Rohde & Schwarz Universal Radio Communication Tester CMU200: PSU problem

[richnormand]

Rohde & Schwarz Universal Radio Communication Tester
Model CMU200

Issue: Front power button Standby (orange) goes to Run (green) but nothing else happens. The unit is totally dead otherwise.

Followed the “service” manual flow chart but this does not fit with the diagnostics.  The orange light indicates the PSU supplies the 12V standby voltage. This checks out with the LED at the back of the motherboard. The green light is supposed to indicate that the 5V line is OK. This is not the case since all the LEDs for the 3.3 v, 5.2v, 6v, 8v, 12v, 12v Fan, etc…. are all OFF. The so called “service” manual is not very useful for this and does not treat that particular case. Either way, it says to replace PSU or replace front module.

Suspect the PSU third stage or the front module controller. Removed all modules separately to see if there was a shorted/faulty plugin unit that could inhibit the “power good” signal and shut down the PSU, but no luck.

Removed the PSU and operated it outside the unit.  Getting a good 280V from the line primary and the 12V standby voltage is present. Third stage with the inverter transformer for all the other voltages does not seem to be active. It goes to board 2 that has a slew of control circuitry.

Checked the electrolytics  MOSFETs and IGBTs etc.. as a first pass,  before digging in further.

Other possibility could be the front module controller that never sends the signal to “turn on” to the PSU or some “power good” handshake, much like in computers.

Any history of PSU failure for the CMU200? Only found one mention on this forum, no details. Nothing else on the web either.

Anyone with the schematics or the plug pinouts to the motherboard? That would save me a lot of tracing signals and reverse engineering.  As you can see from the photo this is not a simple straightforward inverter supply! I do have the “operating” and “service” manuals for it but did not find more detailed information.



Help appreciated!

Cheers and thanks.


UPDATE:
Reassembled the PSU. Removed all plugins and anything attached to the motherboard except the front control panel. No luck. Same issue.

Took apart the front control module (a custom PC basically). Checked all the smd fuses and all is OK. Changed the 3V lithium memory backup battery while in there.
Only suspicious item is what seems to be a temperature indicator as shown in the photo attached.

[ZL1CVD]

I haven't had a faulty CMU200 supply but have repaired many HP 8924 & E8285A supplies. I had put aside an E8285A supply for a couple of years because it got the better of me until a ham radio contact told me to read this http://www.repairfaq.org/sam/smpsfaq.htm - It didn't specifically help fix the supply but put my mind into SMPSU repair mode which enabled me to nut it out. I also have read somewhere that CMU200 supplies also fail from dry electrolytic's & opern circuit inductor. 
73s om de chris zl1cvd

[richnormand]

I haven't had a faulty CMU200 supply but have repaired many HP 8924 & E8285A supplies. I had put aside an E8285A supply for a couple of years because it got the better of me until a ham radio contact told me to read this http://www.repairfaq.org/sam/smpsfaq.htm - It didn't specifically help fix the supply but put my mind into SMPSU repair mode which enabled me to nut it out. I also have read somewhere that CMU200 supplies also fail from dry electrolytic's & opern circuit inductor. 
73s om de chris zl1cvd

Thanks ZL1CVD,

This is a nice reference/link, in particular the several cases presented and the supplementary  material near the end. I printed most of it and will go to the various references given in there.

I tested pretty much the board three for shorts and the ESR of the various big caps. All is OK. I still have to remove some of the power devices off circuit to test them properly without the low impedance of the second  inverter transformer winding across. Nothing obvious either on power resistors, no sign of over heating. Also board twu is a maze of opamps, opto-isolators and the control chips for the dozen of voltages supplied by this PSU.

Your comment about open inductors is an excellent one. I'll do that tonight.

Next thing to do, since I can power the unit on the bench and get the 12V "standby" voltage, will be to locate the "power on" pin as well as the "power good" pin on the bloody connector. Its a future buss type connector with a zillion pins. A proper pinout for the plug would be so useful here to locate the temps limits and control lines!

This might be the only way to see if the fault might be located on the front module not sending the "power on" command to start the PSU secondary.  That scenario would be similar to an ATX computer PSU when you ground the green wire to get it to start. I repaired some ATX PSUs in the past with shorted power MOSFETS and bad caps, but the complexity of this particular  R&S supply is pretty awesome. 

Thanks again for the info, will chime back when I go though the information you linked me to.

[richnormand]

Looked at inductors and such, no luck. It's really starting to look like the PSU is not getting the command to start compared to "start and collapse" due to a short (no click noise) .
Looks like I have to identify the "power good" line and the "main start" line at the connectors and scope any activity on these.

