model 7700 Australian dollars ticket price crikey and on the back here we've got good old old-school GPIB we got 10100 Ethernet LAN we've got USB host we've got in active. whatever that is out in I don't know some sort of control thing. Maybe we've got a cap on here.

That's the remote programming interface. We've got some dip switches to set some various things. We've got the sense lines. Yes, you've got these sort of currents.

you got to have four terminal remote sensing. And here's our voltage output. Our: there's a big-ass terminal block there. and there's actually our two connections for the positive and two connections for the negative.

So if you want the sense line, if you want to sense it at the low, then you've got to take it from here. But as you can see there as is quite common, if you're just using it basic without any sense then you just put in loopback pins in there to do local what sense feedback. And here's the trick and why it released the magic smoke. Here's the three-phase inputs, right? So you've got L1, L2, L3 plus your our Earth terminal.

Okay, no problem. Standard three phase input: 192 240 volts Okay, yep, it's designed for the Australian market. Look up here: 190 240 volts Ac 3 kilowatts plus no worries, right? were wrong. This one's actually a 120 volt Us model because we're talking three-phase and in a Y arrangement that can be 208 volts Or it can be 240 volts depending on the particular configuration.

Fret on a per phase for a hundred and 20 volt. You know nominal country. Like in the US, there are various standards and I won't go into them. You know it depends on what type of factory and installation and what type of three-phase system, how its wired, etc.

etc. We won't go into it, but basically know you could be fooled into thinking this is a 240-volt unit and you can hook it up in a country like Australia which is a 240-volt country and you can hook it up to our three-phase system here, but you can't here. It's actually 400 volts for a three-phase art system typically. so that's exactly what they did here: I Don't know where they got this thing from I Don't know the history, but Charles tells me that.

Sure enough, they actually hooked this puppy up to a local three-phase system here and the magic smoke escaped or some. Let's take a look inside. So let that be a lesson to you on three-phase systems. Know exactly what you're doing.

Don't be fooled by labels like this. But yes, even though I don't have three-phase here in the lab, I could actually technically hook this up to a regular 240 volt outlet I could fit. You know there's no reason why I can't feed my single-phase 240 volts into it and parallel it on all three inputs like that. It's just that I wouldn't get the available power of course I wouldn't get the full 5,800 Watts that this puppy is capable of.

but you know, hey, it still allows me to troubleshoot and do everything else. Actually, technically I would not do that here in the lab because my lab voltage here is actually 245 volts at the power point. and actually I'll prove that. And there you go.

I'm actually pretty close to 247, so that's really quite at almost the extreme limit of what they're allowed to actually give us here. So yeah, not necessarily such a good thing. So even though it's only at seven volts over the nominal operator limit of 240 volts here, and it'd probably work, you know, I'd be incredibly surprised if it was an issue. Still, just as a matter of course, you wouldn't do it.

So I would use my troubleshooting this thing I'd use my adjustable mains power supply. So here we go. Let's pop the lid on this puppy and this thing is going to be first class. I Can guarantee it.

They will have spared no expense on the parts and the construction will likely be absent. Lutely drop-dead gorgeous. let's hope. Agilent Don't let us down.

sorry keysight does have actually bench on the front though. sorry Tada. Oh yeah baby, look at that. that is gorgeous.

Wow And even though this puppy has been powered off for a long time and I'm sure it's got bleeder resistors in there and not to store any residual charge on the caps. You know when you're mucking around with something like this, you just want to check first. So I'll get in here to this lovely looking bus bar and I'll give that a probe and yet bugger all. and then you can go probe the main side but it's all discharged.

You just wanna make sure this whole thing is discharged both the secondary high voltage DC side and the primary side over here. or discharge before you go poking around in here. And just remember if you want to discharge some things, if you've got a multimeter with today one of those that Losey things then well that'll do the business. It'll put a resistor in series in parallel with the Cos series parallel with the cap and it will discharge it so and it'll display the voltage as it discharges.

Beauty. Now first thing I'm gonna do is give it an old sniff test. No I can't smell anything really bad. Um so yeah.

and visually at first glance I don't see anything wrong with this thing. Oh I can see yep down here. Well it's not actually something wrong that's blowing up. but two of the fuses for two of the phases here are missing.

These are our look. 30 amp wah, big Beast HRC ones - 30 amps are 250 volt rated fuse. they're missing. So I don't know whether or not somebody's you know, how'd it go with this thing or whether or not the fuses just blew out.

I mean look, there's something something happened there. Perhaps that's yes. Actually, you see and the side of these caps here look at all that there's all this crud. I mean this is a very clean unit.

