On 10 Mar 2016 23:49:17 -0800, you wrote:

I'm pretty sure that these canned hybrids are always plugged into Berg sockets that sit nearly flush with the board surface. These are among the most reliable IC pin connector types I know of, despite their low profile and simplicity. I've never seen a bad one except where abused by the wrong lead diameter or chemical attack.

I have always been suspicious of the mini-Berg sockets and have run
into a couple cases where good contact was not being made. I think
the machined collet pins which you mention below are more reliable.

The pins on the solder side of the board are fat, indicating they are likely Berg sockets. Tomorrow I'll see if I can pull the hybrids out of one of the spare amplifier boards. This should also tell whether or not there's thermal grease underneath.

There are at least two types of low profile socket pins - the Bergs that I often refer to are made from tiny drawn and crimped sheet metal components, and are usually pressed into the board early in the build process. They often have a white silicone elastomer glob seal on the opening, which is punctured by the part lead when installed. Tek often used these for transistor and IC pins, and also as the center conductor receptacle for board-mounted Peltola connectors.

The other kind are high quality machined sockets, originated by Augat, I think, and commonly used in very good IC sockets and low profile in-board pins and strips. I think HP used a lot of these.

Ed

I cut those out of sockets and SIPs to use them as low profile collet
pins for ICs, transistors, and any other leaded part I might want to
replace later.

http://www.mouser.com/Search/ProductDetail.aspx?R=310-93-164-41-001000virtualkey57510000virtualkey575-193164

On 11 Mar 2016 12:01:09 -0800, you wrote:

I looked at that vertical amp board again and found that the cans are plugged into Berg sockets, and there is thermal grease underneath for better cooling. Also, I hadn't noticed before that two of the pins on the bigger output amplifier are much longer, and were passed though unplated, un-socketed holes, for connection to the deflection plates. There's no solder on them either, and I can't remember if I clipped them off from other leads (very unlikely), or if they were plugged into loose Berg sockets attached to the CRT lead wiring.

What color was the thermal grease? I assume it was white indicating
zinc oxide?

I never removed one of the hybrids from its board.

If the 7854 hookup is made that way, that's another possible location for intermittent connections.

I had my 7834 vertical amplifier board removed several times when I
was messing with the CRT to try and fix its geometry problem. The
first time I did not realize that the leads from the output hybrid to
the deflection plate pins had collet socket pins on the CRT side and I
unsoldered the wires from the collet socket pins leaving the collets
pins in place. The wires were soldered to the thick leads from the
output hybrid.

I must have taken the vertical amplifier board off of my 7854 to
replace R160 but the only thing I remember about it is the lesson from
my 7834; I pulled the collets straight off of the CRT leads instead of
unsoldering the wires. I think the wires were soldered directly the
thick leads from the output hybrid. (I just verified this with visual
inspection.)

On 11 Mar 2016 16:49:17 -0800, you wrote:

Thank you for looking into the TO-8 amplifier arrangement for me, Ed & David. Now I know exactly what to expect.

From your descriptions, I feel the best way to go about pinpointing and solving the issue is going to be to pull the board, carefully pull the U50 driver, inspect, clean off the old paste and the leads with IPA, inspect again, reapply thermal paste, then gently re-insert the amp into the sockets. I might as well do the same for U150 while the board is off.

First and foremost, I am going to put magnification to the board and check all the solder joints of surrounding components and reflow any suspect ones.

Also, only U50 on mine has the plastic clamp.

I checked and all of mine, 7904, 7834, and 7854, are built the same
way. They have the plastic clamp and ground connection on U50,
insulated leads on the output hybrid soldered to wires, and a stud on
the output hybrid bolted to the aluminum bracket under the board.

Hopefully this solves the problem. I did 'spudge' the potentiometers and resistors on the board, but only tapping or pressing on U50 induced the fault. I don't think this alone can rule out the possibility of a faulty component in the driver or amps surroundings. Ensuring all of the connections are positive is definitely the #1 step.

The only major question I have here is about thermal paste.... all I have on hand is Arctic Silver from RS. It's the higher end stuff that they carry. Do you think that will be sufficient? Should I find a resource for a Tek recommended thermal compound - or try to salvage the 25-35yr old stuff that's already there?

Don't reuse the old thermal paste. Tektronix probably used a standard
silicon oil and zinc oxide thermal paste like Wakefield 120. It is
not critical so use anything you happen to already have as long as it
is not electrically conductive.