Tektronix 453A, no trace and hard to find short

[theor]

Hi everyone. My Tektronix 453A started showing a broad, flashing trace, which turned broader and dimmer until it vanished entirely. This lasted about a minute or so.
The fan and various lights still turn on, but I have no calibration signal, some lights are off, and no trace can be seen with the beam finder.

I checked the voltages on the low-voltage regulator board and am missing the -12V rail. I also notice a voltage drop across R1129, which points to a short in the -12V circuit.

The short goes away when I disconnect the H pin on the low-voltage regulator board, then I no longer have continuity across C1121. So at least I know the first stages (transformer and regulator) are fine.

I then traced the -12V rail through the power distribution schematics according to the manual:
https://www.manualslib.com/manual/1899308/Tektronix-453a.html?page=274#manual

The -12V wires are easily identified with their tanned background instead of white. Each of the capacitors along the rail, on the six boards, had continuity across them, which is expected. But they also proved fine once I disconnected the -12V pin on each board.

I'm now left with all the -12V pins disconnected on the boards (pins AD, T, A, H, AN, CC), none of the boards shows a sign of a short, yet I still have a short between the -12V wires and ground, except for the wire that goes to pin T on the A1 board (that wire is the only I can leave connected to the board without having continuity across its -12V caps).

Just to be sure, I've dozed the front panel with contact cleaner while activating the switches back and forth, but it made no difference.

I'm not quite sure where to go from there to find this short. Any insight?

[wn1fju]

Something doesn't add up in your description, or I am not understanding it correctly.  Are you saying that if you have disconnected the -12V output from the P/S and you have disconnected all the -12V supply wires to the boards, you still have a short?  On which wires - one of them, all of them?
Would point to a wiring problem.  Jiggle all of the wires while an ohmmeter is attached.

But my read of the schematic shows the -12V lines different than the connections you listed (AD, T, A, H, AN, CC).  My schematic shows the following: AD to A1, K to A5, D to A2, AN to A4, CC to A3, H to A6.

Normally when I have a short on a power rail, I first isolate the P/S and check the output with an ohmmeter.  If there is a short, the problem is on the P/S.  If the short goes away, then the problem is on the other boards.  I reattach the P/S and leave an ohmmeter hooked up across the P/S output (which reads zero).  Then, one by one, I start removing the connections to the various boards until the ohmmeter moves off of zero.  It is somewhat easy to do this when there is a backplane and the boards all plug in so you can simply lift the boards out of their slots, but in your case, you have to disconnect wires.  When the short goes away, you've found the offending board and it is almost something like a shorted tantalum capacitor that's the culprit.

There is always the possibility that the service manual/schematic is incorrect too.  I would leave the ohmmeter attached at your connection H and then I would start pulling ALL the connection points on a board while watching the meter.

[theor]

Are you saying that if you have disconnected the -12V output from the P/S and you have disconnected all the -12V supply wires to the boards, you still have a short?

Yes, that's right. The boards are fine, but the -12V wires are shorted to ground.

On which wires - one of them, all of them?

Yes, all of them. I mentioned the one that goes to the T pin on the A1 board is fine, but it's an output pin coming from the AD input, so obviously it's fine once I disconnect AD.

Would point to a wiring problem.  Jiggle all of the wires while an ohmmeter is attached.

Indeed. I think I may have a clue: when I bought this oscilloscope, the fan wasn't spinning. Upon removing the rear panel, I noticed a screw was missing from the fan support. I suspect the previous owner tried to fix it and either a screw (or more likely the nut behind), felt inside. Since I carry this unit around, this metallic part could have moved and now be stuck somewhere making a contact to ground.

So I guess I'll have to disassemble the whole thing to understand what's going on...

But my read of the schematic shows the -12V lines different than the connections you listed (AD, T, A, H, AN, CC).  My schematic shows the following: AD to A1, K to A5, D to A2, AN to A4, CC to A3, H to A6.

You're right, I missed the K pin to A5, and as mentioned the T pin to A1 is an output pin. Thanks for your insights!

[theor]

While disassembling the rear of the scope, something became loose and I heard a metallic piece dropping. A nut!



This missing nut should have alerted me when I replaced the fan...

Unfortunately the short is still there but I think I'm on something. Maybe the screw is down there also, shorting something.
Nothing seems to rattle so it's probably stuck somewhere, like the nut was.

The wires are insulated and the pins are in plain sight, so it must be making contact somewhere else. If you were a lost screw trying to wreck havoc in a scope, where would you hide?

[George Edmonds]

Hi

It looks like a screw on the opposite side of the tag strip to the nut

G Edmonds

[theor]

Closer yet. I desoldered each of the eight -12V wire from the tag strip, and eventually found the culprit wire. No more short on the rail except for this wire.



This connects to pin AN on the A4 board. Now I'm puzzled because I don't see what could be causing a short on this wire.
More disassembling will be needed...

[theor]

I couldn't find why this wire has shorted. Maybe a faulty insulation. I couldn't find any foreign metallic piece either.
To solve the problem, I soldered a new wire from the tag strip to the AN pin. This won't win an elegance price but it works.




I notice the power light indicator and the calibration signal are missing for some reason, I will troubleshoot that next. Thanks for your help, case closed!

[theor]

Quick follow-up in case this could help someone in the future. I found the short: a washer was biting the -12V in a braid of wires. The braid was clamped just beside the screw, so I couldn't easily move it to see the stuck wire.

Also a coaxial ground braid was making slight contact with the power lamp and calibrator pins, pushing it away and adding a bit of kapton tape solved that last problem.