Rohde & Schwarz CMU200 Bad Video & More

[pcwrangler]

I recently purchased a used “for parts” Rohde & Schwarz CMU200 that showed the video working but had a missing front foot and had an error on boot (pic 1). When I received it the video was rolling like it didn’t have sync(pic 2). Luckily the external VGA port worked great and I saw the CMOS battery was bad. When I reset the BIOS to CMU defaults and had the same result I started looking at other settings in the BIOS but nothing was helping so I moved on to the software error.

Turning it on a few times, it never got past the error to fully boot. Good, I hate intermittent faults. I used Alt-F4 for DOS to run scandisk and defrag(full drive, not just fragmented files). Defrag showed 2 blocks that were previously marked as bad but overall there were no new issues. I then backed up the drive with ddrescue and it showed the same results as defrag; 2 bad blocks on the drive. The speed was fine unless it was reading just those 2 bad blocks and it never made any weird noises. However, I despise HDDs so I tried to replace it with a CF card to IDE adapter with the HDD image(pic 3). The BIOS recognized the device but it POSTed very slowly and ultimately couldn’t find the O.S. Back into the CMU it went until I could order a PATA SSD.

Since the error happened toward the end of the loading process I was able to choose the option at the bottom left of the screen to load defaults. Surprisingly, that worked! I rebooted again out of disbelief but it was working fine. I suppose a bad block may have corrupted some setting that factory defaults resolved. Preliminary tests using SA, Power, etc. all function as expected. I will run full self-tests and alignment when I have it fully assembled again.

To do:
1. Replace HDD with SSD
2. Fix video sync issue. Probably have to take apart the front panel 
3. Replace CMOS battery.
4. Check front panel fans and cpu heatsink/thermal paste (the CPU showed a high temp in BIOS)
5. I searched high and low but noticed there are no options for people with a completely empty/new drive to install R&S software. The answer is always: “Clone the old HDD”. I would like to find a way to copy Version Manager to an empty DOS 6.22 drive to facilitate the rest of the R&S software install.
6. Find a replacement front foot with extender.
7. Make the wife jealous by the way I stare at the new addition on my bench. 

** Edited: I had a picture of the wrong CF card. Now pic 3 is accurate. SanDisk Extreme 32GB UDMA7

[pcwrangler]

I took off the front panel and the contacts looked ok, not bent(pic 4). Some weren’t perfect but were definitely within tolerance. I straightened the pins that looked slightly off and reattached the panel. Unfortunately, I’m just not that lucky. Time to take it apart. This thing was more of a pain than I expected(pic 5). All I can say is keep track of screw locations because each layer has their own. I finally got the panel open slowly and carefully because I’ve read others have damaged their ribbon cables(pic 6 and 7). Here is a pic of the control/CPU board(pic 8 ). The seller packed it well but the box was very beat-up. Thanks Fex-Ed. Still, I didn’t see any loose connections, etc. 
Now that I am looking at the original post by the seller(pic 1), the ghosting/bleeding is visible. I never noticed that before. I’m still not sure how they had it stable and not scrolling. If anyone has any ideas on this display please let me know.

The battery is a CR2032 but I recall reading the manual indicating about a soldered 3.4v. Is this a later version or did someone already mod this? Either way, I am thankful for the coin holder. The old battery measured <100mV so I replaced it.

The panel fan was next. I noticed it doesn’t spin unless my finger was on it. Testing it with my bench PSU it refused to spin and took >3A. I checked for a short on the board’s fan header but it looked good so I removed the fan. Flushing it with contact cleaner revealed very dirty liquid draining out and it started to spin better. It worked on my PSU but it still made funny noises in most positions while the current wasn’t steady. I don’t trust this thing at all so I will be looking for a replacement. The fan is an ebm-papst 405F(pic 9) 40x40x10mm.

[richnormand]

I have a cloned copy of my HDD. Could send you a copy if needed providing the version is compatible (hddguru.com).

Random Compact Flash card might not boot if it is not a UDMA  (6 or 7, I forget). It might be recognised  but might not boot depending on your adapter. I had issues with that on my CMU200.
There is a post on eevblog to that effect that I could not find now.  Not all cards/ adapter combo can boot:  https://www.transcend-info.com/Support/FAQ-391

If in there replace cmos battery.

