Well, I have now finished the first and most important calibration procedure. The rest is more related to diodes and shorts, but the actual circuit you use most of the time is now calibrated. It was kind of a pain, and the army manual, while having all the information in there, puts the step that should be first near the last, IMHO. The plate voltage adjustment step really adjusts the meter circuit, and therefore all of the prior calibration steps that required you to be at line set condition are now screwed up because you have to play with the meter to make the plate voltage correct.
See, it turns out that the plate voltage, like the other voltages, are directly related to the the line voltage adjustment. However, unlike the bias and other voltages checked during calibration, there's nothing you adjust to change the plate voltage except the line adjust. In other words, the plate voltage stays a constant and everything else gets adjusted around it. If the meter doesn't show 60 at line check when the plate voltage is at 150, then you adjust the meter circuit, not the plate voltage.
So the army manual totally wastes your time in my opinion, by having you do tests and adjustments that will get wiped out should the meter circuit not be calibrated. That's military intelligence for you.
Knowing what I now know, the things I did in this restore would be completely rearranged in the following way. Let's go over that before I outline the final steps in calibrating this thing and how I did it.
First, I should have checked R123 and R125. If those weren't close to 1K, but more importantly closely matched to one another, I'd replace them.
Second, I'd inspect and/or replace the meter diodes, CR101 if need be.
Now, knowing that the meter circuit is out of whack from new resistors and diodes with a likely higher output voltage than the original, I'd skip straight to the step where we compare plate voltage to the meter at line check.
Now with the new diodes, the meter is reading a high line check when the plate voltage is at 150v. It should read 60, and in my case it's reading 68. Because it is too high, I have to compensate by shunting a resistor across the outputs of the meter diodes. The manual suggests 40k-60k, but this turned out to be too low. Probably because of the voltage from these diodes, I achieved the desired 60 reading with a resistor of 8.2k. Here it is in place:Loading Image...
I just added it right on the output terminals for the diode. I think it looks right at home there.
So, now having gotten the meter down to a proper 60 at line check, with 150 volts coming from the plate, then I'd do the screen grid check. In my case it came out to 130 on the nose, which is exactly what it's supposed to be. Looky here, no adjustment was necessary now that the meter is calibrated! Why did they put this near the end???
The next test, you press diode and mut. buttons at the same time (buttons 2 & 3) and you should get 56v +- 1.5 I wound up with 68.9 and so had to adjust clamp B on R130. R130 is shown here:Loading Image...
Clamp A is the bottom clamp, the one we adjust for the bias calibration. Here we adjust clamp B, the top one, until 56v is obtained. I managed to get it to 56.2 so, I'll call that done.
Now having gotten that right, I had to go back and re-do the first steps where I calibrated the bias knob, adjusting clamp A on R130 a few times until the bias voltages were correct.
After this was all done, it is now calibrated as far as I can tell. Tubes that tested below acceptable before (when I knew they were good) now test quite well, and give readings that I would expect them to give. Obviously, with meter deflection never being very linear, and tolerances being what they are, you're lucky if any two TV7 testers are in 5% of each other when calibrated. So only time will really tell me how I did. But all of the voltages are now correct, the critical resistors are tightly matched now, and the diodes are brand new. It should be smooth sailing from here on out.
At some point I will update this with adjusting the shunt portion of the circuit for diode testing, etc. and using a resistor to verify the shorts tester is working correctly. Then there will eventually be the cosmetic end of things.
That's all for the moment. I'll look at this again soon and try to follow up with anything that I've missed or gotten wrong. Probably will need to do that when I'm less tired.