OK, spent many an hour on the thing again...
Played with the NOS trubes, swapping here, swapping there... I can't seem to get the exact DC levels indicated in the schematics. I can get very close, almost spot on.. (within a few tenths of a volt), but never spot on. .... in these circumstances, the best I can get is less than a 2 major divisions offset on the screen. Still a lot, but can be easily compensated with only a 1/8th of a turn on the vertical control knob.
So I started to think that probably I am chasing a red herring again (a trademark of mine for sure...
)and that the figures indicated on the schematics are merely given as a guidance, rather than accurate values that require to be achieve for the scope to work properly. Plus, people on TekScope said that these scopes were designed such that they would will work just fine even with pretty worn out tubes that would even fail in a tube tester, but work just fine in the scope ! A bit like the good practice design that dictactes that a good transistor circuit should be able to operate properly regardless of the huge variations of Beta from one transistor to another.
So, at least for a moment, I took the problem the other way around (<which I should have done from the start I guess, truth be told...) : start from the last/fourth stage of the amp, and work my way backwards.
So, I grounded the input, then centered the vertical control knob on the front panel, that brought the trace one or two DIV below the center line. The I measured voltages. Across the vertical deflection plates, I get like 15V or so. Let's say 15,8V for the case /round of measures for which I did take notes on paper.
I get that voltage upstream : at the input of the delay line, ie at the output of the last/fourth stage of the amp (V244 / V254). The upper tube gives 158,7V and the lower tube outputs 174,5V. Take the average of that and you get 166,6, not far from the 168V quiescent voltage indicated on the schematic. So, Looks like it works fine ? The output is symmetrical, with each tube pulling or pushing an equal amount form the 168V bias point.
Well... that would be fine if I had an a signal at the input that is ! But there isn't any of course. So how comes the last stage produces such a large signal in the absence of an input signal...
So I checked the input of that stage, the control grids. I do a get a voltage difference there, a small one : gird of the upper tube sees 88,9V and the lower tube sees 88,15V. That's not much of a difference is it... 0,75V.
How can 0,75 turn into 15V... well when you do the math, it's a x20 amplification, spot on. Sound like a nice roudn number, not too big, not too small... maybe it is the actual amplification that this stage is supposed to provide... so that would mean it works just fine ? Probably.... no luck then, still no answer to my problem ! :-/
So, I move upstream one step : the third stage (cathode followers, not amplifiers). These two cathode followers (V233 A and B) appear to drop about 3Volts each. The upper one gives 88,9V as siad earlier, and at its control gird I measure 85,0V. The schematics suggest 84V , one volt off.
As for the lower cathode follower, 88,15 at tit's cathode/output, and at the input/control grid, I measure 85,5V. Schematics suggests 86V, so 0.5V off.
So, the upper cathode follow drops (88,9V - 85V ) = 3,9V.
the lower tube drops (88,15V - 85,5V) = 2,65V.
No idea if this is normal or a sign of a problem. Both triodes are in the same tube technically, so you would think they are decently "matched". so either the mismatch comes from the external circuitry and is normal (since I measured all the resistors and they are fine), or from a defective tube.
I pulled the datasheet (attached) for the 6CL6 tube (amplifier/fourth stage)... boy I can't believe it but it's actualyl available. First time I look at a tube data sheet... not exactly my era. Still, it looks surprisingly familiar... data sheets are all allike I guess, be ti tubes or integrated circuits ! LOL
Looks like the transconductance is of the thing is huge, 10,000+ at the least... so given that we have onlky a gain of 20, that means that must have built bags of negative feedback in, plenty enough I guess to indeed make the amp/scope immune to variations in transconductance from one tube to another, and for a given tube as it ages/degrades. So, I guess this is one more clue that this stage is operating properly, I guess... so back to square one : WHAT is wrong in this amp....
Looks like I will need to seek help from the tube expert on TekScope...
Was starting to make me a little depressed, so to keep moving, I decided to check the calibration of the vertical section. After all, the offset can be easily compensated for, using the verticak control knob on the front pane, and the calibration procedure doesn't state that it is mandatory to have zero offset in order to be able to calibrate the thing.... maube it was implied, oh well... The calibration is pretty simple/quick to do anyway, so I didn't mind having to do it all over again, should fixing the offset problem happen to affect the calibration. At least it would give me a first idea of where the scope stands on the vertical side of things.
So I did that. Followed the calibration procedure very closely, no skipping any step fo course, doing it all by the book.. while looking at the schematics to try and understand why the calibration procedure was sequenced in the way it was. This way you get to learn a thing or two on how the scope works/is designed, always interesting...
First thing yo need to do is calibrate... the "calibrator" (probe compensation signal), because that is what they then use a reference to perform the calibration.
