Rohde and Schwarz CMU200 freeze when RF generator starts.

[electronic_guy]

Hi,

I've got a CMU200 that becomes non-responsive when the RF generator option is turned ON. I've dig into the problem and so far I found the following. I have replaced the original HDD with a SSD by copying it with HDDGuru Raw copy tool and that seems to work and CMU boots normally.

   1. The CMU200 takes a long time to load the options and stays at BaseDiscoveryOptionsDevice. This is true for both hard drives and it makes no difference in drive being HDD or SSD.
      a. I have noticed that on another blog thread this has been noticed. The link is below.
      https://www.eevblog.com/forum/testgear/rohde-schwarz-cmu200/150/
      
      The mentioned CMU200 in the above thread had a broken DIG module (TXDSP MODULE in R&S jargon) and replacing that corrected the RF generator freeze.
      
   2. System self test on RF front end shows some voltages out of range. Check picture-1
   3. System self testing RXTX  board shows the same voltages referred above are out of range Check picture-2
   4. System self test shows that the CMU-B68 board is not usable. I think this is the main culprit here. Check picture-3
                I've already taken out the CMU-68 and CMU-B21 combined board and reseated it, in case to see if that might be the problem. But no change. The board is still not usable according to the self test.

   5. The CMU200 becomes non-responsive when the system self test reaches the CMU-B56Var14 test. Check picture-4
   6. FE self test shows over voltages. Check picture-5
   7. REF self test is okay. Check picture-6
   8. DIG self test is okay. Check picture-7
   9. RXTX self test shows over voltages. Check picture-8
   10. Hardware data verification test does not show passed for the CMU-B68. Check picture-9
   11. Options at the start up shows the CMU-B68 is available. Check picture-10
   12. Initializing the WCDMA FDD option shows a software error and it says one of the C++ files are missing. Check picture-11
   13. L1CoPro Level calibration makes the CMU to freeze and become non-responsive.

Please support to solve this problem.

Thank you in advance.

[electronic_guy]

I'm attaching the remaining photos with this reply.

[R-1125F]

I think you are on the right track, but the B68 error is strange.   Try doing what the other thread that you reference tried; remove the DIG board and restart the CMU200.  See if it recognizes the B68 module.   A bad digital board is common. 

[R-1125F]

You may also want to join https://groups.io/g/Rohde-and-Schwarz

[richnormand]

Your first photo showing 26V are worrisome as this grossly out of range (9V)
voltage could be harmful to a lot of boards and plugins.
Also some board/plugins make their own local voltage. Could be a bad local regulator.

The cmu psu can work with all the loads disconnected and I would first measure all the voltages there supplied to the motherboard before doing anything downstream.
There is a long posting on this site with all the info for the main motherboard psu plug:
https://www.eevblog.com/forum/repair/rohde-schwartz-universal-radio-communication-tester-cmu200-psu-issues/25/
See post #41
Also, if I remember correctly, at the corner of the motherboard the are test points near the leds.

If the psu voltages are OK I would remove all the non essential plugins back to the most bare configuration and not add anything until it is resolved.

Good luck with it. You have to love it when companies do not supply schematics.

[electronic_guy]

Sorry for not getting back sooner,

I did the tests you have asked and got the following results,

When the B68 and B21 combo is removed, the 5V reference went to some acceptable range ( picture-12)

But the RxTx board voltages are still not within the value and some more voltages went out of range after removing the B68 and B21 (picture-13) 

I removed the board and inspected it. Some modules have gold finger contacts rusted. I don't know if that could be an effect. I will clean up them and try again.

But I removed the small DIG board and then put the B68 and B21 back and still the board is not usable (picture-14)

I measured the voltage references on the board. Specially at the input from the supply and found all the voltages are nearly the value expected. Below are the measured voltages in the order of those voltage points on the PCB. Where the power supply connects there are 9 LEDs in two rows, so I have listed the voltages in that order. I have no idea where those voltages power up the main board.