This crud is not anywhere. It's not dust or anything like that. so it looks like you know this fuse here is likely like totally exploded. and yeah, all that crap and is just yeah.

it's not charred everywhere, but it's yeah. I don't know. Has the ceramic exploded? I I Don't know but I mean HRC fuses are supposed to take huge surge currents like that. So I'm surprised I Can't immediately see anything else visually blown in this thing that's funny-looking Check out that they've got some silicon stuff on one side of this.

Um, this cap here. Are you kidding me? Like why? it's that's not? No, it hasn't like ooze down at the bottom of something. I'm actually can see this. actually.

you know quite a few places around in here actually gluing components together, but why they're just stuck it on the side of there. I Got no idea because they haven't done that to these ones down here. So go figure and just look inside. This thing, Isn't it? Gorgeous.

All separate ball construction. Here's the mains input here. it's going to have. it's actually got.

It's a big clunk and switch on the front, but it's just going down to there so it looks like we might have some relay switching to actually turn that off and on. Even though it's got a big clunking and switch, it doesn't seem to be. It's certainly not switching the three-phase the actual mains input. so it's doing that electronically and we've got I'll show you inside in a minute.

We've got three bigger side bridge rectifiers here, so see 3 separate fuses. We've got some caps. We've got our common mode jokes here. so with a three-phase common mode choke arrangement, I've got through our three bridge rectifiers here and I assume these the big-ass relays that looks like a little transformer, but we'll have to get some numbers off those puppies anyway.

It's got a power factor correction as well, so it's going to be doing some power factor correction is that down on this board here. And then. these are our two big output boards. Even though this is only a single output, it's obviously getting it like there's This is like the highest power model.

avail. So they're obviously paralleling two of them and you can see that on the busbar arrangement here. these are the two outputs here of the supply. These two here see: I can touch that because I've already measured it.

Um, the two outputs here and the two outputs here. These are identical power supplies. They just put them in parallel with these huge bus bars and then these bus bars. It's quite nice.

they come up into this top board here. You'll notice there remove some of the solar mass to increase the current handling capability of the copper on there. but that's probably your two-ounce copper board anyway and just going to the output terminals here. Oh nice little spark gap in there.

Let me show you there we go. It's always nice to see a little spark gap arrangement like that. I've done a video actually arcing these across it. It's really cool.

I'll link it in if I remember and no wonder this thing is gorgeous. Well, it'll be gorgeous anyway because Agilent / Keysight know how to define products and they design really good ones. but the best in the business has done this. TD Calendar Well, one of the best, if not potentially the best power supply manufacturer in the world.

So it looks like Keysight overlook everything subcontracted. TDK Lamda: No surprise. I Mean you know, leave it up to the experts to do these sort of things and they would have charged a pretty penny to let me tell you. So it looks like every power supply board in here is TD Calendar: This one is this one.

this one and you two main power supply outputs as well. So it looks like Keysight of Subcontracted Out that I did. probably maybe this output board here might be. you know custom Keysight and may you know the digital side of things.

I mean that wouldn't be TDK Lamda. that would be all our. Agilent So they've got the GPIB board here and processor and it looks like a control board down the bottom and it looks like someone in production has written some stuff on here after testing. Now what is that? p 205 K So is that maybe five? five kilowatts they tested this puppy at at 200 volts? Maybe.

So yeah, they've written those on top of the heat sinks after they've done some production stress testing on this. Mmm. sorry for the hand footage, but it's easiest way to get this yet. Here's one of the relays on the main power supply input: Tyco Relay I'm not going to look up that number, but it certainly is.

And there's two of those relays in there. that one and that one. and as I thought this one is a transformer, so that's doing some you know, tapping off a signal for cents or something like that. Now one thing I'm a bit disappointed that just from a our service and teardown troubleshooting point of view is that you know, yeah, they've labeled all the connectors you know, Cm-1 and everything else.

but like, why can't they just label them like with their system function like you know, five volt rail, 15 volt rail, whatever you know, like actually be more descriptive on the connector outputs, but you know there's nothing like that on here. There's no test point voltages or anything like that, so you know it's not exactly service friendly and I can't I Had a quick look, but I cannot find the service or service manual or schematic for this thing, so if anyone has it, I'm at please leave it in the comments because that would be absolutely fascinating. We'd be able to tell a lot more without having to work reverse-engineer all this jazz. but yeah, what? I Think yeah, we've got our input mains input switching here.

