Jumpy [B&W] composite video signal from 1974 dumb terminal

[MightyFrame]

I'm not sure if this might be beyond the scope of this forum, so if it is, please forgive me.

I'm very much a newbie to the technical understandings of vintage video, and I'm hoping that some of the experts here can point me in the right direction, maybe give me some pointers or explain what I might be doing wrong here.

Basically, I'm looking at a B&W composite video signal coming from a very rare vintage dumb terminal from about 1974. I'm using a USB composite video capture device to view it, but the video signal is very jumpy.

It's apparently a PAL signal at 625 lines, but as far as my research shows, this modern equipment should be able to handle a standard PAL composite video signal no problem, without this jumpiness.

Here's a very detailed video showing the problem, and what I hope is [too much?] information about everything I know about it, and can possibly share up to this point.
Video of the problem detail:
https://youtu.be/nTPZRe_TK2U

If you're interested in diving into this, I very much appreciate your willingness to watch this video.

Some of the resources I show in the video are hyperlinked right here for convenience:

My page with all technicals about the terminal
http://bit.ly/2XpxETd

Terminal Motherboard Schematics (that create the composite video signal out)
http://bit.ly/2r1m0Sj

Motorola Monitor-TV Chassis Schmetics
http://bit.ly/35dx11L

Logic Capture File of the signal for 8 seconds
http://bit.ly/2Xso8i3

Saleae 1.1.15 install (for Windows, if needed, to view the logic capture file)
http://bit.ly/2NUJjWU

Video of the actual CRT chassis assembly that was made for this terminal:
https://www.youtube.com/watch?v=qPjS...ature=emb_logo

Video Capture Card StarTech.com SVID2USB23
http://bit.ly/32S2l4p

Thanks so much everyone, and I really appreciate your help and feedback on this!

Best always,
AJ

[maxhifi]

Without reading all of the links, or knowing anything about this specific terminal, I will say the following:

Some devices made in the past, which could show a picture on a TV of the era actually produced a non-standard video signal, which was close enough to work on the TV sets of the era.

If at all possible, try to use the terminal with an era appropriate monitor, and see if you have the same problem. My guess is that you will be able to tweak the controls on the monitor a bit to make it stable. My guess is the converter likely expects a signal exactly conforming to standard, and the terminal just can't/won't do this.

[old_tv_nut]

Originally Posted by maxhifi

Without reading all of the links, or knowing anything about this specific terminal, I will say the following:

Some devices made in the past, which could show a picture on a TV of the era actually produced a non-standard video signal, which was close enough to work on the TV sets of the era.

If at all possible, try to use the terminal with an era appropriate monitor, and see if you have the same problem. My guess is that you will be able to tweak the controls on the monitor a bit to make it stable. My guess is the converter likely expects a signal exactly conforming to standard, and the terminal just can't/won't do this.

Ditto. The sync signal may be a non-standard waveform, plus it is likely that the video is progressive scan at a horizontal rate close to PAL, rather than interlaced. Both of these things are likely to confuse your video capture device if it is looking for a standard PAL signal.

[old_tv_nut]

The data analyzer will not be much help, as the video signal consists of at least three levels: white, black, and sync. You need a scope to display it.

The monitor is a typical simple analog circuit in which the horizontal and vertical are independent and will sync whether the incoming signal is interlaced or progressive or even free-running ("random interlace," as it would be with an industrial video camera). Your video capture device is likely trying to lock the vertical and horizontal to each other at exactly the PAL standard, whereas the sync signal from the video generator board is likely progressive (meaning the horizontal frequency and/or the vertical is off from the PAL standard and their ratio is also different from PAL). My bet is that there is nothing wrong with the generator card, and it's just that your capture device doesn't like it.

[old_tv_nut]

By the way, is your goal to capture the output, or just to view it? If the latter, it is possible to add a video input to a small black and white set of this era.

[MightyFrame]

maxhifi and old_tv_nut, thank you VERY much for this very helpful feedback. I'll see what I can do to get the old crt working.

My goal is to do both, capture AND view, but knowing what the problem may be is certainly a great place to start.

I'll work on this, and report back here when I have some progress.

Thanks again!
AJ

[ppppenguin]

old_tv_nut has got this right. A lot of older video generating devices produced a non-standard waveform. For example I have a TV test pattern generator c1970, from a reputable manufacturer, that gives PAL colour bars, crosshatch and a couple of other patterns. The field/vertical sync is non-standard and there is no interlace. The subcarrier to H and V realtionship is wrong too. None of this matters much if you're setting up an older TV set. I've not tried it with any modern sets or other kit with digitally based video decoders. I wouldn't expect it to work very well.

It may be worse than this. The output may not be 625 lines (or 624 if not interlaced) at all. It may be somethng completely non-standard. But this is less likely.

For viewing purposes find an old portable TV with video input. Likely to be colour as few if any B&W sets had AV inputs. An NTSC set should lock to the roughly 625 line signal. The vertical hold and height might need adjusting. Alternatively an old CRT based video monitor. The sort that might have been used in a TV studio or CCTV system.

As others have said, you'll need to look at the video waveform with an oscilloscope to find out much more. Even then, unless you're familiar with TV waveforms, it can be tricky to find all the quirks. As for using a digital capture device, the answer is probably no. You can try different devices and you may strike lucky but more likely you'll get nowhere fast.

[AlanInSitges]

AJ do you have access to an oscilloscope?

[MightyFrame]

Originally Posted by AlanInSitges

AJ do you have access to an oscilloscope?

Sorry for the delay, but YES, indeed I do have access to a scope.

I waited to reply until I could actually record a short video showing the composite video output on this scope.

It's a short 7-minute video where I go through various settings on the horizontal to hopefully show a good full-spectrum of the signal.

https://youtu.be/ZphFzpkUtto

Does this help to figure out what might be going on with this jumpy signal?

Or, should I do some additional tests on the scope, at your suggestions?

Thank you!
AJ

[ppppenguin]

What a revolting video waveform. at 10us/cm the H sync pulses are OK-ish but there's loads of other stuff hanging down below black level during th active lines. It would be a miracle if anything locked to that.

Actually chroma information can go below black but that doesn't look like chroma to me.

[zeno]

This may help you
https://www.ni.com/en-us/innovations...video-101.html

Everything I have noticed ran on 1V P-P ( after the video detector). I
cant say for sure on all monitors, there are different schemes. The second
higher signal may be sync. BTW tradition is to show two complete cycles
( lines ) thats what we are used to seeing.
Good luck

73 Zeno
LFOD !

[old_tv_nut]

I skipped quickly through your video, so may have missed if you adjusted the sync level knob. Youmneed not only to use NORMAL sync, but also turn the level knob toward negative a bit until you are triggering only on the sync (if it's there) and not the video pulses.

Refer to the link that zeno posted to see that the sync should be the most negative part of the signal.

[MightyFrame]

It's been 3 years, and I'm revisiting this project, hopefully to create a device that "normalizes" this "revolting video waveform" (well said ppppenguin!)

I'm going to start with a build of this guy's composite video adapter circuit: https://youtu.be/5FYF5uhCzAM

Hopefully, I'll see if I can develop it into a "normalizer", but still have composite video output to use with my computer/USB capture devices. That way I can record nice clean signals from these very rare system terminals.

I welcome any input from the experts here!

[dishdude]

I think you'd be better off trying to find a multiformat VCR to run the signal through.

[old_tv_nut]

Did you ever determine if the video actually has composite sync on it, or if there are separate vertical and horizontal pulse outputs?