Xantrex/Sorensen XHR 40-25 schematic?

[ifonlyeverything]

I've recently come into possession of a non-working Xantrex/Sorensen XHR 40-25 power supply. Does anyone happen to have a schematic? I tried searching Google but I can only seem to find an operating manual which doesn't include a schematic. The XKW series service manual/schematic is available but just looking at the board reveals significant differences...

[coromonadalix]

EDIT   see further away

so far  nothing from xantrex has ever leaked, or maybe some  as you suggest,   witch is rare ...

apparently  maybe  some lambda TDK  may have some similarities,  not sure for Sorensen

[ifonlyeverything]

That's unfortunate, I'll look around for anything TDK that might be comparable.

For anyone who has never seen the inside of these.... very EFFICIENT use of space near the output. 

[ifonlyeverything]

If anyone is familiar with these beasts, I'd appreciate some hints on how the main DC-DC topology works. I believe there's a PFC boost upstream. I can see the mains rectifier and then two FETs and two diodes in parallel on the PFC side. On the downstream DC-DC I see four FETs with bootstrapped gate drivers (IR2110), followed by the transformer, output inductor, and output capacitors. But I'm not seeing any rectifiers?

[coromonadalix]

some power supplies use some kind of a buck boost / buck down configuration / pwm  configuration like / or ac phase control and have another section called the main regulator section

a mix of switching and linear modes ...  not sure if xantrex ones use something like this ??

on the bus bar you have a big black rectangle component,  could be an scr, triac  or bridge rectifier or even an big IGBT   witch would made sense  with the nearby ic's, and the thansformer who is in reality an huge inductor seems to confirm the operation


the linfinity  sg3524 is a classic  pwm ic
the uc3842 is a 500khz pwm ic too

IR2110/IR2113 are high voltage, high speed power MOSFET and IGBT drivers

[ifonlyeverything]

on the bus bar you have a big black rectangle component,  could be an scr, triac  or bridge rectifier or even an big IGBT

Wow, nice call. I thought it was just some sort of insulator but I checked it with my multimeter and it shows a 0.33V forward voltage on diode mode. So it apparently works, I guess, but I'd like to know WHAT part it is. Of course the identifier is on the fucking bottom, where it can't be seen without taking off the (+) bus bar... which means breaking the EMI ferrite glued around the bus bar in half and de-soldering and unbolting the clusterfuck that is downstream of the EMI ferrite. The engineers definitely did not design this with serviceability in mind.

[edpalmer42]

Here's the service manual for the XHR series.

http://www.ko4bb.com/getsimple/index.php?id=download&file=06_Misc_Test_Equipment/Xantrex/Xantrex_XFR_Power_Supply_Manuals_and_NI_Drivers/Xantrex_XHR_Xantrex_XHR_600W_1KW_Service_Manual_TM-XR01-XN.pdf

Ed

[ifonlyeverything]

Here's the service manual for the XHR series.

http://www.ko4bb.com/getsimple/index.php?id=download&file=06_Misc_Test_Equipment/Xantrex/Xantrex_XFR_Power_Supply_Manuals_and_NI_Drivers/Xantrex_XHR_Xantrex_XHR_600W_1KW_Service_Manual_TM-XR01-XN.pdf

Ed

THANK YOU. 

edit: The manual doesn't disclose the exact part # but their internal part # (CR-2040-PT) and description (ultrafast 200A 400V) appears to correspond to MURP20040CT.

[coromonadalix]

basicly  the psu topology is:  they rectify the AC,  PFC control/stabilize it in some cases, and with this huge DC line they PWM it ...


oh ho    check  "xantrex" on the given website  there is more   
i did  use xantrex psu's in the past,  very efficient and compact, while being heavy  loll

years ago  when they broke, it was a dud to repair them, now i see  that we can do repairs at some point  ... i admit  i never redone any search  my bad 

and you seem to have some software too  for visa or control ??

