Hey Vince.....have fun.
I will, and I have ! It's all fixed now !!
So the problem in that final, 5s stage was indeed again the cap.
As I said it's not a Sprague, it's a General Electric, big metallic can bolted to the deck. It looks very fancy, it looks sophisticated, it looks the business, its looks... expensive. It looks like the last cap on earth you would incriminate.
But, every evidence points to it so... I looked in the parts list, it's listed as type " PMC " , which stands for " Paper Metal Can " . Yes,
PAPER !!!
So I disconnected one of its wires so I can measure it. It measured just fine, close to its 1uF rating.
But because it's paper it's likely to leak a lot so... I replaced it with a film cap, and that fixed the problem !
So AGAIN a cap that measured perfectly good with a DMM, but leaks badly when in service.
So that's it, I have now diagnosed and fixed all of the 14 stages of this puppy. Now works just fine.
So in the end we had 3 bad Mica caps, a bad expensive looking can cap, a few Sprague that although were not "bad", had drifted too much so the trim pot had run out of adjustment range so the stage could not be made to function properly. Oh, and a broken wire, and a dead/white tube.
So this Tek was a nice adventure. A new kind of TE I had never owned / worked on before. Learned some cool stuff, and of course learned that YES, Mica caps can test just fine yet leak badly !
It has to be said that working on all stages starting from 100ms time interval, was... not fun. Signal is so slow that you can't use an analogue scope anymore, you can't see squat.
Switching the Combiscope to digital mode was a saviour, as you get a nice trace no matter how slow the sweep speed... HOWEVER it was useless at first, because at such slow sweeps, the SAMPLING rate is ATROCIOUSLY slow as well, which means the scope can't capture the markers/spikes ! So many markers would either not show up because they happened to be in between two samples, or for the "lucky" markers that got blessed with a sample, depending where exactly the sample landed on the marker, marker would have a, seemingly, amplitude that jumps all over the place from zero to 100%. So basically you just see markers dancing all over the place like vu-meter in an audio console...or not showing up at all. So you can't know what is due to the scope or what is due to the instrument itself, for real.
Thank Goodness, I tried the " Peak Detect " acquisition mode, and it improved things dramatically. All markers would now show up, and their amplitude would be stable, where they should be. I could then at last pretend to troubleshoot the instrument
However in " Run " mode, at these ultra mega slow sweep speeds... go figure, the scope can't trigger reliably ! So the markers would move all over the place horizontally, making it again impossible to trouble shoot the freaking thing !
But.... I noticed that if I trigger it manually, using the Single Shot button / feature... somehow that fixes the problem
So I was single shooting repeatedly, to mimic the "run" mode by hand. Doing that a dozen times in a row, I always got a 100% consistent picture, never an odd result. So I had a decent degree of confidence that I got it all working and adjusted properly. But... it was not fun. It was boring... having to wait ages or two to get ONE screen... then do it all over again many times in a row... wow.....
Since this old Tek was of course meant to be used on analog scopes back then, and since any marker above 100ms is unusable in practice... I wondered what was even the point of having them available on the instrument to begin with ?!
Sure, the glowing Tek scopes of that era have sweep speeds up to 5s, so makes sense for the time marker to go up to 5, I agree, makes sense but again... you can't even see the freaking markers so....
Then it occurred to me that ... well, even back then you had STORAGE scopes... I have a couple type 549 with a storage tube.
With these yes of course, you can get a nice stable trace on the screen no matter how slow the signal.
So, this now gives me motivation to fix one of my type 549 so I can see if I can get them to display a cool super slow marker picture !!
One more project on the pile !
So, now that this marker generator is all finished.. it was time to put it to the test : display SEVERAL markers at the same time, to display a "comb " thing. Would it work at all ? If so, how good or bad would it look on the screen ?
So I made this lovely picture just for you people. A comb of 3 markers, the 3 slowest ones : 5s + 1s + 500ms.
It looks absolutely..... fantastic !
Their amplitude adds up perfectly, one third, two thirds, 100%.
When 2 or 3 markers combine, they look like one single marker, not some "soup" ugly Frankenstein.
It looks as good as I could possibly dream them to look like.
That's what I call a nice piece of TE, a new tool in the lab one can actually use ! It's not an ornament in the living room !
To begin with, simply looking at that comb let's me check my Combiscope itself !
Looks like it's well calibrated. It's slightly of on the extremities, but it's just parallax error I guess, not actual error. The graticule is not built into the CRT.
I very much enjoyed fixing that thing, learned stuff and now have a cool new kind of TE in the lab, instead of just yet another, 35th scope !
So now just need to order a few film caps to tidy it all up, then a good clean and that will be it, it's good for service now !
The mechanical restoration and paint job will have to wait until the garage is built.
Have a good Discord !