-group-therapy-thread/?action=dlattach;attach=1451218;image)
Vince, good deal! 
But obviously you can't expect it to be fully functional when you have a dead tube. 
Worked some more on it...
Checked the power rails. ONCE AGAIN, the manual show the WRONG test points on the ceramic strip ! It is completely unreliable/useless. Had the same problem in the manual for my tpye 502A !
Why can't Tek freaking put arrows on a pictures on the correct terminals ?! Is it THAT HARD ?!
So instead I looked the power supply schematic to figure out by myself with 100% certainty where to probe....
All rails were well within spec. The -150V reference was at -150.9V. For the heck of it I adjusted the best I could. That would be -149.99V.
Ripple I did a quick check with a DMM, could not be bothered to look at it on the scope. First tried with my MX 56 DMM to get good accuracy, but its RMS converter is so freaking slow, takes at least 30 seconds, at least, to slowly converge to the correct value. Drives me nuts. So I resorted to using my Fluke 11 instead. Resolution only 1mV, but it's not RMS of course, so I could get instant readings at least, to keep my sanity. Its B/W is ridiculous of course, but for 120Hz ripple it does the job. No real need for a scope either... I mean it's not a SMPS with mega huge spikes all over the place... it's a linear supply, it's good enough measuring it with a DMM. At least to ballpark it. I multiply the DMM reading by about 3 to get an idea of the peak to peak value, since ripple in the manual of course is expressed peak to peak.
So, doing that... all rails are just fine, ripple well within spec. It's all a no brainer. So the big filter can caps can stay for now. They are not causing my problems.
Then I replaced the dead / white tube. A 12AU7 it was. When I went to pull it, the glass envelope came right off !
Had no glass left to speak off, to get a hold of the tube... so had to pry the remains out of the socket, using a small flat head screw driver...
While I was in there, I resocket every tube in the thing. Took a while, there are a ton of them...
One my 6 type 317 donated one of its two 12AU7...
Then I tested all the ranges again.
The few that (kinda) worked, still do. No better or worse. However I now have life in most other ranges (bar the 1 and 5s ones) !
Very crude form of life, however : amplitude often extremely unstable, makes the signal look like a plot on a spectrum analyzer with tons of harmonics of random amplitude. Very funky.
Time wise it's very funky too, with periods being anything but they are supposed to be. Well, in most cases any way.
1ms setting gives 160us
5ms setting gives 6ms.
10ms gives 17ms.
50ms gives 80ms.
100ms gives 260ms.
500us, which used to give between 250 and 400us depending on mood, now gives only 80us, same as the 100us range.
It's baffling. It does not make sense... the OCXO runs fine, so you would think that the frequency dividers would give exact multiples of the OCXO period...
How can a flip-flop produce random divisions like that ?!
So me thinks.. there is no flip-flop. Not enough tubes to implement flip-flops in there... so me think it might using this clever analogy "divide by N" trick, using just a few discrete analog components. We were discussing that about ancient counters some time ago. The divide by N functionality was achieved by using the time base pulse, of course of known shape/duration/amplitude, to charge a capacitor. After 'N' pulses charging the cap, it gets you to a known voltage level which you can detect using a comparator.
So maybe the Tek uses something along those lines, which would explain how the time intervals can be completely random and independent from one another.
I see that every range has a trim pot associated with it ! That further reinforces my theory.... I guess one can use it to adjust the analog divider to make it divide properly... I am sure all these trimmers are just fine. The problem is probably tired tubes.. they all look original, very filthy and dark... I fear in order to get proper operation on all ranges, I might have no choice but replace a bunch of them, which would make this restoration very expensive indeed... hence not going to happen in the near future.
Will RTFM to learn all about it, but I think I might be on the right track here...
What I can do / afford for now (barely...) , is buy a new can of contact cleaner to clean all those push buttons, as well as replace all the crappy black/red Sprague caps, and that ancient looking, paper shrouded " Cornell Dubilier " cap as well.....and see where that gets me. It needs to be done anyway, it's basic stuff. So let's do that first, and go from there...
Stay tuned....
Home
Help
Search
Login
Register
