Topics

7854 revived with a new CRT - how to correctly calibrate Z circuit (was: #157693 - https://groups.io/g/TekScopes/message/157693)


unclebanjoman
 

Hi all,
more than a year ago, I initiated that discussion: https://groups.io/g/TekScopes/message/157693 about my 7854 with severe focus/intensity problem.

Finally I got a replacement CRT guaranteed working some days ago.
I replace old CRT with the new one and.. finally!!! Trace is brilliant, sharp and focused. The readout also is very well focused, even at high intensity.

Undoubtedly Albert Otten was right when he hypothesized the phenomenon of CRT's double-peaking for my case!

Now, with this new CRT, some re-calibration is necessary, due to slight changes in deflection sensitivity, step response, etc. So I started from scratch.
I want to clarify that the scope actually works well, I have not noticed obvious anomalies (for now).

I follow the calibration step-by-step as per manual.
I have come to section B, "Z axis and display - page 4-47" and here I found two things that baffled me.

Section B-3 "Adjust Z-axis DC levels": I correctly managed steps a thru g obtaining an 80 Vp.p. signal at TP122, board A21.
Continuining to next steps I read this:

i. Connect the test oscilloscope 10X probe to TP183 (located on the Z-Axis Board) with the probe ground connected to chassis ground.
j. Set the 7854 B INTENSITY control fully clockwise.
k. EXAMINE—Test oscilloscope display for a waveform baseline between 8 and 12 volts above ground reference.
I. ADJUST—The Output Level adjustment, R135 (located on the Z-Axis Board A21), for a waveform base line at 10 volts above ground reference.

Question: what is "baseline"? The bottom portion of the waveform? If this is the case, R135 has no effects on the level of bottom portion.

Continuing:

m EXAMINE— The test oscilloscope display for a 68 to 72 volt peak-to-peak vvaveform. (Do not move the test oscilloscope vertical Position control.)
n ADJUST—The Z-Axis Ampl Gain adjustment, R125 (located on the Z-Axis Board A21) for a 70 volt peak-topeak vvaveform displayed on the test oscilloscope.

Here I can adjust: I set for 70 Vp.p. But the lower portion of waveform is at approx 7-8 V about GND.

Can be the manual wrong? I think the the regulation at points j/k/l should be made witch INTENSITY fully CCW, adjusting the R135 (Z axis output level) for top waveform level about 10 V from GND. I'm thinking wrong?

The second problem arose during the calibration of the Good SIgnal Flip-FLop GSF (page 4-50, section B5 of the manual): using a DVM I set the voltage at TP103 (board A21) 40 mV less than the voltage at TP122 with no problems.
Using two identical timebases 7B53A in slot A and B, I connected the probe of my test oscilloscope at TP106, board A21.
7854 set to Horizontal mode B: I see a good pulse signal, with levels approx. -0.2V and +3V. The signal is stable while rotating B INTEN through its range.
7854 set to Horizontal mode A: I see a the same pulse signal, same levels, but it's "slippery". For a few instants is stable and triggered, then become jittery, slips and then return somewhat stable. Adjusting GSF Level, R102 (located on Z-Axis Board A21) has no noticeable effects.
Thinking about some problem in the 7B53 in A slot, I swapped the two plugins: same result: GSF signal is stable in HORZ MODE B. Unstable when set to HORZ MODE A .

I don't see any motivation for this behavior since the FF circuit is very simple.
Really I don't know why this happens, since the scope seems to work normally. I will follow the Z signal back to schematics <4>

The 7854 manual seems very unclear in this cases. I do not have another 7854 to compare the signals and I don't remember what the signals on A21 Z axis board was, since my last tests two years ago...

What do you think? Am I doing something wrong or is the manual to be wrong?
Any suggestion is welcome.

Cheers,
Max


Dan G
 

On Fri, Jan 29, 2021 at 09:39 AM, unclebanjoman wrote:

Question: what is "baseline"? The bottom portion of the waveform? If this is
the case, R135 has no effects on the level of bottom portion.
"baseline" is the bottom portion of the waveform. The service manual is right,
but the confusion may have come from the labels on the A21 PCB: there
are TWO trimpots labeled "OUTPUT LEVEL". One is in the lower right
corner of the board -- that's R70, and the other is near the center of
the board -- that's R135.