This looks like our As I said the power factor correction. That's important when talking about these sort of loads or your professional power supply is going to have PFC in it. It looks like we've got another power supply board over here, which looks like it does the low voltage rails for your digital and stuff like that and our secondaries. Actually, this PFC board here has the old anemic Lander name on it, and that's the same as the mains input as well.

but the low voltage DC supply has TDK Lambda, as do the two main power output ones as well. They're also TDK Lambda. So yeah, maybe a different group within Lambda Lambda Lambda Lambda. The tri-lambs eighties aficionados will get that one.

And of course, you expect nothing less than Nippon Chemi-con There you go. The real deal. 105 degrees C Of course. So of course these are all Nippon Chemi-con All the major ones in the DC power supply and the input.

Now the output side and the input side as well. And they've put that gray silastic in between those two stop and flapping around in the breeze. Don't want to vibrating. Nothing worse than vibrating capacitors.

Really annoying. So those caps are the output side. Of course they are. of course our 450 volts so this is a 600 volt capable supply it.

So they're going to have those puppies in that series. So they'd be a series. so there'd be two in series and then two in parallel. Like that.

So two groups in parallel. the output diodes here. They've got eight of those on this directly tied to the busbar output. which and they're as see because they've got no insulating washer, no sill pad on the back of that.

So they've got yeah, four on one side, four on the other, feeding the direct rail which goes right out there right to the bus bar. Very nice. And then there's our primary sized switching transistors. We've got four of them down in there.

They'd be big-ass MOSFETs big cell pad on the back of that. Don't need too much heat sinking on there. This is a pretty efficient design, but as you can see, they've got the fans right here, so drive in the air directly over these. The heat sinks are oriented in the correct orientation, except it sort of would have been nice to put this in the path of the fan here instead of like just slightly outside the path of it.

There would have been nice to get that air going over the fins, but you know they've done their homework. They've done their testing and that's adequate. So that's a small s oh yeah, air being sucked in the front and pushed right out the back directly across. All of these would have been a real faux pas if they mounted the heatsink like that orientation and they're just blocking all the air.

That's a trap for young players that one. And you can see that they've done this absolutely perfectly designed up on the power factor correction board up here. they've got the fan here. It's sucking the air in from the front, blowing straight over the capacitors first, so it's keeping those really cool and then the two heat sinks are aligned in the correct orientation so that the air can flow over the fins and straight out.

So also it's you know if you had it the other way so that these capacitors we're on, you know, over here like this then you'd be blowing the hot air onto the capacitors and that's not good for their life. So you want these caps on the cool side of the heatsink and your airflow like that? Nicely designed and they've got their cells a little temperature sensor there right on the heatsink. that's very nice. That's a token brand NEC own those I think or vitamin I don't know.

Anyway, it's a little I'm any see token temperature sensor very nice. directly sense in the primary side output heating. You can see here that the secondary side in addition to the four other diodes on the other side of the bus bar. Here there's addition for another four in there so that'll be for a higher current model and I think you can see down there.

there's a jumper in one position there and that would be to our configure these output capacitors in series or parallel our combination for once again for the different voltage current models. So lower voltage model they might work all these and higher current they would work all of these four capacitors in parallel whereas the higher voltage one like this is six hundred volt one, they would actually wack them well as I believe I think they are two in parallel, two in parallel and then the that group in series with that group to get your higher voltage. and they'd probably have a ballast resistor on there I'd be guessing just to equalize the voltage across the two. All those well looky what we have here There we go.

There's where a magic smoke escape from looks like we have. Is that a moth? We've had the arse blown out of that one? Only one of them. Only one. So yeah, that's interesting.

That looks like the only visible, visibly damaged component in this thing is. see where it's just blasted. Look, it's blasted right out here and just spewed all its guts right against this heatsink. So yeah, that's the first thing you need to fix.

And that puppy is actually a Nippon chemi-con moth. That's 470 volt nominal rating. hence the up: 471 K There, that's a dead giveaway and this one's a 30 jewel. Yeah, obviously I'm because you plugged in 400 volts.

So I See in here when you rectify that. Oopsie. Now, just looking at the failure mode here, we know where you put 400 volts per phase onto this puppy. So when you are for wave bridge rectifier that of course you're going to convert it to RDC Multiply that by 1.41 times and it's going to jump over here via these big jumper cables.

I Didn't show it's going to jump over this power factor correction circuit and you know it. Let's soak them like we've got a little more bound here which is blowing on this side, but it's only for the low energy stuff going over to this secondary one. So yes, oh, that's not like a main mod. I Don't think it looks like might have to get the board out to check the wiring.

