Tektronix 422 PSU all voltages too high

Thread Starter

DarthVolta

Joined Jan 27, 2015
521
It's a 1965 transistorized oscilloscope. I got it off ebay, and I did have a trace for a moment once, then shorted the AC somehow when wiring on a new plug, other things could have happened right before that too. I really don't know how it's aged or history. It looks incredible inside,very clean and gold and silver everywhere.I should have re-capped it 1st.

Here's the PSU
http://w140.com/tek_422_ac_powersupply.pdf
big schematic on p35

Since then I re-capped the PSU side, which is fully detachable from the rest of the scope. However I notice a bunch of voltages are way off. Wrote them all down. (some transistors aren't on ) All the main transformer voltages are also high. My household AC is at 123V. So I wired the T601 for 125V, before it was for 115V. It barely lowered any regulated voltages.

The -12V rail is -13.9, and with 100Ohm load went to -14.4V
The +12V rail is 17.7V, never loaded it yet
They should be +/-0.25V
The un-regulated 55V rail should be about 61V, I'm at 81V

So I put it on the auto-transformer, and the +12v at 17.7 pretty much stayed above 16-17V until I was under 80VAC for less, I check in more detail again later.

I had just put in a new D622, a 9V1 Zener, it's the reference voltage for the -12V reg. .It and the other 22V zener D655 both read fine.


For now , on T601 on page 35, I have un-soldered all windings except the -12V reg. So I'm going to apply an external +12V to R634 for the base and collector of Q633/34. Then see if the -12V reg. works or not, and check with the auto-trans too, if not.


Also, I modeled the PSU in LTspice, and it regulates just fine. However I really don't get where some of the listed voltages on the schematic come from. And even in the circuit theory part of the pdf. They have the top of the 9V zener at -6.9V on the sch., but say -6.4V in the text (on p12). But neither of those even makes sense for a 9V zener and a 390ohm resistor across 12V. LTspice agrees with me.

In real life it measures -4.6V, with -13.9V as (-12V) rail. and the 9V1 zener at 9.25V
 

Thread Starter

DarthVolta

Joined Jan 27, 2015
521
Yeah its not that. So with 32VAC on the -12V rectifier, and 12.5V acting as the 12V rail, I'm getting -12.7V, with the ref. zener at 9.22V. Thats as low as I can adjust the 1.5k pot. I'll try with exactly 12V and see if it matters.

So -12.7V is what it's aiming for, with the various resistors and silicon worn in as they are I guess.

Changing the AC input voltage only changes the rail by tens of milivolts. Now I'll try w/ a load, and then try just the +12V rail with a stand in for -12 and 55V rails
 

Thread Starter

DarthVolta

Joined Jan 27, 2015
521
Thanks. With a 100 Ohm load I get -14.7V, so that's way off. With 200R it's -14.6V. Changing the 12V battery level didn't help, with 9V the -12V rail is even wider, at -12.8V. I checked the beta's of the diff pair, and they are about 40-50 apart iirc. So I'm going to try and match a pair of 2n3904 and try again.

No way is this thing meant to be that far off, it has to be the silicon or the resistors, which more or less are right.
 

Thread Starter

DarthVolta

Joined Jan 27, 2015
521
IDK I'm afraid to try, and it needs re-caping or testing at least. But I will try loading down all the rails. It only has 40W power rating, that's nothing.
 

dl324

Joined Mar 30, 2015
16,846
Yeah its not that. So with 32VAC on the -12V rectifier, and 12.5V acting as the 12V rail, I'm getting -12.7V, with the ref. zener at 9.22V. Thats as low as I can adjust the 1.5k pot. I'll try with exactly 12V and see if it matters.

So -12.7V is what it's aiming for, with the various resistors and silicon worn in as they are I guess.

Changing the AC input voltage only changes the rail by tens of milivolts. Now I'll try w/ a load, and then try just the +12V rail with a stand in for -12 and 55V rails
You're not following the calibration procedure in the manual.