[Chris56000]

 K04BB are taking a hell of a long time to sort out the manuals from their "Recent Uploads" directory – some of the files in there haven't been made available for download in over two months!

Chris Williams

[edpalmer42]

K04BB are taking a hell of a long time to sort out the manuals from their "Recent Uploads" directory – some of the files in there haven't been made available for download in over two months!

Chris Williams

Remember that KO4BB isn't a group or a company.  It's a guy who pays for the storage space and bandwidth on his own and maintains the system in his spare time when the rest of his life allows it.  All things considered, I think he's doing a great job!

By the way, when did you last send a donation to help him cover his costs?  I sent him $50 a while back and he seemed very grateful.  Makes me wonder if that was an unusual occurance.

Ed

[armandine2]

If you don't want a scan, another link, worked for me - manuals - but no schematics

https://www.programmablepower.com/

https://www.programmablepower.com/products/dc-bench-power/xhr-series



[edited to manufacturer's links]

[coppercone2]

lol check those sockets my sorensen (30V30A) had a bad socket on a op amp , even after I resocket it and clean the chip out etc.

nonsense happens until you push on the chip then it worked fine. if just wanna try some random whatever replace those socket
put some berylium copper or bronze pin sockets in there

i wonder if it could be because they got bus bar and big heat sinks like on the PCB

you can also disconnect the limiter circuitry totally (trimpot stuff) by disconnecting a resistor maybe. it should work fine. gets rid of that variable


as usual check the rails on the IC first before you start messing with the high current portion. I also took mine apart and cleaned/greased  the bus bars and bolts etc with deoxit grease or liquid depending on ease

but If I had to do another sorsensen switcher, I would literarly replace all the sockets before working on anything else, and I would not bother trying to test the socket (probobly lies when you probe it, mine did)... the whole thing with the probe on the bottom of the leg or under the PCB and a probe on the chip to measure continuity is BS for dodgy sockets that is not reliable troubleshooting IMO. i know there is some socket master voodoo priests in old computer repair but just IMO fucking replace that BS its not worth trying to analyze, you aint trying to preserve the vintage of some atari lol

here, take this

https://www.digikey.com/en/products/filter/sockets-for-ics-transistors/ic-sockets/409?s=N4IgjCBcoGwJxVAYygMwIYBsDOBTANCAPZQDaIALGGABxwDsIAuoQA4AuUIAyuwE4BLAHYBzEAF9CtAMyIQKSBhwFiZENOlwwAJkaENW7duZtOkHv2FjJIAKx05CpXkIlI5CgAZbnz8ZYgHFy8gqISUhqejmhYLqrulNT02hQg%2BjQ09GAI6TS22jRp6hkU%2BSaBZhah1lL0WtGKsSpu5NI%2BdTBFXgXSEFLlQeYhVuHgdbLQ8jHKrmq9cBRZRdI0MLYUCAGDVSM2YHWMk05Nswnzi33FaxsDlcNhNgC0xkdQ-ACuzWq2zOI2P5NWFAwGwgZAfDZOpMBAATLhgSJFbb3MSEdgAT1YuC4MOwKD%2BQA

[ifonlyeverything]

Cleaned up the PCB a bit, didn't see any blown FETs, PCB scorch marks, or blown fuses. Need to get a light bulb limiter set up before I start investigating the power electronics further, and I think I want to test the control board separate since it's low voltage stuff (or just build an op amp test jig since it's all socket mount.) coppercone2 I'll be sure to focus on the sockets, some of them do look a little janky. Also managed to drop a fucking screw between that output capacitor clusterfuck, so I had to pull a few caps off to get the screw out -- needle nose pliers just jammed it in deeper. Might as well replace the output caps while I'm at it. 

In the meantime I'm still studying the schematic, trying to understand how it all works. I noticed the marketing material claims ZVS. Is this actually ZVS/soft switching? I'm no SMPS expert but the full bridge IC is just an SG3524 and I'm not seeing any evidence of any sort of ZVS detection circuitry... unless I'm missing something obvious?