I have just replicated the adjustment procedure from the manual, using
a 7B92A as the time base in Horizontal Slot B, and a 7A26 in Left
Vertical Slot (adjusted to make sure the trace is as far off the screen
as possible!).

R135 (the correct one!) does indeed directly control the level of the
bottom portion of the waveform, as expected. In my case, the levels
were already correct (10V baseline, 70V p-p), so there was no need
to adjust.


I hope this helps,
dan


Dan G
 

On Fri, Jan 29, 2021 at 09:39 AM, unclebanjoman wrote:

The second problem arose during the calibration of the Good SIgnal Flip-FLop
GSF (page 4-50, section B5 of the manual): using a DVM I set the voltage at
TP103 (board A21) 40 mV less than the voltage at TP122 with no problems.
Using two identical timebases 7B53A in slot A and B, I connected the probe of
my test oscilloscope at TP106, board A21.
7854 set to Horizontal mode B: I see a good pulse signal, with levels approx.
-0.2V and +3V. The signal is stable while rotating B INTEN through its range.
7854 set to Horizontal mode A: I see a the same pulse signal, same levels, but
it's "slippery". For a few instants is stable and triggered, then become
jittery, slips and then return somewhat stable. Adjusting GSF Level, R102
(located on Z-Axis Board A21) has no noticeable effects.
Thinking about some problem in the 7B53 in A slot, I swapped the two plugins:
same result: GSF signal is stable in HORZ MODE B. Unstable when set to HORZ
MODE A .
This is an interesting observation, and one that I was able to reproduce with my
7854 using 7B85 and 7B92A time bases. HORZ MODE A is the key -- it doesn't
matter which time base is installed in which slot.

Observing TP122 with 7A13 (using a different mainframe, of course) in voltage
comparison mode allowed me to see what is happening. Namely, with Horizontal
Mode set to A, there are occasional spurious pulses (about 20ns wide) that
show up out of sequence. They get converted to TTL levels along with the real
pulses, and so they throw off the display and make it look like it is "slipping".

As far as I can tell, U105 is just doing its job here and converting these spurious
pulses along with the rest, since they have the same amplitude. Hence, I don't
think this "slipping" indicates an incorrectly set GSF level. I think as long as you
can observe all the real pulses at TP106, with A and B INTENSITY varied
throughout their range, then the GSF level is OK.


dan


unclebanjoman
 

Hi Dan,
thanks for your explanations. Very clear and concise.

I've rechecked carefully the adjustments of R135 and finally I adjusted the baseline to 10 V.
Regarding the GSF, I readjusted to get TP103 40 mV less than the voltage on TP122.
Varying the A and B INTEN controls through their range shows stable TTL pulses between -0.2 V and +2.4 V, so I can assume that the Z section is perfectly calibrated, disregarding the strange behavior when switching to A timebase that you confirmed.

Now I'm facing with another problem: since I changed the CRT,I had to recalibrate all the vertical and horizontal offset, gain etc. in the analog section.
I used the 067-0587 standardizer. All things goes o.k.

Then I made some measurements using the digitizer part.
It turned out that the digitized waveforms had wrong amplitude (and some offset) compared with the real waveform (using BOTH command to view the acquired and real waveform).
So I managed to quicky recalibrate the vertical pickoff system, precisely the vertical sampler offset (A25R380) and vertical sampler gain (A25R290).

Now the digitized waveforms looks o.k. in amplitude and step response, but they are wrong in horizontal scale.
Superimposing the real and digitized waveforms the latter is more expanded than the former (shows an error of almost 0.8 div over the complete 10 div of graticule).
Moreover, I noticed that the first acquired points seems not filled (shows a perfect flat line).
I assume that surely the horizondal must be recalibrated also.
Should I perform all the tedious steps in the H section of the service manual, starting from H1 and up to H6 in strict sequence, or is there a more quickly procedure?

Thanks again for your advice and observations, Dan.

Max


unclebanjoman
 

After careful study I decided to touch and adjust only the Digitizer Horizontal Gain, step H11, R42 on board A9 (very difficult to reach the trimpot, augh!).

Now the horizontal scale is in agreement with the real signal scale, and the two waveforms (real and digitized) superimposes almost perfectly.
So I prefer do not touch all other various calibration points, since the digitized waveforms shows no gross and obvious aberrations.

The 7854 is a splendid scope but calibrating it gives me a headache!

Max