Link from ZL1CVD is really worth the read. 73s

[PartialDischarge]

Hello Rich

today I opened my CRTU which I guess has the same PS (PULS SN250) and 'probably' the same connectors at the control side. I believe at least the former is true, so here's what I did:

1) check voltages at PSU connector with unit turned OFF (but powered)
2) check voltages at PSU connector with unit turned ON
3) identify possible "power on" lines at PSU connector
3) turn off and disconnect unit, find those 'candidate' lines at control side (X12 connector in my unit)

See attached pictures and PDF. Pins pictured in green, output some voltage even in the OFF state.

So, I'd say B1, C1 and D1 are pretty good candidates, especially B1 given that the ON logic could be negative.

If i were you I'd measure the voltage at B1 with the PSU disconnected from the equipment, and if you get some voltage, drive it to zero with a
10K to 100Kohm resistor

[richnormand]

Hello Rich

today I opened my CRTU which I guess has the same PS (PULS SN250) and 'probably' the same connectors at the control side. I believe at least the former is true, so here's what I did:

1) check voltages at PSU connector with unit turned OFF (but powered)
2) check voltages at PSU connector with unit turned ON
3) identify possible "power on" lines at PSU connector
3) turn off and disconnect unit, find those 'candidate' lines at control side (X12 connector in my unit)

See attached pictures and PDF. Pins pictured in green, output some voltage even in the OFF state.

So, I'd say B1, C1 and D1 are pretty good candidates, especially B1 given that the ON logic could be negative.

If i were you I'd measure the voltage at B1 with the PSU disconnected from the equipment, and if you get some voltage, drive it to zero with a
10K to 100Kohm resistor

Hi MasterTech,

thanks for the response and the work you put into this.
It would be nice to have the voltages on the pins while it is working and while it is in standby.
Looks like the same connector alright, may even be the same motherboard too.

Here is a sketch of the connector I did last week. At the lower left and at the far lower right are the pins I had marked "to check" up to the front CPU module. Much like you suggested.
As you can see B1 C1 and D1 are on the list. Also are A2 A3 D3 A24 B24 and D24 that need to be traced on the MB.

Also is a pix of the PSU label with the type number. Looks like it is the same PSU. Not a surprise considering the lineage of the two units.
There is also another separate connector that goes to the first plug-in. It looks like and old pc hard drive connector.

I have the PSU powered on the bench with the standby 12V line active.I'll try a 10k resistor to drive the B1 line to zero as you suggested and try other combinations.

Cheers and thanks. This is an on-going saga......

UPDATE:Just measured the voltages. PSU on the bench (not connected to the motherboard but under power):

(pin#, voltage)

a1     10.7V
b1     11.1
c1     0
d1     0
a2     0.1
a3     0
d3     0.5
a24   1.0
b24   0
d24   12.4


UPDATE2:
Drove a1 and b1 to near ground with a current limit resistor (about 1k). No effect. They are high internal resistance lines as 10k brought them to about 5V.
Also d24 is a solid 12V. Probably the 12V standby.
I used two voltmeters to look at the two major set of pins that look like high current output, most likely the system 5V and such. I would have noticed if the PSU came to life.

So does not look like a1 or b1 are negative logic to start it. Perhaps some other pin are positive logic and require the 12V to start the PSU?

Would be nice to take a full readings on a good unit while in "standby" and then "on" and see what are the difference on these lines indeed.

[PartialDischarge]

I think you missed the PDF I posted earlier. there are the voltages at standbby and power on.

from what i see your unit looks good in standby state. Can you look into the logical circuitry inside the PC that drives the signals I refer to in the X12 coonector? If at least we knew if those signals are inputs or outputs to,the psu

[richnormand]

I think you missed the PDF I posted earlier. there are the voltages at standbby and power on.

from what i see your unit looks good in standby state. Can you look into the logical circuitry inside the PC that drives the signals I refer to in the X12 coonector? If at least we knew if those signals are inputs or outputs to,the psu

WOW! You are right. I totally missed the PDF in your original post.       Sorry about that MasterTech, my fault 
Nice job too.

I'll plug the PSU back on the motherboard and move to looking at the PC X12 connector side.

Thanks again. Hope we meet some day, I owe you a beer!

[PartialDischarge]

Good news, just decided to test the B1 pin by myself. While the CRTU was in standby I drove that line low with an 820ohm resistor (just to go a bit farther than you...) using the grounded screw next to it and voila it turns on 
I was confident about B1 because the 0.2V present when the unit is on, resembles the saturation voltage of a transistor driving that line low.