It doesn't look like it. that doesn't look like it's you know, huge directly in the path, but it's still there and a blue. But anyway, you know you would suspect something on here. I mean your DC Filter Cats: We've only got four hundred volt caps 420 volt rated caps there on our main DC rectified output of this thing.

So because we're basically four wave rectified and the mains input so that that is not high enough rating to cater for the 400 volts per phase that we got here. but there looks to be no stress damage to those caps. But you know, like B Yeah, I wouldn't probably fix this thing without replacing those and putting it back into our production environment or something like that, perhaps. But anyway, there's no physical stress on those that don't.

Doesn't look to be any physical stress on any of the other parts in here. None of the four transistors in here have any blow holes or signs of physical stress. but yeah, we don't necessarily know that. But visually, the only thing I can find is that moth now in theory because the the main output of these DC cabs.

If it stressed these ones, then it's going to stress these ones up here as well on the input because there we've got our wires coming here. they go straight over to the input. They've got the same caps here over here, so they're effectively in parallel, but they're like doing some local local bulk storage down here for the channel, so you know these ones could have been stressed as well. But yeah, I can't see anything at all.

but after that, I mean I wouldn't expect anything on the secondary side over here to be blowing. So if anything's going to be blowing, it's the primary side over here. and all this power factor correction stuff in here, so who knows. We might have got lucky and the fuses here and the Marv's actually are protected or mob singular what the others would have kicked in as well.

but they didn't have their ass blown out of them so all the energy was directed to that middle one by the looks of it. But yeah, they could have taken out the fuses. Quick, smart, before any. you know, before the energy could build up in.

these are caps here and well, that's what you know and damage anything else, that's what you'd be hoping. Anyway, this puppy's interesting. Check it out little transformer there that's got this big ass like secondary winding here with like a coiled well, a wrapped like sense wire around that just going on the secondary side. there.

It's got four pins, but it doesn't look like there's any coil on the second side. They're only on the primary side here and they've got a similar sort of thing happening there on the output board. Now that that puppy is just flapping around in the breeze, they're not sure why they've done that. And I Popped the main output board.

here. you can have a look at it, look at these big ass. once again. the same nip on that chemic on the same series.

These are our big 450 volt caps and they're whack these in series because this is a 600 volt power supply and you can see they've actually circled 600 volt model here and these four resistors here. These would be the ballast resistors that ensure that the voltages are shared equally between these - caps here: I Mean you know you can't actually rely on the caps - what self balance? but you know using the internally. Sr But yeah, you know that's a bit how you're doing so you know there. You can see that they've done it, are properly and put in the proper ballast resistors.

and they've even got them stood off from the board there. Lots of air spacing under there. No workers got ourselves four sets of suppression caps here and there are silastic together. No worries whatsoever.

and that's it. There's a that looks like a common mode choke. It's not very impressive if it's only got a single wire turn on each side there. So, but anyway, that common mode choke is actually.

yeah. So that's what they've got. The spark gaps across the common mode choke there. Now this is really interesting.

Check this out. There's a big-ass bridge rectifier here on the output board. Now, Is it actually doing any bridge rectification? Well, let's follow, follow the money and see what we're doing. This is the positive pin here.

Okay, the two middle ones you can see there. the AC input and the negative pin is over here. Let's take a look at this positive pin. The positive pin here is connected to this terminal here, which is I've actually traced out.

This is actually the negative output terminal. Okay, so the negative output terminal to the positive pin of the bridge rectifier. Look at these two AC input pins. They're going nowhere.

They flapping around in the breeze and I've actually confirmed on the top. You may not be able to see that, but I've confirmed that those pins actually go nowhere. Okay, so they're completely isolated and then the negative terminal of the bridge rectifier is going to over to here to our negative output. Right over here to our negative output terminal here.

What's going on? Let's go to: Dave CAD So what they're doing here, here's the negative output, terminal on? the back. and they're actually basically put in two series diodes like this in and then two of those in parallel because these AC terminals here are just floating. Okay, they're not connected to anything. so you basically got two diodes in series two in series and then those in parallel going down to the negative terminal if your power supply output.

So yeah, all of the output current is flowing through this diode bridge. Why they're doing that in the negative line like that? I don't know some sort of sense thing. reverse protection somehow and it doesn't actually make sense until you have a look at the second party which looks like another for terminal bridge rectifier. but it's not.

It's actually a high value for terminal current shunt resistor. Uh-huh Let's go back to Dave Ken And Bingo! Our sense resistor is in parallel with the diode bridge here. so this is our output current sensor resistor. They're doing low side current sensing.