If you turn the calibrator knob ot the "OFF" position, you get a 100BV DC voltage on a test point in side the scope. Because it's DC and not a square wave anymore, you can easily adjust accurately with a DMM, which I did. the cal pot allows for an accurate setting, took only seconds to bring it spot on to 100V. Was only a tad too low, just a hair.
Then you can start to calibrate the vertical amp :
- First the "Variable Attenuation Balance" control, which I talked about in my previous post.
- Then the (main) amplifier gain.. was quite a bit off, 20% or so. Now fixed. It's trouble-less : once set on one range, it stays correct on all the other ranges as well.
- Then we adjust the gain of the pre-amp (for the AC only, high sensitivity ranges). Easy enough again.
- Then the attenuator high-frequency compensation : you can't use the internal calibrator signal anymore, since its edges are not fast enough. So I used an external signal generator. Looked pretty decent to me so I didn't touch anything... I left all those little ceramic trimmer caps alone...
- Then the attenuator Input capacitance : goal is to compensate the scope to work with the bespoke x10 probe. Yes you compensate the scope itself.. not the probe, as you would do with more modern scopes.
I do have that probe, somehow still works fine 55 years later... a tad under copensated it seems, buit nothing dramatic... I chose to play it safe and not mess with those ceramic trimmers...
- Pre-amp low frequency compensation : was impossible to do because there is too much "humming" in the way. In this case the manual suggests to put the cabinet covers back on, but I won't do that until the scope is fully restored/cleaned, and ready to be put back together. So, I might come back to play to this particular step, later on. From the looks of it though, it again seems a little under compensated.
- Then there is ..... there is.... oh my god.... there the delay line !!! I guess you need a PhD in astrophysics to master the calibration of such a thing, so I didn't even consider, not even for a split second, messing with this thing ! LOL Especially since the scope works just fine as it is...
Then... I went to check the horizontal side of things, the time base.
Overall it was half a major division too short. Not a big deal, but still plenty noticeable, so I adjusted that.
Manual tell you to use a "Time Marker" generator... never seen on of these, so instead I used my programmable pulse genrator. allows to prodcued calibrated pulses, and can run in "continuous" mode, making it kinda like a signal generator.
Only problem is that said pulse generator is defective : the fequency is very unstable, jumps all over the place. So in order to keep an eye on it, I added a frequency counter to the mix : I thought my old 8 digits Nixie tube counter would be fitting with that old tube scope. The pulse generator was kind enough to at least give me the first 4 digits stable. Plenty good enough for the 1% accuracy we can about dream of, for the scope. So I used this setup, while going through the long and tedious calibration procedure...
The first few steps are pretty straight forward : adjust the gain in "Magnified" position, the adjust the gain in "normal" mode. this takes seconds and allowed me to get an accurate time base easily. Then comes the more strange things, like this "Magnifier registration" setting... boy if you don't know about it, no way you could figure out how to calibrate it. The the less fun stuff : going though the fast sweep speed settings, and fine adjusting the sweep speed AND linearity, via ceramic trimmer caps. The sweep speed and linearity interact with each other, and no matter how long you play with them, you just can't get set it 100% right. you have to come up with the best compromise and accept it as it is. The manual doesn't even pretend otherwise, they just tell you basically : "good luck with it !!! " LOL I tried it for "fun", to get a feel of it, on the 1us/DIV setting... saw how much of a headaches I was getting, and promised myself not to mess with the other sweep settings as a consequence ! LOL Well unlike I had too... fortunately I didn't have too, they looked good enough to my taste.
So, it's only a 10MHz little scope yet there are quite a bunch of trimmers to play with and it's a nightmare to get everything just right, as you get into the high-frequency settings. Which leads me to : in the same era, I gather the 500 series scope could go 10 times faster, 100MHz BW ?! Now I am scared just to think of how much more of a nightmare these beats must be to calibrate at high frequencies !! ..... or maybe not ? I mean, unlike the compact/portable 317, the 500 scopes were very large, which means Tek could stuff a lot more tubes into their cabinet, which means they could afford to come up with more complex/sophisticated/refined circuitry... maybe that meant they could produce designs they were easier to calibrate ? Don't know, but I hope so.. for the sanity of any 500 owner ! ....
Sooooo.... looks like this scope is getting and better. Vertical is calibrated, horizontal is calibrated and working fine too, so is triggering. CRT HV is stable and strong...
The only problem is this vertical offset which I would like to get to the bottom of, if just out of pride... but the truth is that it can now be compensated easily with the vertical position knob, and that the scope is actually perfectly usable now.
So I guess I could now start to clean/refurbish it.... I can always keep working on this offset problem after it's been cleaned. Cleaning was postponed at first, because I was in the dark and didn't even know if I could diagnose the various problems, let alone fix them. But now the scope is clearly working and just need minor debugging for this offset problem. The thing is basically operational now.
So, let's now start making this thing shiny again....