Author Topic: Fixing/Replacing bad transformer in logic analyzer SMPS.  (Read 16262 times)

0 Members and 1 Guest are viewing this topic.

Offline ArtemisGoldfishTopic starter

  • Regular Contributor
  • *
  • Posts: 58
Fixing/Replacing bad transformer in logic analyzer SMPS.
« on: August 29, 2010, 05:27:15 pm »
Story time!

I was poking around my boss's office looking for a heat gun to solder a no-lead package IC onto a board when I saw a big, rather old logic analyzer (80 channel, HP 1652B with built-in oscilloscope.) Anyways, he was like, "I was using it about 3 months ago and suddenly magic blue smoke came out, burning electronics smell came from it, and it died." Curious, I thought. Probably the power supply, and the power supply is the bread-and-butter of a technician's skills. I was hoping it'd be linearly regulated for simplicity, but it's an SMPS = Whatever.

Anyways, I took it home, plugged it in, and it turned on and seemed fine for a few minutes. Then, no magic smoke, but it turned off and had burning electronics smell. I isolated the cause to be the power supply, it kept making a clicking noise and emitting the burning smell, so I asked around and it seems like it's a bad transformer.

TL;DR: Bad transformer in switching power supply.

The OEM which made the power supply doesn't seem to exist anymore (Boschert), and it's very difficult - even on eBay and the like - to find a replacement HP power supply (they were also used in 1653Bs and some oscilloscopes, I think 54501A uses it, but those are also kinda expensive.

So, how feasible is it to actually re-wind a transformer for a power supply? That seems like my only real option at this point, really.

Thanks much.
John, Hardware Technician, F5 Networks
 

Offline Varal

  • Contributor
  • Posts: 23
Re: Fixing/Replacing bad transformer in logic analyzer SMPS.
« Reply #1 on: August 29, 2010, 06:02:16 pm »
Wouldn't it be easier to simply buy a new transformer? (I assume that if you can read who made it you can also read other info like current or voltage)

Not that I neglect the educational value of rewinding a transformer but getting a new one seems safer and gives a bigger chance for trouble-free operation. :)
 

Offline NiHaoMike

  • Super Contributor
  • ***
  • Posts: 9331
  • Country: us
  • "Don't turn it on - Take it apart!"
    • Facebook Page
Re: Fixing/Replacing bad transformer in logic analyzer SMPS.
« Reply #2 on: August 29, 2010, 08:04:02 pm »
What voltages does the power supply output? There's probably an off-the-shelf unit that can be modified, or even used as-is.
Cryptocurrency has taught me to love math and at the same time be baffled by it.

Cryptocurrency lesson 0: Altcoins and Bitcoin are not the same thing.
 

Offline mikeselectricstuff

  • Super Contributor
  • ***
  • Posts: 14133
  • Country: gb
    • Mike's Electric Stuff
Re: Fixing/Replacing bad transformer in logic analyzer SMPS.
« Reply #3 on: August 29, 2010, 09:58:51 pm »
Transformer itself seems fairly unliklely - more like something else making it dissipate too much power. Electrolytics are a common failure mode - if the main HV reservoir cap is dodgy it's conceivable the extra ripple is causeing problems.
You won't find a replacement - SMPS transformers are invariably custom made for a particular PSU.
Rewinding is rarely likely as the cores are generally glued together.
Best bet may be fo find what voltages the PSU outputs and replace the whole thing.
While you can still get it running, measure all the PSU output voltages and currents - if it completely dies it will be harder to figure them out.
Post up a hi-res photo of the PSU here & maybe we can figure out some details.
May also be worth positing on Agilent's forum as there are some helpful ex-HP guys there :
https://forums.tm.agilent.com/community/viewforum.php?f=78

Youtube channel:Taking wierd stuff apart. Very apart.
Mike's Electric Stuff: High voltage, vintage electronics etc.
Day Job: Mostly LEDs
 

Offline ArtemisGoldfishTopic starter

  • Regular Contributor
  • *
  • Posts: 58
Re: Fixing/Replacing bad transformer in logic analyzer SMPS.
« Reply #4 on: August 29, 2010, 10:45:43 pm »
The PSU outputs, if I remember correctly, 12V, -12V, 5.2V, -5.2V, and possibly something like 3.3V and -3.3V.