So next thing you'll have to test the same, and you'll know if its the PSU or the control unit


Cheers

[msraya]

Very Interesting Information.

I my CMU power supply dies suddenly.. I will need that information...
Thanks All!!
Manuel

[richnormand]

Good news, just decided to test the B1 pin by myself. While the CRTU was in standby I drove that line low with an 820ohm resistor (just to go a bit farther than you...) using the grounded screw next to it and voila it turns on 
I was confident about B1 because the 0.2V present when the unit is on, resembles the saturation voltage of a transistor driving that line low.

So next thing you'll have to test the same, and you'll know if its the PSU or the control unit


Cheers

Thanks for the feedback MasterTech, much appreciated.

Just did that with a 820 ohms and even went to a near dead short (10 ohms) with no luck. The PSU does not start.
Only other factor I can think of is the fact that my PSU is on the bench. Once plugged in the motherboard/PC it would have a load and perhaps another control signal ("service" manual mentions things like over temperature and power good signals for example)?

Was yours on the bench unloaded or in circuit when you did the test??

At any rate it looks like you are right and that B1 is the start signal.   I'll trace it in the PSU before reassembling the whole thing and bring B1 low while in the CMU200.

[PartialDischarge]

Was yours on the bench unloaded or in circuit when you did the test??

It was inside the crtu, it may have another signals that need to be connected before power up.... so all is not lost for you 

[richnormand]

Was yours on the bench unloaded or in circuit when you did the test??

It was inside the crtu, it may have another signals that need to be connected before power up.... so all is not lost for you 

OK I reinstalled the PSU in the unit. When I go from "standby" to "run" the B1 line goes from  11.0 volts down to 0.2 volts as it should. But "no cigar"...
So either there is something else preventing the startup or the problem is inside the PSU....

[PartialDischarge]

Theres one more test you should do. put a scope probe at b21 and see whether the line is driven low by pressing the power up button or not. I suspect it will, and you have a faulty PSU.

I recall from another eevblog thread about the cmu or crtu ( cant remember) that a guy fixed his psu by changing a switched regulator. It was something like TOP2000Y or TOPY2000, at the time I looked at ebay and it was available for sale. Besides trying to 'debug' that psu Id suggest you change that chip too....

Keep it up, that unit deserves to be repaired! 



EDIT: it was a TOP200YN...let me know if you identify that part in your unit

[richnormand]

Theres one more test you should do. put a scope probe at b21 and see whether the line is driven low by pressing the power up button or not. I suspect it will, and you have a faulty PSU.

I recall from another eevblog thread about the cmu or crtu ( cant remember) that a guy fixed his psu by changing a switched regulator. It was something like TOP2000Y or TOPY2000, at the time I looked at ebay and it was available for sale. Besides trying to 'debug' that psu Id suggest you change that chip too....

Keep it up, that unit deserves to be repaired! 








EDIT: it was a TOP200YN...let me know if you identify that part in your unit

Thanks again.


Status: PSU is plugged to the motherboard with all units. The front button is initially amber (standby) and then green (on) when pressed

The b21 is of course the whole line abcd21.



voltages are first in standby mode and then in on mode.
A1 10.8 then 0.77 NOTE: did not change with PSU on bench and in your measurements it was 10.6V off and 11.3 on... TO DO CHECK!
B1 11.8 then 0.23, seems OK
B21 0.33 then 0.33 Note: see scope trace later
D3 0.4 then 0.4 Note: same as on bench
D1 0 then 0
C1 0 then 0

Scope traces. Low speed  Tek 7D20 with direct probe.
B21 line: initial is 0. Then plug probe in standby, then push on button. NOTE: initial rise in voltage and then a shutdown. Remove probe. Voltage consistent with previous measurements. The other outputs react the same...
A1 line : initial 0, plug probe in standby 10.8V, then push on button drops to 0.77V. Remove probe.Voltages not consistent with previous measurements on bench.  Note vertical v/div, same horz.

At my first tear-down of the PSU I noticed the TOP200YN and got the spec-sheet. The one I noticed was on the primary at the 120 and 220V first inverter. The initial 280VDC stage was ok, as well as the 12V standby. On the other hand I did not check if there are others for the third supply. I will check for that.

[PartialDischarge]

At this point all you have to do is find the faulty item within the PSU  , otherwise the signals look ok. If I were you I'd find a substitute TOP200YN...Meanwhile keep looking and see whether you can test and identify any other broken parts

[PartialDischarge]

Its actually fairly easy to test that top200, see attached pictures from the PowerIntegrations specs...

