TDS5054B Artesyn CVN300 PSU Low -12V Rail Hi 2.7V Rail U3843 Based..HELP

[denimdragon]

I have been pulling my hair out for the last week trying to figure out why my -12V Rail (Pin 14) is -9.39V, and the Pin 8 is .123V. Also on the 10 Pin Molex no# 2 pin is 11.4V when it should be 2.7 according to the service manual and the Pin 5, and 10 are 2.47V and 1.84V when they should be 7mVrms. 

I've changed all the capacitors on those traces and it hasn't helped. I figured that a bypass cap might have went belly up but that doesn't seem to be the case. All the caps I replaced checked out in spec. I never new that an ATX would be such a pain. I have considered swapping this PSU out with a solid Corsair VX550 PSU but i haven't because I'm not sure what the 10 pin molex does and it's connected to the PowerPC board.

Also, while repairing this unit, I first had to replace a mosfet and .025 SMD resistor on the PowerPC board and that resistor and mosfet are by the 24 Pin connector. I plugged the scope up when I first received it, and the 945G (Advantech) motherboard wouldn't get past the splash screen, so I changed out the caps on the motherboard and tested it with the above mention Corsair PSU and got it work properly. Now the PSU is the only thing that is holding this up.

I had 2 of these scopes with the same issue. The first one is working like a champ after replacing the caps on the 945G motherboard and recapping the PSU. Now this one is giving me the Willies.

I have a theory but it's just enough to be dangerous. I think the problem might be in the PFC circuit, but I don't know enough about it to know what exactly to look for. Not a SMPS kinda guy, but I'm learning quickly that I have to be one.

All the other voltages are in spec. One last thing, is the 10 Pin Molex for PFC? It's seems to be so because the 7mVrms pins 5, 10 are linked to the mains Live and neutral and are color matched (Brown, Blue), but they are only shrink wrapped to the lines so there's no physical connections at all. I've read some where that this is to get a sine wave for VA or something or another. I'm assuming that this is part of the whole PFC thingy.

One last note, When I was testing the PSU standalone, and put a load on it (8ohm 20 watt resistor on the +12V line, the PSU cut off. That's only a 1.5A load so I know something is screwy somewhere. I'm starting to think that it it one of the ICs (Dips).

[Jwalling]

Power supply datasheet: https://www.artesyn.com/power/assets/cvn300_02_trn_1195688538_techref.pdf

That should help as it gives you the 10 pin AUX connector pinout used for fan control and 3.3V sense.

[Dacke]

I see some similarities to ATX but it doesn't look like a typical one.   As far as the PFC goes,  measure the voltage across the large main filter cap.  It should be between 375-390V,  if it is then the PFC is most likely fine.   

A few of the rails have gone higher than normal voltage so I'd probably suspect the feedback circuit first.  I would check passives in the feedback circuit, the optos, ICs, the TL431 or whatever other reference it might be using,  etc.   Did you happen to find any schematics for this power supply?

E:  Ah, never mind,  I guess it is ATX.  So no schematics then.

[denimdragon]

I see some similarities to ATX but it doesn't look like a typical one.   As far as the PFC goes,  measure the voltage across the large main filter cap.  It should be between 375-390V,  if it is then the PFC is most likely fine.   

A few of the rails have gone higher than normal voltage so I'd probably suspect the feedback circuit first.  I would check passives in the feedback circuit, the optos, ICs, the TL431 or whatever other reference it might be using,  etc.   Did you happen to find any schematics for this power supply?

E:  Ah, never mind,  I guess it is ATX.  So no schematics then.

I thought about the feedback circuit but didn't take a look at it too much. I will look at that now and check the bulk cap for the voltage, and yeah, no schematic    I'm going to use this time to learn about ATX PSUs once I get the scope up and running (new daily driver... YAY). I have some books on their design so this whole experience has been and indication that I need to master these, and SMPS in general. Oh how I love me simple but bulky linear supplies, let me count the ways...

Thanks for your input.

Power supply datasheet: https://www.artesyn.com/power/assets/cvn300_02_trn_1195688538_techref.pdf

That should help as it gives you the 10 pin AUX connector pinout used for fan control and 3.3V sense.

