Re: 468 sweep offset issue


 

On 24 Feb 2016 15:53:50 -0800, you wrote:

I am not sure what you mean by voltage offset. With no sweep, there
is always a voltage offset because the blanked beam is located at the
left side of the CRT.
Is the bow on the right 1/4 of the CRT visible in your photos?
It is only slightly visible in the close up photo. It is most noticeable when I have the scope set up per the calibration procedure for Y axis and Geometry adjustment. I will upload a photo of the display symptoms shortly.
This may be just a characteristic of the CRT. Tektronix sometimes
specified a tolerance for geometric distortion inside and outside of
the center area. Toward the edges, you can usually find some geometry
error if you look closely enough.

According to the schematic for the horizontal amp, when the scope is set up in XY mode (as per the voltage measurement conditions) both plates should measure close to 52V. My 110V rail is 2V high, so it should measure 54V, but either way they should both be the same with the dot centered. Left plate measures 56.7V and Right plate measures 54.3V.
It is normal for the CRT to have some vertical and horizontal offset
to produce a centered beam. To compensate for this, the CRT
amplifiers may have a centering adjustments. x1 horizontal centering
on the 468 is adjusted with R551 shown on schematic 10 and x10
horizontal centering is adjusted with R578 shown on schematic 10.
There are separate vertical centering adjustments for each input which
do the same thing.

The horizontal position control is off-center to the right side of the display by approx 2 divisions.
Also, when I center the trace in a normal timebase setting and then switch it into XY, the dot is positioned that same 2 division to right offset.
This should be corrected by the procedure on page 4-64 if the
horizontal position control is already centered. Tektronix appears to
have left that out.

Are you referring to the start point of the A and B sweeps not being
exactly aligned?
The alignment depends on circuit tolerances and there may be no
adjustment for it. At high sweep speeds, it is normal for the start
of the sweep to creep to the right somewhat.
I can probably suggest where it can be trimmed out if we determine
that there is no actual fault.
Yes, I am referring to the A and B sweeps not being aligned. My other 468 doesn't have this problem at any timebase settings. (That one needs more work than the one being discussed.)

What really bothers me about it, and also has me slightly confused, is how to properly adjust the sweep start and sweep cal settings when taking this problem into account.

As per the horizontal cal procedure (it's the first step in this section):
-Set timebase A to 1ms and B to 5us
-Set delay time to 1.00
-Input 1ms time markers
-Adjust so the pulse displayed on B sweep starts at the beginning of the sweep & start of 2nd time marker from A sweep display**

The pulse on B sweep is not displayed if I set it to where I believe it should be - which is coinciding with the start of A sweep (which is the first time marker)
If I don't have it set exactly there, won't B sweep end up being out of tolerance?

I suppose I'll have to go through the waveforms again now that I have a digital scope and see if everything syncs up with the sync pulse, A and/or B sweep start, and blanking/unblanking. When I had checked them with my 465, everything appeared to be correct. Some waveforms just had minor differences in their voltage levels when compared against the manual spec, but seemed OK when compared against the other 468.
Their calibration procedure is pretty complicated and Tektronix does
not say what the goal is unfortunately. I suspect some of the steps
could be done in a different order for better results.

I definitely want to fix this and get them aligned. I guess I'm a bit OCD...
I sometimes trim the values of various parts or match transistors to
fix problems like these.

There is only one Sweep Start adjustment for both sweeps and based on
the design shown on schematic 8, there could be 100s of millivolts of
difference between the start points of the sweep generators do to the
matching of CR953/CR954 and CRsomething/CRsomething shown on schematic
8. Differential amplifier/comparators made up of Q958/Q960 and
Q987A/Q987B control the sweep start voltages. Oddly enough, the A
sweep uses a dual transistor which will have better matching while the
B sweep uses two separate transistors; I wonder why.

With an estimated CRT deflection of 1 volt per division at the sweep
output, a difference of 1 minor division or more is possible and your
first photograph shows an offset between the sweeps which is within
the design tolerances.

We have diagnosed other 465B type oscilloscopes here which had as much
offset between the sweeps.

If you want to trim the offset out, then I can make suggestion for how
to do it but it will require modifying the circuit slightly.

I just have a couple things to do and then I can get another picture or two uploaded, then do some waveform checking.

Join TekScopes@groups.io to automatically receive all group messages.