What is my next step to debugging an old NTSC video generator circuit that doesn't sync?

Question

I know this is somewhat broad but I hope it's appropriate here. I have a 1970s PCB that is designed to output a sort of bitmap video image via a standard NTSC signal. It came to me as a half-built kit (abandoned in the 70s) and I've completed it by adding all the ICs. (It had all the passive components and a few transistors already present.) The output video signal does not sync to the monitor I have (an old analog monochrome monitor, very forgiving). The trace of the video signal being output looks incorrect to me. The timing is reasonable, but the signal never drops below the 0V level as you'd expect it to to trigger a line sync. Here is the signal it outputs:

Here is a signal with the same scope settings from a standard NTSC signal generator-- this is the 5 stair setup greyscale signal with colorburst but you can clearly see the drop below 0 for the sync:

I'm certainly not surprised that this old board would put out a messy sort of signal, but it doesn't look valid which is why I suspect the monitor can't lock on to anything. I've checked the resistors which all seem to measure fine. There are two old electrolytics but they also measure OK. I'm doing this as a learning experience, so I'm fine probing around and checking stuff but I'm not really sure where to look for the part of the circuit that "would bring the level down below DC zero for the sync".

This is the part of the schematic I think is most relevant (shows the timing generation stuff and the video out line is near Q4 in the lower left center):

If you were me, what would you be looking for or checking out next here? I don't want to overload here but there's some theory of operation on page 14 here.

Thanks so much for any suggestions.

Answer 1

First you should understand how the analog video output is generated. The instructions say

Video data from IC3 pin 6 is routed to IC15-D where it is mixed with the appropriate blanking pulse. The outputs of IC15-B and D are then mixed in Q4 to give the required composite video output.

But exactly how the signals are 'mixed' is not explained. It works like this:-

^{simulate this circuit – Schematic created using CircuitLab}

IC15b and IC15d are AND gates with open-collector outputs. When an output is low it acts like a switch to ground, when high it is open circuit.

If both gate outputs are high then the voltage at Q4 Base is pulled up to 5 V via R24. This is the 'white' level.

When IC15d is low it connects R25 to ground, forming a resistor voltage divider with R24 that lowers the voltage on Q4 Base. This is the 'black' level.

When IC15b output is low it pulls the voltage down to ground. This is the 'sync' level.

Q4 acts as a buffer which amplifies current to ensure that the termination (typically 75 Ω) doesn't load down the voltage divider.

R26 sets the output impedance to (hopefully) around 75 Ω, and R27 provides a load to ground to pull the output low when going down (Q4 only pulls up).

Looking at your scope trace it looks like the signal is having trouble pulling both down and up, and the high level has a lot of noise on it. Assuming you are looking at the signal without being terminated (by eg. a monitor) this suggests that:-

1. R27 is not doing the job. If R27 was not present then the only path to ground might be via the scope probe, which has a very high resistance.

2. R24 is not pulling up properly, or Q4 is not amplifying. Perhaps the value of R24 is too high, or the transistor is installed backwards.

If a monitor is connected then some of the distortion may be caused by the DC restore circuit inside the monitor. It would be particularly bad if R27 was a high value or missing.

Another possibility is that IC15 is the wrong part. If it did not have open-collector outputs then the outputs would fight each other. Or the fault might be just a mundane dry joint, eg. on the Collector of Q4.

First thing I would do is check that all the part values are correct, transistor installed correctly and all joints good. If that doesn't reveal any anomalies then scope the Base of Q4 with no termination on the video output. If the ICs are working this should show a good waveform.