exxos's Atari, Amiga & retro forum

stephen_usher wrote: 02 Apr 2022 16:54 Yeah, welcome to the wonderful world of analogue pain. :-)

Similarly on my C64 the colour banding gets worse as the machine warms up.

Both are probably due to some sort of resonance somewhere that drifting components hit when they both drift to when they're warm.

Yeah, driving me mad, I really thought I had it and then as soon as I had the mobo back in the case to test before I put the lid on, off it went again. Very, very annoying.

What's somewhat more annoying is that I can't easily test with just the Antic, because I have no alternate video out as I had to remove some items to allow fitting of the 13 Pin video socket, and also the modulator was huge and in the way of the VBXL. I do know that before I fitted the VBXL the composite out was pretty decent. Perhaps I can remove the VBXL and revert to composite, and for the Pokeymax stereo just have a flylead through the RF out hole.

Lets see if it can be fixed first, at least it kind of seemed like there was an improvement of sorts.

Heatsinks on order for the VIC-20 once the new VIC arrives. Hopefully here soon and in the meantime I can look at the 600XL some more.

It's very odd, I am using the same wiring harness that was fine in the 800XL, and having swapped out the Antic, there is no change either.

I really don't want to have to remove the VBXL, so really scratching my head at the moment, but it may become a reality and will require the original video out to be installed again.

I managed to get a better video of the shimmer:


https://www.youtube.com/shorts/5LpZocVyeRI


So if anybody has any ideas, they'd be gratefully received!!

There are multiple frequencies in there. There's a high frequency component, which shows up as the interlaced lines being offset and the low frequency ripple, which is probably a beat frequency between the horizontal scan rate and whatever the noise is, and that's a pretty close frequency due to the slow speed of the ripple.

It's almost as if the original and new video outputs are interfering with each other. How is the original video signal generated on the motherboard?

P.S. I think that you said that you removed the original modulator... I wonder if you need to terminate the inputs with resistors (75 ohm?) to stop the traces ringing.

stephen_usher wrote: 05 Apr 2022 11:43 There are multiple frequencies in there. There's a high frequency component, which shows up as the interlaced lines being offset and the low frequency ripple, which is probably a beat frequency between the horizontal scan rate and whatever the noise is, and that's a pretty close frequency due to the slow speed of the ripple.

It's almost as if the original and new video outputs are interfering with each other. How is the original video signal generated on the motherboard?

P.S. I think that you said that you removed the original modulator... I wonder if you need to terminate the inputs with resistors (75 ohm?) to stop the traces ringing.

So the original video circuitry is still present bar 3 components, the original DIN socket, Q8 and R135 to match what is removed in the 800XL VBXL instructions (800XL- Q8 & R100 - minor differences between the two), and the video is sent directly to the new 13 pin DIN from the VBXL board.

BUT you're right in that I have removed the modulator, something I didn't need to do in the 800XL on account of the extra space in the 800XL case, so your theory might be spot on, especially since the VBXL is over the modulator inputs. I did place the shield between the tracks and the VBXL, and it did seem to improve a little which might be a clue.

I'll see if have some 75R resistors handy, I presume that I should connect the resistors to ground from the 2 modulator inputs?

I would guess that it's luma and chroma, so 75 ohms to ground would probably be a good idea. The resistance probably doesn't matter too much if it's in the few hundreds of ohm it should probably stop reflections.

stephen_usher wrote: 05 Apr 2022 12:09 I would guess that it's luma and chroma, so 75 ohms to ground would probably be a good idea. The resistance probably doesn't matter too much if it's in the few hundreds of ohm it should probably stop reflections.

I'll give it a go and let you know 😁 Thanks, much appreciate the suggestion.

stephen_usher wrote: 05 Apr 2022 12:09 I would guess that it's luma and chroma, so 75 ohms to ground would probably be a good idea. The resistance probably doesn't matter too much if it's in the few hundreds of ohm it should probably stop reflections.

So there are 3 connections to the modulator, I've added a 75R resistor from each connection to GND, and no difference. Will leave them in place for now.

Harumph!!

Hmm, now wondering if the new socket is where the interference is being picked up. It is right over the top of the old 5 pin DIN socket, as in the 800XL, but the socket was positioned 180 degrees rotated, so wiring on the left side not the right.

It was also a very long way away from the power socket and in the 600XL it is right next to it. I think I need to remove the socket (It's only held in with hot glue) and see if it helps.

It's a pity that the termination didn't help.

Given the frequencies of the interference the power is highly unlikely to be the source.

stephen_usher wrote: 05 Apr 2022 15:54 It's a pity that the termination didn't help.

Given the frequencies of the interference the power is highly unlikely to be the source.

I doubt it's the power too, but since removing the socket is trivial I may as well try.