Jay, thanks for the link. I'll take a look at it. Based on it being just an AUX socket, I should be able to get away with swapping the PSU out later and use it instead of this one... I think. Seems like a fun project I can do for school. I'm switching to this scope because I can use it with my TLA714 Logic analyzer on the newer 5.8 TLA remote software. My TDS784D isn't supported by the newer versions of TLA software. I'm starting to learn about embedded systems and the power of the 5054B with the 714 will be really nice. The little Rigol 1054Z is starting to show some of it's boundaries when it comes to advanced stuff. I don't want any handicaps to my learning so I figured I'd get this out of the way before school starts again. 

[nctnico]

I'd start by looking at a -12V post regulator and any diodes from the transformer.

[denimdragon]

I'd start by looking at a -12V post regulator and any diodes from the transformer.

I pulled these 2 out and they checked good. I'm doing a bit more hunting. I'm looking to see how the -12V is made in these ATX PSUs. Back to digging. Thanks for the advice also 

[nctnico]

Also be aware that these kind of power supplies need a minimum amount of load to work.

[denimdragon]

Also be aware that these kind of power supplies need a minimum amount of load to work.

I'm using a 8 ohm 20W resistor for the 12V lines. I'm really leaning toward it being the feedback loop as suggested by Dacke. Also I measured the optocouplers (3) and to have the same Vdrop and one is lower than the other two. I get 1.056V on 2 of them and .995V on the one. Might not have anything to do with this but it seems odd that two would be the exact same and one is different. Wrong thinking process ?

[Dacke]

I'm using a 8 ohm 20W resistor for the 12V lines. I'm really leaning toward it being the feedback loop as suggested by Dacke. Also I measured the optocouplers (3) and to have the same Vdrop and one is lower than the other two. I get 1.056V on 2 of them and .995V on the one. Might not have anything to do with this but it seems odd that two would be the exact same and one is different. Wrong thinking process ?

The difference in the voltage drop is probably not a big deal,  but the LED in an opto fails generally a lot less than the phototransistor,  which can fail short.   Optos usually have to be pulled from circuit to be tested properly,  but they aren't really high failure points in my experience.

The -12V rail isn't critical nowadays in standard ATX power supplies used in PCs and in fact,  much of the time it isn't even used at all.  But this isn't a standard ATX power supply application you have,  and I'm not sure what Tektronix is using it for,  aside from maybe serial ports.  The -12V may or may not be generated from a negative regulator,  depends on the power supply.

I have more questions but... Would it be possible for you to take a decent res picture of the power supply from the top? And maybe sides as well since the heatsink may block the view...

And I might have missed it,  but the voltages you measured on the 10 pin connector were with the power supply plugged into the scope and the scope turned on correct?

[denimdragon]

The difference in the voltage drop is probably not a big deal,  but the LED in an opto fails generally a lot less than the phototransistor,  which can fail short.   Optos usually have to be pulled from circuit to be tested properly,  but they aren't really high failure points in my experience.

The -12V rail isn't critical nowadays in standard ATX power supplies used in PCs and in fact,  much of the time it isn't even used at all.  But this isn't a standard ATX power supply application you have,  and I'm not sure what Tektronix is using it for,  aside from maybe serial ports.  The -12V may or may not be generated from a negative regulator,  depends on the power supply.


I have more questions but... Would it be possible for you to take a decent res picture of the power supply from the top? And maybe sides as well since the heatsink may block the view...

And I might have missed it,  but the voltages you measured on the 10 pin connector were with the power supply plugged into the scope and the scope turned on correct?

The -12V is actually attached to the B25 Pin on the PowerPC board where it interconnects to the Acq Board. I'm assuming it's for differential signaling. Also I tested the PSU in the 5054B and standalone with the same results. The PSU is actually connected to the PowerPC board (VXWorks RTOS) and the on/off switch on the front panel of the scope goes through it first before going to the mATX motherboard. The mATX then boots from the custom Tektronix BIOS then hands control back to the PowerPC board for the BIST.

I will try to take better pictures but there is an upload limit here and I'm not sure how to put the high res photos hosted on another server in the post, but I will see if I can find out. Give me a few to figure out how to get the pictures up. I also desoldered the two daughter boards from the main board on the PSU so I could take a look at them both (I already desoldered one so I could recap it).