Small fan tend to seize and result in over temps. Disassemble and oil.  Mine was stopped with the temp indicator saying it was over specs. Do the main fan while at it.

Saw a sync issue on the display and it turned out to be a flat flex connector issue.

Hope that helps.

Cheers.

[pcwrangler]

Thank you for the insight Rich. Did you have to replace the ribbon? If so, where did you source it?

[pcwrangler]

I found a puncture (red) in the video ribbon cable (pic 10) but it was on one of 6 shared lines (green) so I think that wouldn’t be a big issue. I also confirmed continuity from one end of the cable to the other on all pins. One end is soldered to the small pcb or I would swap it out as a quick test. However I think it is ok.

The contacts on the board to board connection look straight and clean (pic 11) but I used a small amount of contact cleaner and inserted repeatedly for good measure. I then checked many of the components around the “PanelLink” QFP ic (pic 12) followed by every solder joint on the QFP by nudging them. So far this all looks ok. Moving on.

I attached a pic (pic 13) of the panel model number for reference. Sharp LQ084V1DG21

[richnormand]

Thank you for the insight Rich. Did you have to replace the ribbon? If so, where did you source it?

No. Just clean the pins. Squeeze the ribbon clamp on the wires and reseat. Must have been a bad contact at the pins, oxidation, bad contact in the cable to connector itself or a badly seated connector.
This was after finding a bad fuse for the LCD display fluo inverter. Been fine for over five years now.

[pcwrangler]

Thank you for the tips. I will do all that when I re-assemble the front panel.

I removed the CPU board and checked capacitors and soldered legs of video chip and connector. Checked everything in the area that I thought was related to video. The CPU thermal paste was dry and crusty as expected in a 20 year old set so I replaced it. The fan and SSD is on order.

Does anyone know what these unused RJ45 jacks are on the CPU board? (pic 14) Is this ethernet? I’m just curious as I have no use for it on the network.

[pcwrangler]

I did my best to squeeze the video ribbon connector onto the cable since I don't have a proper crimper for that type. I also used contact cleaner on the same connector. The screen really sharpened up but still had the roll/sync issue. The good news is with the new fan installed and fresh cpu thermal paste the unit is running much cooler. 

One step at a time but still progress. I'll have to dig deeper into the sync signal. More to come.

[pcwrangler]

SUCCESS! After datasheet hunting I confirmed the vertical sync signal is coming from the SiI141BCT80 on the board mounted directly to the display panel(pic 11, 12 above).  I couldn’t find anything that would stop the signal on the board and I couldn’t probe while it was powered on since the board mounts face down. Since most of the video issues I have read about involve the cabling, I inspected the white connector more closely(pic 11 above). When I looked before, I used a magnifying glass and didn’t notice anything but cleaned both sides with a brush anyway. This time I used my microscope and knew instantly what the problem was; filthy contacts on both ends. I scraped every pin with a needle-point probe and some of them even had their plating flake off. After a thorough brushing with IPA they were solid connections again.

Quick Tip: If using liquid like IPA on the panel side, be very conservative with it. You don’t want it flowing into the panel and leaving liquid marks.

It worked perfectly after that. Not one jitter or roll and the screen looks brand new(pic 15, 16, 17). My phone’s camera doesn’t do it justice but it’s incredible. I recommend anyone having any video problems to inspect this connector under a microscope. Remember mine was also color bleeding/blurry.

On to the self-tests: The “Internal RF Loop Path 1” fails on every line with values from -40dBmISH to -80dBmISH(pic 19). I could use some advice on this one as I just don’t have the experience. I don’t find much on the forum or elsewhere online either. Any help will be greatly appreciated. 

BTW: I realized the “load defaults” option I used in post #1 just bypasses the self-test after it finishes loading modules and booting the system. It didn’t fix the error, it just bypassed it. And since the unit remembers what screen you were last using and boots to it, it no longer shows the error unless you run the self test again. My software error is still present (pic 18) but I will address that when my SSD arrives.