To be more vivid, whenever the power supply is plugged in and on, it makes the clicking noises, and a very slight sizzling sound. However, the transformers don't get hot at all, but I don't know if that helps at all.
John, Hardware Technician, F5 Networks
 

Offline retched

  • Newbie
  • Posts: 3
  • Nothings Shocking
Re: Fixing/Replacing bad transformer in logic analyzer SMPS.
« Reply #5 on: August 30, 2010, 12:52:01 am »
I would probe the transformer output.  If the windings dont look burnt, and you are getting proper voltage, I would move on.

This clicking/sizzling could be a few things.  The caps are probably older than dirt,and by not reforming or replacing them prior to use, you could have done other damage.

The clicking you mention.  Is it the standard o'scope relay clicks?  It may be a short to the coil or a fused diode from a spike from super leaky caps.

I would pull and replace the caps after probing the secondary of the transformer.

If your lucky the bad caps caused the relay to smoke and arc from bouncing or operating the relay coil at a higher frequency than intended.
-rm-
 

Offline ArtemisGoldfishTopic starter

  • Regular Contributor
  • *
  • Posts: 58
Re: Fixing/Replacing bad transformer in logic analyzer SMPS.
« Reply #6 on: August 30, 2010, 01:03:17 am »
I would probe the transformer output.  If the windings dont look burnt, and you are getting proper voltage, I would move on.

This clicking/sizzling could be a few things.  The caps are probably older than dirt,and by not reforming or replacing them prior to use, you could have done other damage.

The clicking you mention.  Is it the standard o'scope relay clicks?  It may be a short to the coil or a fused diode from a spike from super leaky caps.

I would pull and replace the caps after probing the secondary of the transformer.

If your lucky the bad caps caused the relay to smoke and arc from bouncing or operating the relay coil at a higher frequency than intended.

The clicks come from the power supply itself (which contains no relays to speak of), completely disconnected from the test equipment. I tried probing the voltage outputs and turning it on (which subsequently blew out the input fuse), and all that came through the voltage outputs were some rather high-frequency (so, probably switching frequency) pulses. This makes me even more suspicious about the capacitors. Maybe this will give me a good excuse to buy an ESR meter of some sorts.
John, Hardware Technician, F5 Networks
 

Offline retched

  • Newbie
  • Posts: 3
  • Nothings Shocking
Re: Fixing/Replacing bad transformer in logic analyzer SMPS.
« Reply #7 on: August 30, 2010, 02:08:34 am »
This is a standard, text book, cap problem.

Are the caps leaking any dielectric liquid or bulging at all?

Regardless, with the age of the unit, the caps NEED to be replaced.  Reforming such old caps CAN be done, but it isn't worth it these days.  The price of higher rated caps are so inexpensive. (Not that you need higher rated caps (voltage wise))

You would still have to pull them to reform them, so might as well replace them.

I am interested to see how it plays out.

And just in case you are thinking of powering it up again before changing the caps, re-watch the EEVBLOG episode where Dave and friend blow up some electrolytic caps.

You don't want that happening in your face...or house. ;)
-rm-
 

Offline ArtemisGoldfishTopic starter

  • Regular Contributor
  • *
  • Posts: 58
Re: Fixing/Replacing bad transformer in logic analyzer SMPS.
« Reply #8 on: August 30, 2010, 02:37:21 am »
Sounds good, I suppose. I'm going to shop around Digikey and price fully recapping the entire supply (30 electrolytic caps.) Hopefully won't be too bad.