[richnormand]

At this point all you have to do is find the faulty item within the PSU  , otherwise the signals look ok. If I were you I'd find a substitute TOP200YN...Meanwhile keep looking and see whether you can test and identify any other broken parts

I'll do the test in the spec sheet.  The first one on the primary side is most likely OK as I have the full voltage on the cap and the 12V standby and it takes care of the switch between 110 and 220AC as a preregulator, but I'll look at all the other boards too.
Looked at Mouser, Allied, Digikey and Gervais (here in Canada) and a few other places for the TOP200YN. All non stock and obsolete. Life cycle seems to have been short for this product. I'll look to see if there is a replacement for it that is compatible. Several TOP200YN are on Ebay and Alibaba from china....most likely counterfeit or recycled, but if they work, for a while OK.

In the meantime I'll finish doing all the measurements I can think of while the PSU in still plugged in. The most different reading from your "standby" and "on" list (apart from the obvious lack of main supply) is the A1 line. Looks like perhaps the "power good" signal sent back from the PC front end?

Cheers

[PartialDischarge]

Yes, c1 d1 are probably power good outputs, and maybe a1 too
The direct replacement, with only a tiny difference from the original is the 3 pin top222yn available from Mouser:

[richnormand]

Yes, c1 d1 are probably power good outputs, and maybe a1 too
The direct replacement, with only a tiny difference from the original is the 3 pin top222yn available from Mouser:

Ordered from Mouser this afternoon in case I need them.

[PartialDischarge]

Very interesting discussion

A trick I also used with a CMU to power it on is to keep the power on button pressed during switching the power on on the back of the CMU. After doing this sevaral times I got the CMU to work again in some instances. Mine also started with green light but stopped during booting to shut down.

I'm also wondering which part of the controller is powered on during standby mode to detect an activated power button. Is there a watchdog sort of reset circuit on the controller which drives the activation pin of the PSU, or is the button directly wirded to the activate PSU pin without any intelligence?

I hope your PSU is still alive.

Hi, the button does not go directly to the psu, there is some logic on the pc board side. I know this because when you power it down, a message appears on then screen and then it shuts down. So the button is sending a message to the pc, and I guess the logic turns off the psu when the pc replies "ok, its safe to turn off now"...

[PartialDischarge]

Just found out :

B1 (activation) on X110 is connected to B24 on X12 (the main connector to the front unit)

next step is to see where that ends up on the board

lol, I think you also missed the PDF i posted earlier, I found that connection too!

[richnormand]

Indeed it's also mentioned on your sheet, didn't noticed it before, I thought I had a break through

I did some more testing, but using another PSU, without front panel attached

Good PSU :

With power switch on the back on the standby led is constantly green
starts up after B1 to GND

Bad PSU

With power switch on the back on, the standby led is going down after few seconds,
Does nothing with B1 to GND

With (probably bad) front panel attached:

Good power supply with B1 to GND doesn't start up,

The latest thing could be of interest for topic starter:

Does it make any difference if you attach the front panel with respect to the activation of your PSU?

Unfortunatly I don't have a poven good spare front unit... (though some good CMU's but dont want to break the seal.

Indeed MasterTech did a good job
The 12V standby is working well and all the primary side of the PSU checks out. The front will go from amber to green. The B1 line goes down as expected.
In my case there is no difference in standby voltages out of the PSU without the front PC or any of the modules.
Enabling the PSU from the front or on the bench does not show any difference. Dead in both cases. But I can see a situation where the front PC could prevent the PSU from starting. Real test would be to force a good PSU to start while on the bench and see if it goes. There is probably more than one signal required, such as over temp, pc logic shutdown etc...

I am waiting for a TOP222YN to replace the TOP200YN. Off-bench test of the TOP200YN seem OK however. By the way out of three boards (not counting the smallish one attached on board one) I only found one TOP device but there are two other switching supplies here I think. I have not totally figured out the control board.

Finished mapping two of the boards, tested all caps and inductors (small resistor types for open in particular) All OK for now.
As I go to each board I remove the clips on the heatsink and note the device numbers for future reference.

Progress will be slow over the next week as other stuff is piling up.

EDIT: just got the TOP222YN and put it in.

[PartialDischarge]

Nice job Rich, eventually youll get the faulty part. Dont forget to post your notes on the psu (ie part numbers, schematics...) as it may help others in the future!

[richnormand]

Nice job Rich, eventually youll get the faulty part. Dont forget to post your notes on the psu (ie part numbers, schematics...) as it may help others in the future!

Will do. Mostly part locations and part numbers. Don't get your hope too high for complete and detailed schematics however. I'll trace what I need for a repair and understanding the problem.