*P.S. The scope does boot, it just hangs after show the Tektronix splash screen and just shows the little Award BIOS ribbon. Here is the link from the first scope I repaired like this. They are literally 8 serial numbers apart and had the exact same issue. A recap fixed the other scope.

https://www.eevblog.com/forum/repair/tektronix-tds5054b-little-ribbon-in-top-leftt-screen-all-leds-lit-fp-no-boot/

[denimdragon]

...I'm so tempted to try my Corsair vx550 but I would have to track down the AUX power on the Corsair and solder the 10 pin Molex to it. In the Arteysn datasheet on the PSU the AUX connector is just for fan control and monitoring. Tek had a couple of wires add to it going to the AC mains but with no physical connect, just shrink wrapped to the outs of each AC line. I haven't traced the pins relations to the PPC board but I think I'll give it a shot. I wanna do this proper, but I still have a lot of work left once I get the scope working to set it up with my TLA714 and work on the communication kinks between the 5054B, 714, and windows 7. I'm not sure how much of a rabbit hole that would be but I've had issues with the TLA714 and windows 7 initially but worked it out.

[denimdragon]

Here's the link to better pictures. I'm not the best photographer ...

https://drive.google.com/drive/folders/1ifMHlXNZepqd4Ra0n68GLHu06KbDQpaX?usp=sharing

[Dacke]

Those are good pictures.  Doing some research on that 10 pin AUX connector for fan control,  I'm not too concerned with the voltage readings there (yet).  I'd start with these issues which seem more important: "-12V Rail (Pin 14) is -9.39V, and the Pin 8 is .123V."  So it seems they are both being loaded down,  without the scope connected.  See if you can see a negative regulator (probably be labelled with 7912) ,  possibly TO-220,  there are a couple in the pictures I can't read.   If there is,  measure the voltage on the output pin.  If its still around -9V,  desolder the output pin and measure it again,  floating the pin in the center of the hole in the board and making sure it doesn't make contact.  Trying to determine if the regulator is at fault,  or it's something else.   

I've got work soon so I gotta get some sleep,  long day and the brain is fried.  I'll brainstorm some more tomorrow and check back.   Hopefully someone more familiar with these particular power supplies/scopes and issues will chime in as well,  that's a really nice scope to not be able to make use of it.

[denimdragon]

Those are good pictures.  Doing some research on that 10 pin AUX connector for fan control,  I'm not too concerned with the voltage readings there (yet).  I'd start with these issues which seem more important: "-12V Rail (Pin 14) is -9.39V, and the Pin 8 is .123V."  So it seems they are both being loaded down,  without the scope connected.  See if you can see a negative regulator (probably be labelled with 7912) ,  possibly TO-220,  there are a couple in the pictures I can't read.   If there is,  measure the voltage on the output pin.  If its still around -9V,  desolder the output pin and measure it again,  floating the pin in the center of the hole in the board and making sure it doesn't make contact.  Trying to determine if the regulator is at fault,  or it's something else.   

I've got work soon so I gotta get some sleep,  long day and the brain is fried.  I'll brainstorm some more tomorrow and check back.   Hopefully someone more familiar with these particular power supplies/scopes and issues will chime in as well,  that's a really nice scope to not be able to make use of it.

Thanks again for the help. I really appreciate it.

I did a bit more digging and this particular PSU is using a L4981A PFC IC, ST U3843B, and a LM393N DIP8 on daughter board #15 (Marked on PCB as "BD#15"). On the other daughter board "BD#16" there are two ST LM339 DIP14 Voltage comparators, TL431 shunt regulator/reference, and a 100-6 SCR TO92 package.

From what I have read so far, BD#15 is the feedback part of the circuit and I'm "assuming" this because of the 3 optocouplers I saw earlier. There are a couple of 10uf 25V Elecs on that board that I haven't changed yet (so damn tiny I didn't even see them between the heatsink and the daughter board) and I'm wondering if the one next to the U3843 is using this for the PWM function. It that cap is bad it might be the culprit. This is a hail Mary as I don't know much about how that works, but I'm going to dig into the datasheet and see what's said about the chips operations. That would be great if it is that simple. Most of the SMPS I have managed to repair are TL494 based or equivalent. This seems to be a very different set-up.

[Dacke]

No problem.   Did you check to see if there is a -12V regulator?  If so then it needs to be tested.  If there isn't one then -12V is probably being generated by the switching transformer,  and you'd need to check the components on this rail. 