When I included the B68 and B21 back (without the small DIG board) the 5V reference on the FE test went back out of range. (picture-15)

                                                                                * +28 V (measured= +28.4)
   • + 3.3 V (measured = +3.385)                          * +5.2 V (measured=+5.3)
   • +6 V (measured = + 6.13)                               * +8 V (measured = +8.51)
   • +12 V (measured= +12.54)                             * +12 V (measured = +12.38)
   • +12 V (measured= +12.36)                             * - 12V (measured = -12.49)

On B21 there is a voltage measuring point and there also the +1.8 V(measured = +1.82) and +3.3V (measured +3.367) are seem to be within an acceptable range)

Now I'm clueless if I want to replace the B68 and B21 only or if I need to replace the RxTx board too.

Your opinion on this is highly appreciated.

[electronic_guy]

More photos

[electronic_guy]

More photos

[electronic_guy]

More photos

[R-1125F]

It is hard to say what is wrong without seeing it.  Clearly, something is wrong with tuning the RX and TX LO.  Is the 110.8 MHz signal from the reference module OK?

[electronic_guy]

Which cable output gives this signal ? I will check if the signal is okay.

[R-1125F]

The signal comes from the reference module, connector X106,X107, and X108.  It connects to the RXTX module.  There are selftests that should show if the reference clocks are present.  If you don't have the service manual, it is available here: https://gotroot.ca/cmu200/CMU_Service.pdf

See page 3.76 for module selftests. 

I think what is odd, is that the RX and TX LO voltages are way off.  It is like it is not locking.  There could be many reasons for this.

[richnormand]

with respect to the different voltage reported (26V) on the RxTx board note that it has it's own regulators on board.

https://www.eevblog.com/forum/repair/rohde-and-schwartz-cmu200-rxtx-module-issues/

On the first post the fourth photo shows the pass transistor I removed (right hand side) with the other one still present on its left.
Mine was open but if your is shorted it could explain the higher voltage that is reported and the loading when the RxTx is plugged?

[electronic_guy]

Thanks for sharing about your troubleshooting on the RXTX board. I will check on the pass transistor on the RXTX board.

Since the CMU B68 is shown as "board is not usable", would it be better to try replacing it? What I wondered is that if it's the fault on RXTX board that's causing the B68 board to be not usable. May be B68 is good. Could you comment on this please. So I can order a B68 board and first try to pass the self test on the B68 first and then go to the RXTX board ?

[richnormand]

if I remember correctly the CMU-B68 = Versatile baseband board for WCDMA. The CMU should boot properly without it in a base configuration (I think).
I would remove all the unnecessary stuff and get the unit working in the minimum base configuration first.

Also note that some boards need the appropriate software installed too.

[electronic_guy]

So which board is responsible for RF generator on CMU200 ? Is it the RXTX board that creates RF signals ? In a minimum configuration the B68 and B21 is not needed right for cmu200 to work as Spectrum analyzer, power meter and RF generator ?

What is the purpose of the digital board and the reference board on the CMU200 ? I just have no clear idea on the purpose of each board. Service manual is quite vague about the function it seems.

[richnormand]

Indeed.

Here is a list of options for the cmu200
https://cdn.rohde-schwarz.com/pws/dl_downloads/dl_common_library/dl_brochures_and_datasheets/pdf_1/CMU200_Options_and_accessories.pdf
At the end please note the mandatory modules (tick in square), tick and optional. The mandatory modules are for a basic operational instrument.

Poor wording of my part.
Yes you need these modules to use the cmu200 as intended but it should boot without them (and give you warnings).

I am worried about your RxTx showing 26V instead of 9V. The lo signal must be absent or the power supply voltage to the card is wrong. I would check the two onboard voltage regulators and I would look for signals at the two inputs of the RxTx card that are supposed to come from the digital board.
With the service manual  block diagram you can see which card depends on a signal supplied from another and tap in the connector.  All depend on making sure the main PSU is OK (seem to be the case from your previous post) but many cards have on board regulators and these also have to supply the card with proper voltages.


The idea is to populate one module at a time and get all the supply voltages OK before getting in to the higher functions.
Without proper voltages some cards might appear to be unusable until the proper voltages or signals are established.

At the end of the day I managed to fix multiple problems on my CMU200 but I needed a well equipped bench with spectrum analyser, oscilloscope, DMM, microwave extensions, made e field and h field probes for the SA, and a lot of help from this forum..... also took about a year to get there.