NIM Shaping Amplifier Issue

[Tyler Christensen]

Upon testing my TC202BLR amplifier, I've noticed a very strange behavior. I'm running a signal input from a pulse generator and then adjust the amplifier so that it is well peaking the +-10V of the amplifier output capability. Then, I leave it, and over time the voltage drops on the scope attached to the amp output. After 20-30 minutes, it's down to less than +-1V. Upon turning it off, it goes up again over time (if left for 5 minutes off and turned back on, it would be a few volts peak to peak). I can only assume this is thermal, but I'm not sure. Any suggestions on this? It behaves the same with a GM tube hooked into a preamp feeding the shaping amplifier, so it's not the pulse generator.

[Doug Coulter]

It does sound thermal. If the waveshape is changing too, probably power supply issues, else something else. I always troubleshoot those with hot/cold - some freeze spray or holding a soldering iron here and there (not touching) while watching. Usually you can find things fairly quick.

It's the intermittent ones that are hard to get to repeat that give service techs gray hair.

Also, the basic divide and conquer strategy -- if a multistage amp, hang a probe someplace in the middle and see if the gain is changing there -- you have then eliminated half the places it could be, repeat. This takes more finesse if there's a negative feedback loop involved, as if the failure is near the end, all the signals before it will go *up* in trying to compensate the loss after them.

Of course, this means you have to have access to the running circuit, so you need an extender (or skip using NIM stuff that's hard to troubleshoot).

[DaveC]

Yes - it could be thermal, certainly the simplest guess. I was guessing an electrical cause... leaky input capacitor, diode, even dirt ...charge transfer could be dropping over time.

Without a schematic... it's hard to help much....

Dave Cooper

[Tyler Christensen]

Are those black potted plastic boxes inside transformers? I've been wondering what they are since they can't be opened. I assume they're common in all the preamps, if not I'll take some pictures of them tonight.

[Jon Rosenstiel]

The preamp socket pin-outs on many older Tennelec amps are Tennelec specific, connecting another makers preamp creates a short circuit on one (or more) of the power supply rails. Result is usually a smoked resistor or two and low (or no) voltage on the amps + and - 12 V rails. I would suggest checking the amps PC board for burned resistors and monitoring the rail voltage to be certain it remains constant over time.

My 2 cents...
Jon R

[Jon Rosenstiel]

Tyler Christensen wrote:
> Are those black potted plastic boxes inside transformers? I've been wondering what they are since they can't be opened. I assume they're common in all the preamps, if not I'll take some pictures of them tonight.

Probably delay lines. (Used in the pulse shaping network).

Jon R

[Richard Hull]

Jon is correct. While Tennelec made great stuff, their 9 pin connectors must not be taken as standard in every unit. (Must be physically checked). The smoked resistors are usually 10 ohm 1/8 watters. Also be aware that a whole gang o' tantalum caps in all Nim modules are old and hyper leaky in many cases.

What happen is that when cold they are OK. the 10 ohm resistors placed in all power supply input lines on some boards will allow the distributed power voltages to sag as the capacitors warm and head towards a short. This drop in power buss voltages or just one buss voltage will show up as bizarre operation without obvious pointers beyond a darkened 10 ohm resistor here or there.

If you plan on locking in a specific older NIM module to your system, you are wise to replace virtually all of those older tantalum electrolytics on the power busses.

Richard Hull

[Tyler Christensen]

Thanks for all the tips. I did some thermal testing and determined the area with the bad power bus, and indeed a +-12V point pair creeps from 12V down to +-2V after warming up. There are a pair of tantalum caps on that bus which I assume are the culprit. In fact, one is cracked at the top upon inspection.

So, my question is, is there any reason they need to be tantalum? Can't I just put an electrolytic in their place being that it's just a DC filtering capacitor? I'm thinking size and type aren't really relevant. Am I wrong in this?

[Wilfried Heil]

You can use aluminum electrolytic caps for power supply bypassing, but you will need 10-20x higher capacitance for the same HF filtering effect as the tantalum caps.

I would replace all the old caps with new tantalum caps of the same or slightly higher value.

Wet electrolytic capacitors will dry out with time and are guaranteed to fail, sooner or later, depending on the temperature. They are used in price sensitive consumer electronics that will be tossed in a few years anyway.

[Richard Hull]

Wilfried is correct. You should endeavor to use tantalums as replacements. Electrolytics are fine for short term operations 1-2 years, which might be all you need. However, modern tantalums don't have the reliability issues the early "jelly bean" tants' used in old NIM gear had and should be good forever.

Richard Hull

[Tyler Christensen]

I replaced the caps that I thought were bad and the problem did not go away, and I managed to hunt the problem down to fairly certainly be within one of those black boxes as pictured, which are cast in epoxy. I assume there's no way to identify and replace it? It's stamped only as AX1. I guess I'll be on the lookout for another to hopefully appear on eBay in the near future at a low price.