Edit: About $31 US. Not too bad, guess I'll order it tonight so I'll get it in a few days. Replacing the bundle of caps in this thing with brand-new Panasonic - ECG caps, which are quite reasonably priced. Most of the old caps in this thing are Chemicon and Rubycon.
« Last Edit: August 30, 2010, 03:10:16 am by ArtemisGoldfish »
John, Hardware Technician, F5 Networks
 

alm

  • Guest
Re: Fixing/Replacing bad transformer in logic analyzer SMPS.
« Reply #9 on: August 30, 2010, 04:50:51 am »
Make sure to get low ESR caps for switching power supplies, otherwise they'll die again soon.
 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 20529
  • Country: gb
  • 0999
Re: Fixing/Replacing bad transformer in logic analyzer SMPS.
« Reply #10 on: August 30, 2010, 07:40:21 am »
I'd be surprised if none of the semiconductors have been damaged.

The switching transistors must be fine, otherwise it would've blown the fuse every time the power was applied. If recapping doesn't solve the problem, then I'd check some of the semiconductors, perhaps the rectifier on the secondary could be faulty.
 

Offline djsb

  • Frequent Contributor
  • **
  • Posts: 998
  • Country: gb
Re: Fixing/Replacing bad transformer in logic analyzer SMPS.
« Reply #11 on: August 30, 2010, 08:12:53 am »
Find out what the problem is THEN replace parts.

David
David
Hertfordshire, UK
University Electronics Technician, London, PIC16/18, CCS PCM C, Arduino UNO, NANO,ESP32, KiCad V8+, Altium Designer 21.4.1, Alibre Design Expert 28 & FreeCAD beginner. LPKF S103,S62 PCB router Operator, Electronics instructor. Credited KiCad French to English translator
 

Offline mikeselectricstuff

  • Super Contributor
  • ***
  • Posts: 14133
  • Country: gb
    • Mike's Electric Stuff
Re: Fixing/Replacing bad transformer in logic analyzer SMPS.
« Reply #12 on: August 30, 2010, 11:19:23 am »
Find out what the problem is THEN replace parts.
David
Although this is a reasonable general rule, cap failures in SMPSs are so common that replacing them regardless, especially in very old kit, is often a very paragmatic solution and can be much quicker than doing detailed faultfinding.
Bad caps can produce multiple simultaneous faults, which are pretty much impossible to faultfind in a tightly coupled system like a SMPS.

There is a particularly common failure mode causing to SMPSU not to start due to a dried cap in the startup circuit, and so common that when presented with this symptom I replace the startup cap without doing anything more then checking the fuse - 90% of the time this fixes it.

ArtemisGoldfish - before replacing everything I'd suggest you look for a low-voltage cap on the mains side of the PSU and replace that  - there is often only one, and this is prime suspect. It will usually be connected via a high-value (100K+) resistor to the +ve side of the reservoir cap - check the value of this resistor, as they can go high or open due to extended stress of high voltage.


 
Youtube channel:Taking wierd stuff apart. Very apart.
Mike's Electric Stuff: High voltage, vintage electronics etc.
Day Job: Mostly LEDs
 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 20529
  • Country: gb
  • 0999
Re: Fixing/Replacing bad transformer in logic analyzer SMPS.
« Reply #13 on: August 30, 2010, 01:04:36 pm »
Yes by the time you've wasted doing all the fault finding for capacitors you might as well have just replaced them.
 

Offline ArtemisGoldfishTopic starter

  • Regular Contributor
  • *
  • Posts: 58
Re: Fixing/Replacing bad transformer in logic analyzer SMPS.
« Reply #14 on: August 31, 2010, 03:34:30 am »
Find out what the problem is THEN replace parts.
David
Although this is a reasonable general rule, cap failures in SMPSs are so common that replacing them regardless, especially in very old kit, is often a very paragmatic solution and can be much quicker than doing detailed faultfinding.
Bad caps can produce multiple simultaneous faults, which are pretty much impossible to faultfind in a tightly coupled system like a SMPS.

There is a particularly common failure mode causing to SMPSU not to start due to a dried cap in the startup circuit, and so common that when presented with this symptom I replace the startup cap without doing anything more then checking the fuse - 90% of the time this fixes it.