It's 4 terminal of course to get the accuracy because these things like point 1 percent accurate on the current over the full range I Believe! So yes, so that's going these two. This sense terminal is going off to a connector which is then going back down to the board and they're measuring that with a diff amp is not on this output board. it's actually back on the main board so that's a bit. Yeah, you doing almost like that.

I Like having my my differential sensor my defect right there. You know you don't want to go in three cables. Mm-hmm Anyway, they've done their homework. They know what they're doing.

So this big ass bridge rectifier is basically protecting the current sense resistor here. and you can see that there they are actually sensing the output voltage right on the output connectors here. This goes back to a connector here which is wired back to the main board. So yeah, you know they're doing it right at the connector.

so they're doing it properly. But of course if you want to sense out the load because this is, well, this is not a huge current I Mean it's eight. You know what is it? Six eight eight amps or something with you know a decent amount. but the ones that are like you know, 400 amps.

Absolutely crazy. You know you want to sense at the load, you don't want to drop it across your wires so you can still do that with this. but at least they're sensing right on the output terminal. so you know your voltage and your output terminals is exactly what you're getting.

And of course this is the As I said high Voltage eight Amp model for the Low Voltage 400 amp model I mean they've probably got a different interconnection out bored here. different output terminals. The whole works so you know they still got the busbar arrangement here. But yeah, they're going to have much higher current capability.

I Mean they say you know 8 amps is? yeah, you know, in the scheme of like you know, high current power supplies is, you know, and nothing. it's down in the noise, you know it's so you know there's an output board like this is adequate. but is this big thing? Yeah, yeah. the other models could be significantly different in their output configuration and that's for this.

GPIB Ethernet and USB board. Well, it looks like a complete kludge here on the top with the lose copper shield in an adhesive copper shielding tape here all over the damn thing. So you know, like did it Oopsie, not pass. Not past their requirements and they did that as an afterthought.

and nice cutouts by the way on the tan alums there, but ya don't like that at all. And yep, they'd be our GPIB drivers on the back. probably I'm not going to bother looking at the number, but yeah, they said this would be an Agilent job. No doubt, we're not talking TDK anymore.

And as for the main processor board under here, well looky what we have here. Old school, we're talking Phillips Isthmus /n XP Um, eighty nine, Five One Eight Oh Five one. Woohoo! in a PLC SI Saket I'm bit of a Plc SI Saket fanboy always loved him. so you've got to think you know how old is this design.

What is a date back to? I Don't know the previous generations of this system power supply, but yeah, it's probably not fairly recent. probably not even the century you'll notice. There you go the little model number ID 600 volts circled as I said from eight volt models up to six hundred. So that's not just a matter of a firmware change, Otherwise, they wouldn't probably wouldn't bother having that on the board.

they just work a sticker on the chip or something like that. By the way, that Micro is our 64 K a flash, fancy-pantsy 805 one, let me tell you. But yeah, so they've obviously got some hardware changes in terms of you know, dividers and you know ADC ranges and you know, like stuff like that voltage sensing stuff so it's going to be significantly different between the different models. So there you go I Hope you enjoyed a look inside this keysight 50 100 watt system DC power supply and well I know what everyone's thinking Oh bet you haven't fixed it Dave I know I haven't sorry I've already got 40 minutes worth of shot footage on this thing and it's what is it.

It's now 5:17 on a Friday afternoon and yeah, I got to go. So that's it. So I won't get a video out this week because I've been extremely busy so. but oh I have to do a follow-up video with this actually? Yeah, I have to take this board out.

Have to replace the my board or take it out. Don't necessarily have to have to replace it to get it actually working. it's just there info in in there for full protection. So when you're actually you know if you power this thing up under controlled conditions, then you know exactly what you're doing.

You know it's fine to actually do it without that so we don't need a replacement move are working to lower current our fuses down here to protect it. I'll pair it up, feed the same phase into all three channels down here, and look for other visual stuff once they get the board out. Maybe there's some track charring on the bottom of the board? perhaps where there's you know being over current but can't see anything on the top? so I don't necessarily expect anything on the bottom. So yeah, I can't see any other visual science probably need to give it.

Maybe somebody's screaming at me, they're watching in HD and they saw something that's blowing that I didn't catch by looking at the I can caught a screen here. but anyway. um so yeah, maybe we can repair this puppy once again. If anyone's got schematics or a service manual for this thing, please link it in.

So I Hope you like that. If you did, please give it a big thumbs up. Catch you next time you.