A1200 Composite Video Problem

GrahamN

New member
Joined
Apr 25, 2015
Posts
9
Country
UK
Region
Essex
I've spent the last hour or so scratching my head with this A1200 problem as it isn't the usual composite video problem with a dodgy delay line.

I have an A1200 board that works fine on RGB, but not on composite. I initially diagnosed a faulty delay line (checked by shorting pins 16 & 18 on the video chip and replaced it with a replacement that I believed good. This again failed to work, so I decided to replace the video chip and associated circuiry with a more modern CXA2075. This however still didn't solve the problem. I can see a video output on pin 20 of the chip (and indeed can connect directly to that chip to get a very bright, slightly unstable display. However, while the video signal gets to C237 +ve, there is no corresponding signal on the -ve pin. I have tried a number of capacitors there and even just connected a capacitor to pin 20 of the video chip with the other side of the capacitor going directly to the monitor with no success whatsover. Removing the capacitor and replacing it with a temporary wire link gives a very bright output, but I simply can't get the video signal to pass though a capacitor. I temporariy removed C236 in case something on the modulator circuit was causing a problem, but again with no better result - indeed disconnecting the output from C237 and connecting it drectly to the monitor still doesn't work.

Seemingly the composite video signal is simply not able to pass through a capacitor - and I don't think I've ever seen that before.

I should add that this is a PAL board, and I am using a DiagROM for testing. Also that the RGB signal is completely fine and no other faults seem to be present on the board.

Any ideas?

Thanks

Graham
 
I would start be checking that the composite socket has a good connection to ground on the outer contact. If the gnd isn't good, the signal has nowhere to go to, so it's not keen on travelling any further (through the capacitor). Also check that you have good continuity between the negative pin of the capacitor and the composite socket centre pin.


Bryce.
 
Thanks for your thoughts. The signal comes from the -ve pin of the capacitor through a 75 ohm resistor to the socket. The path to the socket and the socket itself was checked and is good. To be absolutely sure I removed the capacitor, connected the +ve lead of the capacitor only to the board (this goes directly to pin 20 of the chip) and ran the -ve lead directly to the monitor. The earth connection was also made to an alternative earth contact. The monitor showed no picture. Trying the same technique but using a wire to where the +ve connection of the capacitor went to the board rather than fitting a capacitor worked fine (though what looks to be too strong a signal). As far as I can see, the signal coming from the CXA2075 (and previously from the original CXA1175) just doesn't like to pass through a capacitor - and that just doesn't make any sense unless I am missing something fundamental!

Graham
 
And you are 100% sure that there isn't a crack in the trace right beside the capacitors pads?

Also what value / type / voltage capacitor are you using?

Bryce.
 
I even tried temporily soldering a lead directly to pin 20 of the CXA2075. I tried the originally specified 470uF electrolytic (Panasonic - of the same type I've used on dozens of recapping A600 / A1200 boards without issues). I also tried a basic 100uF electrolytic capacitor that I've used hundreds of times to convert Spectrum computers to output composite video. It's as if the capacitors are blocking the AC component of the composite signal or the signal is too weak to get through - and neither makes sense!

When I started trying to diagnose the problem, I used the components fitted to the PCB and temporarily connecting a resistor of up to about 1K straight across the capacitor gave me a picture - so I'm pretty confident anyway that there is no problem with the PCB itself.

Graham
 
And the 75R resistor measures ok?

Bryce.
 
Yes - but even a capacitor only from the composite video out pin should surely show something on the monitor?
 
True. Then you need to measure the voltage and power of the signal. As you say you do recaps, I assume you have a scope?

Bryce.
 
Yes - I get a signal that looks OK (from the chip), but I don't currently have a working Amiga here to compare it with. I am starting to wonder if I have a faulty CXA2075 - I only have the one at the moment, but I've ordered some more, so may try replacing it when they arrive.
 
Does the signal still look ok if you load the circuit?

Bryce.
 
Hello,

i have done this with CXA IC

https://www.amibay.com/showthread.php?48994-RGB-gt-CVideo-S-Video-and-VGA-self-made!

i know a little how output look like, on a working device.

First of all you can check VREF, that is most import on this type of device; and that Ground and +5V have not crack on the solder joints.

You can measure with the scope this output

sync
luminace
chrominance before filtering
cvideo

I have here, more or less, the right values of this signals.

Regards
Stefano
 
Back
Top Bottom