TDS-744a screen rolls when warmed up

[cncjerry]

I'll take a picture tomorrow, but the wire connects to two pins around that cap, C399, I asked about previously. The board connector is labeled as J399 on my 744a board and J400 on others. Now I am not sure if that is the same connector. On the 544 board (j400), only one pin is square but on mine, both pins are square.  Tek standards are that the square pin would be the red pin or pin one.

Edit:
I have to look at my notes again, because I found that connector on the 544 schematic, and though the 744a service manual drawings look like that is the right connector for that wire, it doesn't make sense when I read the schematic for the 544.  The red/black wire is coming off the yoke separate from the deflection cables.

Thanks

[Tantratron]

Let's see your pictures then compare with my own references of real boards (four units).

One quick method to eventually help you out, most of these 2 pins cables (connectors) at least in older CRT boards have always one pin GND. So just use a DMM ohmeter to find who is who, which is which on both ends including processor board then you'll plug the connector correctly.

[cncjerry]

I checked the connecter where I think the cable goes and both pins were floating..

[Tantratron]

Ok then if really you have a red + black wire connector coming directly from the CRT tube wether Black&White or NuColor, this is just TRACE ROTATION trim which might have no polarity, just a floating potentiometer (look schematics).

[TERRA Operative]

It's just the screen rotation connector, it's a coil around the crt to rotate the image.

I don't remember which way is correct but just plug it in one way, and then the other. Whichever is the straightest image is correct, then tweak the associated trimpot to get it perfect.

[cncjerry]

Thanks all for confirming the trace rotation.  I'm going to plug it in this morning, out of 58 caps replaced, I'll probably have at least one in backwards.  I checked them four of five times now. I hope this thing comes alive.  It was my primary scope for a long time, just love that deep black screen.  Wish me luck. 

[cncjerry]

Well, I thought I had sent this already, but I plugged it in, it was fine for 5 minutes and then started doing the same thing.  Basically the colors roll all over the place, interesting thing is that the trace seems to be consistent. The background grid and everything seems to be rolling color wise, but not the trace which seems ok.  The only cap i didn't replace was the large 470uf, 450v.  Thoughts?  I can't believe this thing survived all those leaking capacitors.

Thanks for all the help of course.

[Tantratron]

Have you tried madao's adjusting suggestion ?

[cncjerry]

I played with the adjustment pots some.   This is not a vertical hold problem where the entire display rolls up, it is like the colors are being impacted by a low frequency.  I'll take a video and put it up on youtube.  Weird thing is that it starts out ok, then after about 3 to 4 minutes, it starts to fail.  Then if you turn if off and, assumably, let it cool, it immediately fails when you turn it back on.  It's like a cap is charging and holds that voltage for a long time.  You would think that if it was heat related, it would sort of follow the same cool, heat cycle.

[Tantratron]

Yes a video to show exactly the issue would help.

Could you at the same time re-connect an external VGA display so both your CRT and VGA shows image at the same time then report if VGA works always whatever CRT degradation ?

You might probe at the same time the 80 MHz quartz which does play a key role to drive the NuColor shutter, maybe it is the quartz which is overheating or about to fail.

Lastly did you fully recap the CRT electronic board and its vertical card in charge of the color Shutter unit ?

[cncjerry]

I was searching to see if my video was public and found this one:

https://youtu.be/Kk2ICQeuf58?si=ZYXKNpEmpINsbEun

exact same problem as mine.  Another person posted asking if the ov had ever fixed it without reply.  So I know there are three scopes with the same issue.

I recapped everything in the display board including the vertical daughter card except for the 470uf 450v in the PS as mentioned below.  Between the power and crt, I replaced 64 caps.

The only one I regret not replacing is the 470uf 450v because I would have to wait until October to get one with the same footprint.  I am going to replace it though, and bodge in the best option but the cap tested as good in circuit.

When it fails, I still have great video on the VGA port.

I tried blowing air in from a portable vac, no luck.

If I crank the brightness up using the pots on the horizontal PCB,  the flickering stops but the display is washed out. There are two pots that impact the brightness, and bringing either down from the washout brings the flicker back.   Ripple on the high voltage?  Maybe it is that large cap (or the display unit). 

The 80Mhz crystal counted 78hz high and doesn't vary.

AI mentioned something about the driver transistors, but I can't get that recommendation back.

Thanks for all the input.

[TERRA Operative]

How reliable is your counter?

78MHz is way out of spec for that crystal, it could be causing your sync issues.
It is rated to +/-100ppm (0.01%) so you should be seeing a lowest frequency of 79.992MHz, your reading is 2.5% low (assuming you meant 78MHz instead of 78Hz).


I attached a datasheet for the S15R8 80MHz crystal on one of my spare boards here.

[Agent24]

I don't know much about these, but are you sure you didn't accidentally set it into disco mode?