[coppercone2]

what was the problem with it to begin with? Not really sure how they work too well. I kinda figured out the feedback loop and figured where the fault might be.

but... power rails ok?

and my socket looked fine, it even removed OK, its just messed up. those socket suck because they are the 2 leaf things not the little hole

but you got a much harder supply because its 1kV. Mine is 30V. Anything over like 100 volts makes it mad complicated. 1kv exceedingly so. new failure modes for parts.  Even resistors can do that shit where they start failing at a high voltage like 750 volts.

Also mine had a mains connector that was not soldered in! it was press fit in for most of its life with almost no solder on the pin. It started arcing on me after I fixed the unit, luckily all it needed was a dash of solder. Sorensen QC must have had a bad day. I had enough HV shit since I fixed the miller dynasty welding machine (it has like 800V rail). Weird shit was going on there, far away insignificant looking components like ceramic caps causing problems.

Honestly the way to test this would be to desolder each part, bias it with HV supply, make sure its OK.. anything on the HV rail IMO can be really suspect but I don't have much experiance with that.

[coppercone2]

lol if someone made me fix this, I would replace all the silicon on this thing and check all passives under HV and hope it works. stomach ache from thinking about the schematic . the only thing I know is that I would have to buy extra pace desoldering tips for this job

but for BS parts like 317, 347, gates, 351... its not really big money if you are patient and replace. that might give confidence enough to feel like you can trouble shoot it. for the gates that are like 10$ on ebay breadboard test it, but honestly with a desoldering tool, like a 1.5$ op amp.. Id just fuckin replace it.

if you can get it to a specific fault, it might limit the amount of trouble shooting.

probobly the diodes too.

only tip is clean the board from flux as you go along but not too clean and do a final clean when you are done

power transistors too, if they are cheap, replace IMO. If you gotta import it from bulgaria test it

if you do this crazyness, i recommend printing out the entire part list, getting two color of highlighter, and marking parts that might be suspicious with 1 color and another color if tested or replaced. but you need to test the HV parts with some voltage probobly or it might be another unknown.

[ifonlyeverything]

what was the problem with it to begin with? Not really sure how they work too well. I kinda figured out the feedback loop and figured where the fault might be.

but... power rails ok?

and my socket looked fine, it even removed OK, its just messed up. those socket suck because they are the 2 leaf things not the little hole

but you got a much harder supply because its 1kV. Mine is 30V. Anything over like 100 volts makes it mad complicated. 1kv exceedingly so. new failure modes for parts.  Even resistors can do that shit where they start failing at a high voltage like 750 volts.

Also mine had a mains connector that was not soldered in! it was press fit in for most of its life with almost no solder on the pin. It started arcing on me after I fixed the unit, luckily all it needed was a dash of solder. Sorensen QC must have had a bad day. I had enough HV shit since I fixed the miller dynasty welding machine (it has like 800V rail). Weird shit was going on there, far away insignificant looking components like ceramic caps causing problems.

Honestly the way to test this would be to desolder each part, bias it with HV supply, make sure its OK.. anything on the HV rail IMO can be really suspect but I don't have much experiance with that.

Previous owner said it would click on and off repeatedly with no apparent power. There's only one relay on the PCB (K139) which switches in/out 20 ohms of resistance to slow inrush current. I however didn't plug it in (yet.) Hopefully the problem is isolated to this relay, its power supply or its coil control logic, because these sections don't look TOO daunting.

Mine is only the 40V 25A model, so no worries about high voltage.

[coromonadalix]

you could be surprised,  dont assume thet psu is rinning on low voltages

normally the 120vac is rectified to at least 160vdc or doubled 320 347vdc  to create a stable dc supply to be  pwm'ed to the control stage ...

[coppercone2]

oh yeah but its not a 1kV regulated output stage thankfully

[ifonlyeverything]

you could be surprised,  dont assume thet psu is rinning on low voltages

normally the 120vac is rectified to at least 160vdc or doubled 320 347vdc  to create a stable dc supply to be  pwm'ed to the control stage ...

Oh yeah definitely, the PFC boost stage looks like it's putting out close to 300V.

[coppercone2]

ok so if the relay is clicking on and off what does that mean.. how to recreate the problem


q215 is a low side driver for the relay. when it closes it energizes. something is causing that to open.u228a.

this is labeled overcurrent

the non inverting input is the input to this relay, the other shit is tied to rail. if rails are fine, so it has some shit going to ground, but the signal comes from r221 475k, which is tied to dc negative master, so the 18V rail goes through 1k1 to opamp input.

that op amp looks like its going into a rC delay for Q215.

so is this just a timer here? i think its also wired to the over voltage or whatever from the chip that connects to the shunt and the PSU

r294 to 297 is maybe a shunt for measuring current, goes into Isense on the chip

yeah idk its getting really complicated.'

But that is a clue, if its going over current protection, that might mean a shorted transistor some where I think... that chip might be measuring fine, and bad transistors are really common on power supply.


you can turn it on and probe that chip to see if its doing what its supposed to be doing. like measuring current and voltage

I think maybe just testing those transistors after the relay might be a good start, to look for damaged power transistors, possibly past the transformer. possibly at the output. i have seen more bad power supplies with just damaged power transistors then damaged control circuits, but I have seen power supplies that have damaged transistors and also control circuits. but the first one is more likely. and its likely if you find a bad transistor the control circuit for it might be damaged not some irrelevant one.

like sorensen has a good name, its considered reliable and high end, more then likely you don't have some deep control fault

it might be a transistor that is also connected to ground.like q114 h bridge


Q166 is like a crowbar or whatever, that abruptly sinks power if there is a fault to protect the output. this is the one you can disconnect and have a functional circuit still. r146 is like the front panel adjuster for the trip point.

test everything semiconductor in the top 1/5 of schematic page 95

[coromonadalix]

Since its driven by u228a   

you may have some overcurrent situation / something is current consuming at some point  OR
its friend with the under voltage ic u228B  .... meaning you seem to / may have a triggering condition

Not sure about r317 r318,  if they act as a slight inrush current limitter or "voltage dropper"  and all of it is before the current sense resistors ... for the primary section

You have r221  who goes near r317 and goes to this ic u228a,  for sure its a sense line return ...

-----

yeah  its a tough cookie  psu  loll,  you have some pre-regulation circuit  and the power circuit

you have 3 power supply sections,  you have the small one (u64) who seems to give gives some clock to U21 ,  and you have its output line who goes on U287 opto isolator,  who goes to U250

U287 must act as an ok supply to start U250   or some clock pulse is driven ??   maybe the 50-60hz detection ??  but it seems autonomous ??       man 

you have to decypher circuit parts and try to figure out how they inter connect,  in some cases say :  you isolate 1 section and subs it with another dc source and try to see where it fails

you have p301  test points ? this is the dc voltage of the 1st power section  ...

and the  ususals checks for all the small voltages in this psu .. when its possible ...   there could be a domino effect,  one affect the other .... 

[factory]

These are known as the Glassman LVP series over here, they are 1kW not kV.
We had about a dozen rack mount version at 150V 7A at work, can't remember any dying, the fan packed up in one.

There was however a very nasty fault with one of them, the voltage set pot would jump the output to maximum, when it got to a dead spot on the ten turn pot. 

There is a bigger 150V 20A version & a newer 300V 9A unit at work, those can burn stuff very nicely, if you have a fault on the unit under test. 

David

[Chris56000]

Hi!

Update to my previous post – Manuals on k04bb now available for download!

Chris Williams

[NikiG]

Hello, you maybe have problem in two components in corner?