What are the voltages at the test points (bases of Q624, Q633, Q637, emitter Q624)?
upload_2019-8-28_7-29-46.png
 

Thread Starter

DarthVolta

Joined Jan 27, 2015
521
Well at 1st I adjusted R639, right now it's at 1.5k (all the way cw anyways) and still only get's me the -12.7V

So with just 7-8-9 wired to the trans, @ 33Vrms , a battery for the 12V rail
The base voltages L-R are,

-3.48 -3.3 -3.76 -4.36, emitter q624= -4.06

and with 100R load they are
-5.65 -4.58 -4.2 -4.83 , q624= -5.25
 

AlbertHall

Joined Jun 4, 2014
12,345
The voltage between the -12V rail and ground is -12.7.
Then you connect 100Ω from the -12V rail to ground and now the voltage is -14.7V?
That's really cuckoo. Loading the supply increases its voltage.
 

Thread Starter

DarthVolta

Joined Jan 27, 2015
521
Yup that's exactly what it's doing. I've been too lazy too try and calculate what some voltages should be, but LTSpice has the 2 transistor types, and it has no problem regulating as the line and/or load changes.

I should check the betas on q633/37, the diff pair are about 560, 610
 

AlbertHall

Joined Jun 4, 2014
12,345
Next bit of cuckoo: with or without the load the bases of both Q633 and Q637 are negative, their base emitter junctions are reverse biased and hence they should be turned off! Where is the current through the load coming from?

I suggest you remove these two transistors from the circuit and measure the output voltages again.
If the voltages are now much lower then replace those two transistors.
If the voltages are still the same then call ghostbusters!
 

Thread Starter

DarthVolta

Joined Jan 27, 2015
521
All the resistor are within 5% or better, the caps are not shorted, IDK if I checked their ESR, that was 1-2years ago. Learned a lot since then.
All 4 transistor read good, Q637Vbe=0.4809 and Vbc=0.472, I guess thats about right for a power transistor.

Q624/Q634 are labeled (A) 618 151 151 . In the manual the Tek label is something else, so I'm guessing someone replaced them at some point. They are supposed to be akin to a 2N2219, which in 1 data sheet says max beta 300, and LTSpice uses 255. These have 560/610.

Let's see what LTSpice does with similarly high beta's

Nothing, it's still at -12V with beta's of 524/530


All the traces and nice and thick, with gold plating, or very gold like, on top.
 
Last edited:

Thread Starter

DarthVolta

Joined Jan 27, 2015
521
Ok I think I know the problem, and it was money, I didn't have a 2,000uF cap and I used a 22uF, and I never bothered looking at it on a scope yet, LSTspice sure doesn't like a 22uF tank cap for the rail. I just noticed, I have to solder a bunch back together, try again
 

Thread Starter

DarthVolta

Joined Jan 27, 2015
521
Yup that was the main problem with the -12V rail, and the rest of it is basically working, or trying to. There's another problem tho, at only around 5Hz (yes five), the thing is hiccuping and breaking regulation, with the -12V going too low, and +12V going up again.

I doubt it supposed to do that either, since LTspice doesn't. The ordinary 60Hz ripple looks fine. More probing required.

Maybe it really wants a higher minimum load to work, . But I never saw anything in the PSU pdf that says to expect this if testing the PSU alone.
 

MrChips

Joined Oct 2, 2009
30,712
I know a bit about 422s. I have a couple and the service manual. I can test a working power supply for you but it will have to be put on hold for a couple of months.
 

Thread Starter

DarthVolta

Joined Jan 27, 2015
521
The other unregulated rails I was worried about that seem so high, will probably drop a lot with a load, and actually there's regulation done on the scope side, so they are probably fine too. I just have to load check the PSU. Hopefully the hiccupping it's doing is normal.

The HV driver doesn't work without the coils on the scope side.

Then the real work begins, I might take take the scope side apart as much as I can, do check all compnents, like a real restoration job. At the very least I'll check the ESR of big caps. Maybe try and build Mr.Carlson's LV cap. tester.

I've never worked on anything higher voltage than a 380V PFC boost converter in SMPS. So I have to learn the basics of CRT HV, this is the perfect project.

Ton's of stuff I can learn from working on a transistor only scope in general.
 
Last edited:
Top