[cncjerry]

The crystal was 78hz high, 80,000,078hz referenced to a cesium beam.  I'm a timenut 

I can't think of anything other than the 470uf 450v cap.  I could also rebuild that video driver daughter card, doesn't look like it'll be that hard.

Any other ideas?  I wish I was able to cable this thing up so I had access to everything.

Thanks for all the help.

[Tantratron]

Since you might put your hands on the video driver daughter card in charge of Shutter system (3x60 Hz NuColor patent), not sure but how trying to disconnect its 2x6 cable coming from the main CRT board. The NuColor shutter system would not operate at all but you might have some monochrome display to verify if you still have disco style effect after warm up.

[TERRA Operative]

The crystal was 78hz high, 80,000,078hz referenced to a cesium beam.  I'm a timenut 

Ah, I misread. That crystal seems fine then.

[cncjerry]

Since you might put your hands on the video driver daughter card in charge of Shutter system (3x60 Hz NuColor patent), not sure but how trying to disconnect its 2x6 cable coming from the main CRT board. The NuColor shutter system would not operate at all but you might have some monochrome display to verify if you still have disco style effect after warm up.

That's a good idea.  The fact that it is 3 x 60hz points at the problem, I wonder where it gets the 60hz, because that's what looks to be failing as it gets out of sync and just ripples.  I'll look at the schematics again.

[cncjerry]

I noticed that not all the colors seems to be rolling.  For instance, trace 4 on my scope is like cyan, and that seems to be ok.  Over on tekscopes, they mentioned the neck board transistors failing...  Of I go.

[cncjerry]

just some more data points:

I mentioned the VGA output looks good and that typically means the 80Mhz video processor crystal is fine.  I've measured it with my counter and it is stable but when I measure it with the scope, it jumps around quite a bit.  The shortest gate on my counter is .1 seconds.  It's strange that my scope finds it varying between 78Mhz and 81Mhz.  I'd have to put a known stable signal into the scope to see if it is just the behavior of the scope or what.  I can also inject the signal into my spectrum analyzer sitting right there.

Putting the scope display in hardcopy mode where it paints the background white, removes the color waves and is useable - everything looks ok.  Seems like the color waves / disco effect only impacts the background and background text.

If I play with the brightness, I can slow or speed up the color waves up and down the screen.  It seems to be impacted by the high voltage.

I'm going to check to see if I have the driver transistors on the scope neck card and replace them.  I see newer versions of that card have a heatsink on one of the drivers.  If I don't have the transistors, maybe I can just swap two and see if there is any impact.  I bought out a tek repair person's stash of spare electronic parts 8yrs ago so I might have them.

Thanks for all the input.

Jerry

[Tantratron]

I mentioned the VGA output looks good and that typically means the 80Mhz video processor crystal is fine.

If your VGA output works, it does not mean that your 80 MHz quartz is fine because for the VGA, there is no need of this quartz. The 80 MHz quartz is used for other parts of the processor board video's section to generate the specific NuColor sytem signalling.

One key thing, the video section of all these TDSxxx are native digital 640x480 (color or mono depending on ID resistor). Back those days, tektronix only degraded crisp pure digital pixel image down to either VGA analog and separatly to CRT shutter (NuColor CRT patent) or monochrome CRT. But the core video generation is pure digital 640x480 which is why CRT to LCD kit works great (i.e. Simmconn labs) wether 80 MHz quartz or not.

Albert

[cncjerry]

well, disco is dead - for now. I was ready to give up, so I pulled it apart again,  took out my shutter jumpers, etc... Just for grins, I rotated the HV pot back and forth and then for more grins, turned it clockwise 1/4 turn.  Hasn't failed since.  Ran for two hours now - used to fail in under 5 minutes.  Doesn't mean it is fixed, but clearly a major improvement.  The display is clear and black with depth and vibrant colors.  It's not washed out at all, very vivid.  I'll post a picture, but I need to setup a hood.

I'm going to let it run this way for a while and see if the problem comes back.

Sometimes it's my turn I guess.

thanks for all the help, I'll let you know if this is a solid fix. 

Jerry

[cncjerry]

well, it didn't happen if there are no pictures.  The picture is a little out of focus, CRT is fine.  I also took a shot of the VGA output on a Dell monitor.  The Dell isn't the best, but it was more expensive than your typical monitor having the ability to raise & lower the arm as well as rotate horizontally and vertically.  I couldn't get the glare off the picture but normal viewing if fine.  This is why I wanted to fix the 744/784 display as I am a sucker for the sthat deep black.

Thanks for all the help.  I believe the scope is fixed for now.

Jerry

edit:  when I uploaded the pictures, they were all washed out, etc.  I posted them in tekscopes:

https://groups.io/g/TekScopes/photo/310116/4071663?p=Created%2C%2C%2C20%2C2%2C0%2C0

[Agent24]

Nice, I couldn't figure out how to view the photos on that website, but glad to hear you got it fixed!

[Tantratron]

Have you tried madao's adjusting suggestion ?

Glad you finally got fixed, after all @madao (Matt) was right from the start