HP 1740A on the bench, help please?

[Specmaster]

The CRT wasn't a different, it just had the blue diffuser missing from the front so I was looking directly at the CRT 

[bd139]

Ah well keep it

[Specmaster]

Donor timebase shafts and knobs, removed, cleaned up and refitted to working scope, nice and easy. Now to read the manual and calibrate it a bit as it is slightly out on some ranges and also the flyback can be seen on certain ranges so perhaps tweaking here and there may cure it. After that its a clean up of the fascia and I reckon after that its a keeper. 

[bd139]

Nice work!

[Specmaster]

Nice work to you as well because you and a few others also helped with loads of useful advice so it was teamwork.

[Specmaster]

Murphy has struck again, replaced the 990k input resistor on Ch A, all was working nicely, perfect square wave etc on both channels volts/div was reading spot on, nice steady trace, so put covers back on the night and switched off for dinner.

Came back after dinner, and thought that maybe I'd look at the offset on some ranges so I switched on and no trace, pushing beam find button had no effect either, using the vertical position controls then operate as screen illumination controls and absolutely no horizontal activity at all.

Back to basics reveals that the power rail -15V is down to 1V, pull of the A14 power bridge board and power on, check -15V and its back. Replace A14 and power, -15V is still up, check the 12v rails, both are fine then trace is back again and is looking fine .... Sweep seems to be working just fine again ARRGHHHH. 

I'm pulling handfuls of hair out wondering just WTF is going on here, do I have a Tant playing up on the -15v rail?? 

[bd139]

That’s an interesting one. Possibly. However 1v is suspiciously high for a tant failure. When mine went, it dropped to about 400mV at the output of the regulator all three times one went.

I think if it’s working now it’s probably an idea to run it for a few minutes, power down and disconnect line then have a feel of the caps on the power supply board and see if anything is getting warm.

[tautech]

Don't forget about checking for failed bridge rectifiers like the one Dave did a vid on.

[Specmaster]

Seem to have a dodgy contact somewhere around the timebase switch boards bt wriggling and push pulling on the delay timebase control, sometimes the trace will go off or come on and sometimes the trace brilliance level changes. Can also repeat this by using the well known and scientific method of tapping the case sharply with an instrument can cause the display to flicker, go dim or bright.

[bd139]

Sounds like the switch needs dismantling and cleaning. The switches on these are notorious for swing dicky. The one I had dropped out if you pushed the timebase knob.

[tggzzz]

The timebase switches are easy to dismantle, but make sure you don't lose any of the internal fingers.

Carefully check the PCB tracks where they leave the switch: the edge of the switch abrades the tracks. If that has happened, simply jumper across the gap with a single piece of thin wire, being careful not to allow solder to creep onto the switch contacts.

I was not amused when I saw that switch design.

[CJay]

That’s an interesting one. Possibly. However 1v is suspiciously high for a tant failure. When mine went, it dropped to about 400mV at the output of the regulator all three times one went.

Just be careful around Tants, always have  colleague still has the scar on his thumb from when one went off, he was prodding around a power supply that was 'singing' as the load changed, he worked out that if he touched this particular tantalum cap the singing changed.

Unfortunately it spontaneously disassembled itself when he prodded it to demonstrate the source of the noise to another colleague.

[bd139]

Yes they're pretty vicious if you get them in the wrong way round. Make sure you check the polarity at least 5 times!

[Specmaster]

Lesson learned the easy way, thanks guys, if I feel the need to check on tantalum caps for temperature I think I'll stick to the good old infra red thermometer, not a bad thing anyway especially on HV circuits.

[Specmaster]

Update.
Certainly glad I got that parts scope, took the time base boards from it, cleaned up the board as it was carpeted with filth, almost as deep as some carpets to  took the switches apart and gave them a really cleaning with Isopropanol, checked for wear on the tracks and also where the traces meet the switch contacts, looked like they have seen very little wear, reassembled the switches and swapped the boards over.

I now have a nice and steady trace with no hint of a bad connection, when the tap test is done, trace remains

[bd139]

Excellent - good news 

[Specmaster]

 :wtf:It went off again

[bd139]

It's only doing that because it wants some attention. Sounds like something is failing. I bet it is tants or the filter caps.

[Specmaster]

Update.
Since it went, I left it off for a few minutes, upon switching on, everything was back to normal and has now been running OK for about 90 minutes so I tempted fate and switched off and then on again, no trace and no trigger and no UNCAL light on the timebase despite the vernier being in the UNCAL position.

Switched off for about 5 seconds and back on and everything is working fine and is still working fine after 15 minutes. Its as if there was a built in timer to prevent the scope from working straight away following a switch off and on again without giving it a few seconds delay? 

None of my other scopes behave in this fashion, any ideas?

[Specmaster]

It's only doing that because it wants some attention. Sounds like something is failing. I bet it is tants or the filter caps.

The supply rails all measure within spec when the display goes.

[bd139]

Are any LEDs still on?

If they are then it's likely the HT side of things. Check the oscillator is running when it goes.

[Specmaster]

All the leds are on with the exception of sweep UNCAL one. The two leds for the input verniers work just fine.

[Specmaster]

Oh, I forgot to mention, the 2 timebase boards I took from the parts mule, cleaned up etc, when I put then in, the flyback lines that were visible on certain settings have all disappeared and I noticed that they had been partially recapped at some point before being left to gather dust, all neglected and left alone in the corner

[Specmaster]

Are any LEDs still on?

If they are then it's likely the HT side of things. Check the oscillator is running when it goes.

I'm not too sure how to do that. Are you talking about one in the HV box above the PSU? I don't have anything to measure high voltages.

[tggzzz]

Switched off for about 5 seconds and back on and everything is working fine and is still working fine after 15 minutes. Its as if there was a built in timer to prevent the scope from working straight away following a switch off and on again without giving it a few seconds delay? 

I had exactly that phenomenon in a Tek 485. It was a b*gg*r to find since I could only do one test every 6 hours. The switching PSU needed to be kickstarted into operation, and that was done by a small inconspicuous electrolytic capacitor. Replacing that permanently cured the problem.

It looks like the only part of the 1740 drive by the SMPS is the CRT, so you could test that hypothesis by looking at the Y-output while it is not working.