PWR_OK signal on pin 8 is missing.  This is a logic level high 5V signal to tell the motherboard that all rails are functional and stable,  it should be high as long as the power supply is on.  If a voltage rail is out of spec,  this signal will go low ( < 0.4V).   If the signal goes low,  the machine cannot boot because the CPU is now forced into a reset state (possibly why the BIOS hangs just before the BIOS firmware version and CPU architecture are identified,  as these are generally the first things displayed on bootup).   This is why I'm starting with the first voltage rail that you listed (-12V),  as this is a standard voltage rail common to all ATX power supplies.  Then I would move on to the AUX 10 pin connector and the feedback circuit if need be.

[andy2000]

I successfully powered a TDS5054 from a standard ATX power supply for testing.  It worked fine, but did give a trigger error because the power line AC signal was missing.  That's what those two wires loosely coupled to the mains are for. 

[denimdragon]

I successfully powered a TDS5054 from a standard ATX power supply for testing.  It worked fine, but did give a trigger error because the power line AC signal was missing.  That's what those two wires loosely coupled to the mains are for. 

Awesome! I was hoping you would show up.

I'm going to give the Corsair a try and see how it goes. I figured that is what Jwalling was hinting at 

I was going to go to bed as it's 4:00 am here and I've been on this PSU all day literally studying it and desoldering components, more for learning about SMPS in general. I'm a bit obsessed with getting good with switchers after this. I have a copy Keith Billings and Taylor Morey's "Switchmode Power Supply Handbook" 3rd edition circa 2011, which is a year or so after this scope was produced. If anyone has any more good titles let me know. I would appreciate it.

I'm in training 

[Dacke]

If you continue to work on the power supply,  I hope to know how it turns out.  But I understand if you need to just put the Corsair or similar in there to get it working again,  especially if it's costing you time/money when the scope is not working.

One of the books that I used a lot in the beginning that I picked up a few years ago is Troubleshooting and Repairing Switch Mode Power Supplies by Jestine Yong.  Very little about math or theory,  a whole lot about efficient troubleshooting of switching supplies,  shortcuts,  case studies, substituting components etc..  English doesn't seem like the author's first language but the information and experience he has makes up for it.

[denimdragon]

If you continue to work on the power supply,  I hope to know how it turns out.  But I understand if you need to just put the Corsair or similar in there to get it working again,  especially if it's costing you time/money when the scope is not working.

One of the books that I used a lot in the beginning that I picked up a few years ago is Troubleshooting and Repairing Switch Mode Power Supplies by Jestine Yong.  Very little about math or theory,  a whole lot about efficient troubleshooting of switching supplies,  shortcuts,  case studies, substituting components etc..  English doesn't seem like the author's first language but the information and experience he has makes up for it.

I have that e-Book. It does have quite a bit of useful information in it. I'll go back and take a look at it and see what it says about low output voltage.

I'm going to use the Corsair. I've already started making the modifications to the AC coupling wires that connect to the mains for AC triggering as andy2000 mentioned. I've been working on this scope since 6/24 (overall not just the PSU) religiously because it will be my daily driver. My Rigol DS1054Z is nice but it's something about it that feels.. can't quite put my finger on it. 

I also plan on doing some modifications to this scope, but currently I have the system loaded on a Samsung 850 Pro 128Gb SSD and added 4 1GB sticks of Corsair XMS RAM from an old PC system build from 2008. I've recapped the 945G motherboard, and gave it a good cleaning. I have the VXworks manuals that I'm going to study so I can know the system in and out. I want to hack the scope as much as possible for educational purposes and also to have a nice mid-range scope. It will also be partnered permanently with my logic analyzer so it's part of a much larger project.

I am repairing the original power supply after I have the scope tested good.

[denimdragon]

Thanks everyone for your input and help. I got my scope going even though it's not the stock unit (Huge thanks to J and andy2000 for that info). I'm running the Diagnostic Loop right now but it passes SPC and everything looks good so for. I still have some work to do to it, but I'm in a good position now.

I'm going to figure out what's wrong with the original units under/over voltage issues for sure. I just need to do a bit more growing in that area. SMPS are challenging and that's fun to me so once I get the 5054B back together I'm going to start back on the Artesyn. I'm going to read Billings book over the next week or so and see if I can get a good root grasp on how these things work. I spent most of last night breadboarding  a UC3844 circuit from the fairchild datasheet for the open loop test circuit. Might be handy to make a little tester based on it. Couple of pics below...

[Jwalling]

Interesting that it didn't complain about the fan signal from the power supply, or does the Corsair have a fan tach?

[denimdragon]

Interesting that it didn't complain about the fan signal from the power supply, or does the Corsair have a fan tach?

Yeah it has a tachometer.