I recently purchased a Tektronix TDS744A that I've been trying to fully repair. I found it at a bargain price on eBay. It wouldn't power on and was otherwise really clean. That seemed very promising to me.
When I received it, it indeed would not power on. Inspection of the power supply revealed badly fried TVS diodes. (One was actually blown in half. The other had outgassed badly from the center, but was still in one piece. The circuit board underneath them was blackened.) Some of the capacitors were also suspect; they were still close to spec, but I figured I may as well replace them all while I was in there. So I replaced the all the electrolytics, the TVS diodes, and some of the regulators and rectifiers that I was able to get replacements for at a local shop cheaply. I also peeled up the trace that was over the worst blackened area and bodge wired in place of it. All of this was enough to get the power supply going strong again. To be clear, simply replacing the TVS diodes might have been enough to get it minimally working, but I want this to last.
Just a note to any of the rest of you repairing a PSU for one of these: the no-load voltages given in the 520B component service manual are useless. When these PSUs are working, they'll vary a lot on their no-load regulated voltages. Even comparing two working PSUs can give you results that are outside of the values given in that manual in opposite directions. So don't assume your PSU is bad if the no-load voltages don't match that manual. I don't know why they vary so much - I just know that they do.I noticed while I was working on the display board that the CRT and acquisition board had service tags from 2008. I believe the CRT and the attenuators for channels 2 and 3 were replaced at that time. They also managed to smash C1410 with their screwdriver while they were replacing the attenuators and didn't bother replace it. For shame! But I replaced it with a new capacitor with the exact same specs.
On power-up, the display mostly looked very good (no obvious bad spots in the LCD filters and the CRT seemed as bright as can be expected from this design), but seemed to flicker a little more than it should sometimes. So I opted to also recap the video board. That was not a fun process, since Tektronix appears to have intended for this board to never be replaced - they spot welded the chassis around it. My recap was a success and I think the picture seems to have less flicker now.
The best way I found to remove the video board was to bend the two tabs at the right of it at the front. If you do that, there's just enough room to slide it out. And remember to take all appropriate precautions when working with stuff connected to the picture tube.Initially, the scope would pass all its startup self-tests, but would fail SPC with the following errors:
ERROR: Cal meas clipped at min dig on ...
(3@ 4000): -5.119 divs
ERROR: Cal meas clipped at min dig on ...
(2@ 4000): -5.119 divs
ERROR: Cal meas clipped at min dig on ...
(1@ 4000): -5.119 divs
ERROR: Cal meas clipped at min dig on ...
(0@ 5000): -5.119 divs
ERROR: Characterized values out of border Vertical
Ch2 trigger gain
ERROR: 2260 Calibration failed
Triggers
The cursor (at the left side of the screen) for channel 2 also seems to float higher above the signal than it should. That's probably useful information to someone. Otherwise, all four channels appear to work.
I found info on testing the attenuator relays somewhere else online. According to that test, 9 of my 16 relays were already going bad. (Some were actually quite terrible.) I opted to just replace all of them (except the TN2E-12V relays, which I left alone for now). I get the impression this may pay off later, but it didn't resolve any of the SPC errors.
I tried swapping the channel 2 and 3 attenuators and I would still get the same behavior on channel 2 with them swapped. This indicated to me that the attenuators are probably good. (Or at least they're probably all broken the same way if they're broken, which seems very unlikely.)
So at this point I was still left with all the errors above and didn't have a lead on what might be wrong. I did some studying of the Zenith scans of the TDS544A schematics, which taught me a bit more about how it all works and how the subsystems are divided up. It's all very interesting, but it hasn't led me to the source of the errors.
Reading about other successful repairs of the acquisition board led me to a case where someone was just outright replacing all the TL072 and TL074 opamps because they're just so cheap (
https://www.eevblog.com/forum/repair/tektronix-tds544a-repair/). (Along with the 74HCT154s and LM311s, which I have not tried replacing yet.) I wasn't sure if this was necessary, but I decided to try it. I also tried testing the 13 10uF bulk ceramic caps and discovered that they tested a bit low. I opted to replace those with higher quality (X8R) ceramic caps. And, since I was getting into surface mount work now, I also decided to replace all the surface mount electrolytics on the communication and processor boards. After making those changes, I was left with even more errors:
ERROR: Cal meas clipped at min dig on ...
(3@ 4000): -5.119 divs
ERROR: Cal meas clipped at min dig on ...
(2@ 4000): -5.119 divs
ERROR: Cal meas clipped at min dig on ...
(1@ 4000): -5.119 divs
ERROR: Cal meas clipped at min dig on ...
(0@ 5000): -5.119 divs
ERROR: Internal adjustment range exceeded
1.1257 => Balance dac #2 (ID #145)
(8 tries)
ERROR: diagnostic test failure
trigBtlConfidenceDiag, main TI > TICLK*2.0
ERROR: Characterized values out of border Vertical
Ch2 trigger gain
ERROR: 2260 Calibration failed
Triggers
ERROR: 2260 Calibration failed
Repeat Cal Failed
And the diagnostic test failure unfortunately causes it to fail its startup self-test. I'm pretty sure replacing the opamps and those bulk ceramic caps didn't improve anything and I probably caused this additional error somehow. But I rechecked my work on replacing them and couldn't find anything wrong.
I opted at this stage to use a heat gun (and some good flux) to reflow all of the components for channel 2 and most of the components that aren't dedicated to a specific channel. After, I gave it a very thorough cleaning with 91% ISA. After the cleaning, I was left with a much cleaner acquisition board and the following set of errors:
Error: diagnostic test failure
slewrateTrigDiag, **MTOUT exp data = 0x1 actual 0x0,
ERROR: Cal meas clipped at min dig on ...
(3@ 4000): -5.119 divs
ERROR: Cal meas clipped at min dig on ...
(2@ 4000): -5.119 divs
ERROR: Cal meas clipped at min dig on ...
(1@ 4000): -5.119 divs
ERROR: Cal meas clipped at min dig on ...
(0@ 5000): -5.119 divs
ERROR: Characterized values out of border Vertical
Ch2 trigger gain
ERROR: 2260 Calibration failed
Triggers
So, this gave me a new diagnostic test failure and eliminated the old one. Other than the diagnostic test failure, I'm back to the exact same set of SPC errors the scope had when I first got it to turn on. I'm just about to the point where I'd be willing to settle for just that initial set of errors for a while if I could eliminate the diagnostic test failure. But I would still really like to see it pass its startup self-tests again and pass SPC.
I've since tried swapping the old ceramic bulk caps back in to see if they were the cause of the remaining diagnostic test failure. But it doesn't seem to matter if I have the new caps or the old caps installed - the new error still happens. I may try reinstalling them once I have the board otherwise working to see if there's any change. Theoretically, they're better caps that should last longer so I would rather use them if they work fine.
At this point, I suspect that the "Cal meas clipped at min dig" errors are all the result of the same root cause. It's possible that cause is also responsible for the "out of border Vertical" error. At least, it wouldn't surprise me. Since the "slewrateTrigDiag" error appeared after I worked on the board, I suspect it has its own unrelated cause and it's probably something I did. But I think it's possible that's being caused by something like a damaged trace, via, or solder joint.
Can anyone who has experience with these please help me narrow down my search for the root causes of the remaining errors? Thanks.
Home
Help
Search
Login
Register