ArtemisGoldfish - before replacing everything I'd suggest you look for a low-voltage cap on the mains side of the PSU and replace that  - there is often only one, and this is prime suspect. It will usually be connected via a high-value (100K+) resistor to the +ve side of the reservoir cap - check the value of this resistor, as they can go high or open due to extended stress of high voltage.


 

Here's a picture. The mains connection is the reddish connector on the top right of the photo. Is the cap circled the low-voltage cap you're talking about?

Also, shotgun replacing all (or most) of the caps isn't a huge deal for me, my time is cheap and I have the money to do it. Might as well, right? :)

Anyways, speaking of which, can anyone recommend any good texts on troubleshooting SMPSes? Or possibly just one on SMPS design in general?
John, Hardware Technician, F5 Networks
 

Offline NiHaoMike

  • Super Contributor
  • ***
  • Posts: 9331
  • Country: us
  • "Don't turn it on - Take it apart!"
    • Facebook Page
Re: Fixing/Replacing bad transformer in logic analyzer SMPS.
« Reply #15 on: August 31, 2010, 05:01:07 am »
http://repairfaq.org/sam/smpsfaq.htm
As usual, the old HP power supplies seem to be built quite well. The new ones are the poorest quality I have seen.
http://ecomodder.com/forum/showthread.php/another-use-kill-watt-10959.html
Cryptocurrency has taught me to love math and at the same time be baffled by it.

Cryptocurrency lesson 0: Altcoins and Bitcoin are not the same thing.
 

Offline mikeselectricstuff

  • Super Contributor
  • ***
  • Posts: 14133
  • Country: gb
    • Mike's Electric Stuff
Re: Fixing/Replacing bad transformer in logic analyzer SMPS.
« Reply #16 on: August 31, 2010, 09:02:07 am »
Quote
Here's a picture. The mains connection is the reddish connector on the top right of the photo. Is the cap circled the low-voltage cap you're talking about?
Probably. Anything rated <100v on the mains side - there is usually only one & looks like it from the pic.
[/quote]
Also, shotgun replacing all (or most) of the caps isn't a huge deal for me, my time is cheap and I have the money to do it. Might as well, right? Smiley
Quote
It won't hurt, but IME the only cap that ever gives trouble is the startup cap - I don't recall ever having to replace any other cap in a SMPSU.
It's the combination of often being close to the dropper resistor, which can run quite hot, and also being sensitive to leakage due to the relatively low charge current from the dropper resistor. 
Don't forget to check the dropper resistor value whle you're at it.

 
Youtube channel:Taking wierd stuff apart. Very apart.
Mike's Electric Stuff: High voltage, vintage electronics etc.
Day Job: Mostly LEDs
 

Offline ArtemisGoldfishTopic starter

  • Regular Contributor
  • *
  • Posts: 58
Re: Fixing/Replacing bad transformer in logic analyzer SMPS.
« Reply #17 on: September 02, 2010, 05:08:16 am »
I want to thank everyone who contributed to this thread. You are all fantastic :)

I got my order in from DigiKey. It took about 4 hours to desolder the old caps (a lot of them were in cramped places, which took longer than it should've), a significant number of them were bulged on the bottom and leaking electrolyte onto the board. After cleaning up the board and putting new pretty Panasonic caps on it, the logic analyzer sprung to life. The number of leaking caps I found? About...40% of them. Mostly 2x4 group of them in the middle of the supply.

Anyways, the device is so much more agreeable now. Here's some action shots. The oscilloscope is...weird. As my boss put it, "The oscilloscope part is an oscilloscope made by logic analyzer people, so it's weird."

Anyways, thanks again :)
John, Hardware Technician, F5 Networks
 

Offline DJPhil

  • Frequent Contributor
  • **
  • Posts: 511
  • Country: 00
Re: Fixing/Replacing bad transformer in logic analyzer SMPS.
« Reply #18 on: September 02, 2010, 06:07:33 am »
Congratulations!

Also, it's good to see an old Eico 377 in service. Much love for those. After a long battle with one I finally had to gut it for parts, but the chassis will live on with more modern Wein bridge oscillator soon.
 

Offline snandu13

  • Newbie
  • Posts: 3
Re: Fixing/Replacing bad transformer in logic analyzer SMPS.
« Reply #19 on: May 11, 2012, 07:05:52 am »
Hi,
           This is snandu13. I am new to this blog. I have the very same power supply which is problamatic. This is used in 54501a Digital Oscilloscope HP.
            The Fuse blows up the moment it is switched on. I do not have schematic. I am not sure whether the low ESR caps will be the issue.
             There is also another module attached vertically to this main board.
             I would greatly appreciate if someone can help to resolve the problem.
             
 

Offline Rerouter

  • Super Contributor
  • ***
  • Posts: 4706
  • Country: au
  • Question Everything... Except This Statement
Re: Fixing/Replacing bad transformer in logic analyzer SMPS.
« Reply #20 on: May 11, 2012, 07:38:30 am »
someone earlier in this thread mentioned swapping the main switching transistors if the fuse was blowing, however it wouldnt hurt to use an ohm meter around some of the capacitors, with years of non-use most electrolytics expecially older ones risk internally shorting and would need to be reformed or replaced, an ohms check just helps to see if its the case, and if you cant afford to replace all of them, then you can generally get an idea on which have,

still perhaps post an image unless its an exact match, if so then.. hopefully someone else can point the switching mosfets in question out
 

Offline snandu13

  • Newbie
  • Posts: 3
Re: Fixing/Replacing bad transformer in logic analyzer SMPS.
« Reply #21 on: May 24, 2012, 06:41:11 am »
I am snandu13 again.  I am still unable to fix  HP Power supply used in Digital Oscilloscope 54501A.

The fuse continues to blow the moment I power on. I removed the Bridge and checked. The AC voltage is present at Bridge input.
I checked the switching transistor ( TIPL755A ) after removing. It is OK.

There is a thermal reed switch MTS90A. I am not sure whether the switch should be initially shorted condition. This is across B-E Junction.

 Can someone help, please? I am desperate.
 

Offline vk6zgo

  • Super Contributor
  • ***
  • Posts: 7872
  • Country: au
Re: Fixing/Replacing bad transformer in logic analyzer SMPS.
« Reply #22 on: May 24, 2012, 08:56:10 am »
What is the fuse that blows?
If it is the  AC mains fuse,have you tried disconnecting everything past that point,then restoring them one by one,till the fuse blows again?
If it is further down the track,then you can say that anything further on than that fuse is suspect.
Without a schematic,it is a long hard slog to fix things,but the important thing is to be systematic.
Taking guesses may work when you have more experience,but step by step is the way to go now.
 

Offline snandu13

  • Newbie
  • Posts: 3
Re: Fixing/Replacing bad transformer in logic analyzer SMPS.
« Reply #23 on: May 25, 2012, 04:00:42 am »
Thanks for replying

The mains fuse provided in the input selector switch blows. It is 3 Amps.
 
 I removed the bridge which comes after common mode filters, Metal Oxide Varistors. Then the fuse does not blow. The 230 V ac comes to bridge input.

 Unfortunately I searched for schematics in the net. I could not locate the same.

snandu13
 

Offline HackedFridgeMagnet

  • Super Contributor
  • ***
  • Posts: 2039
  • Country: au
Re: Fixing/Replacing bad transformer in logic analyzer SMPS.
« Reply #24 on: May 25, 2012, 06:09:48 am »
Fixing mains powered switched modes supplies is potentially hazardous.
I hope you can recognise where the dangers are in your circuit.
I would make sure that you are on a RCD protected circuit or use an isolation transformer.
I know not everybody has one but you can help fault find a switched mode supply by using a Variable autotransformer, start at 0 volts and wind up slowly, work at a voltage below failure and find the short or whatever, by observing waveforms.

Or pull out each component and check it ex circuit.


 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf