Background interference/patterning with VBXE

[morelenmir]

Since I first fitted a VBXE to my 800XL I have noticed very marked, running bands of diagonal interference or patterning against the background. It is most noticeable when the background is a medium blue. I think it is likely there all the time, but only noticeable when using the ultra-sharp RGB-SCART out. The interference gets much, much worse when I connect the 800XL to my PC via an 'SIO2PC-USB' cable.

 

I have attempted to solve it, off and on for over five months. Firstly I ensured the 'SYNC' signal came directly from Pin15 of the 4050 chip. Next I removed all the stock single-wipe sockets and replaced them with precision-engineered units. I then remade the connection from the J4 connector on the VBXE to the DIN13 socket - paying especially careful attention the 'G' line was securely attached with good connections on both ends. Over the past few weeks I have also replaced every single electrolytic capacitor on the motherboard, suspecting one had partially failed and introduced the interference. Finally, I reworked an old repair to two tracks than run from the POKEY to the PIA chip - previously the broken track had been glued down slightly out of place and touched a pad that it ran close to. It was still insulated, but when testing continuity there was a definite incapacitate response so I thought the errant signal could have been getting in from there.

 

Sadly all of these attempted fixes have done nothing to cure the problem. I know the Atari motherboard can be extremely noisy, but I wonder if anyone can suggest something else I can rework or replace in order to cure this patterning.

[flashjazzcat]

Candle told me this was a grounding issue (ground loop, IIRC), and in my case it was because I'd naively tied the audio L/R GND to the RGB GND. That's all I can suggest looking at, anyway: grounding.

[morelenmir]

Candle told me this was a grounding issue (ground loop, IIRC), and in my case it was because I'd naively tied the audio L/R GND to the RGB GND. That's all I can suggest looking at, anyway: grounding.

 

The audio is a good place to look FJC... Stupidly that is one thing I have not tried out and gone all around the houses instead. I guess just unplug the Lotharek-Stereo board and see if it goes away. That is the first thing I should have tried!!!

 

I will report back.

 

It is also worth mentioning that I have also tried a completely different power-supply. Other than producing a slightly more compact, narrower interference pattern it did nothing to help!

[morelenmir]

Damnit - I thought that might be it, especially as there is not an actual audio-GND line inside the DIN13->SCART cable I bought. However, even when the Lotharek-Stereo board is totally disconnected and a single POKEY set in its place the interference is still there.

 

Is it possible that by removing the metal RF shielding there is a GND track that was connected around the motherboard through it and now is not?

[Kyle22]

 

...

 

It is also worth mentioning that I have also tried a completely different power-supply. Other than producing a slightly more compact, narrower interference pattern it did nothing to help!

 

Are you saying that the interference pattern changed (got narrower) when you changed power supplies? That would suggest to me that power supply noise may be involved. Do you have any way to power it from a 12v battery connected to a 5v regulator (for testing)? If the pattern goes away, then you have 2 noisy power supplies.

 

Is your power supply connected to the same earth ground (from the mains) as your monitor (such as by using a power strip) or are they plugged in to different outlets?

 

Just some ideas...

[Rybags]

What revision board is it?

 

Mine is practically perfect. Can't remember though if I still have the bottom part of the shielding on.

You could try making up some wiring to provide ground connection between the screw holes in lieu of the shielding but I don't think it'd make much difference.

 

If you have diagonal fairly stable line through the display, then it's probably something close to line frequency, ie ~ 16 KHz.

 

Another thing to try might be a fresh 4050. With my install I soldered a header pin for the pickoff point which is on the board, not directly from that chip.

Edited December 16, 2015 by Rybags

[morelenmir]

 

Are you saying that the interference pattern changed (got narrower) when you changed power supplies? That would suggest to me that power supply noise may be involved. Do you have any way to power it from a 12v battery connected to a 5v regulator (for testing)? If the pattern goes away, then you have 2 noisy power supplies.

 

Is your power supply connected to the same earth ground (from the mains) as your monitor (such as by using a power strip) or are they plugged in to different outlets?

 

Just some ideas...

 

That is a good point - it was just a purely subjective opinion, but it did look like the diagonal 'bars' were closer together when I tested the replacement power unit. I have an old ATX power supply I could give a try with, although I like the idea of the battery as well - just to totally eliminate that source. Both monitor and XL are powered from the same strip... Would some leakage to ground from the mains cause this issue?

 

What revision board is it?

 

Mine is practically perfect. Can't remember though if I still have the bottom part of the shielding on.

You could try making up some wiring to provide ground connection between the screw holes in lieu of the shielding but I don't think it'd make much difference.

 

If you have diagonal fairly stable line through the display, then it's probably something close to line frequency, ie ~ 16 KHz.

 

Another thing to try might be a fresh 4050. With my install I soldered a header pin for the pickoff point which is on the board, not directly from that chip.

 

It is a 'REVC' board - that has been worked on in the past by its previous owner. All the DRAM chips have a bodge resistor soldered down across a broken track to emulate the layout of the later 'REVD' board I believe.

 

I should have mentioned that I replaced the 4050 as well. The SYNC lead is not soldered directly to the pin, but to the track that leads immediately from it to a resistor.

 

It is interesting you should mention the movement of the patterning - it does move, but fairly slowly from bottom right to top left.

[morelenmir]

Is it possible that this patterning could be caused by some sort of conflict between the 'Monitor' output and the 'RGB SCART' from the VBXE?

 

Right from when I first installed the VBXE I kept the 'Monitor' output also attached to my TV in order to test the screen. I have never disconnected it and never even thought about doing so. However, earlier I was trying to - unsuccessfully - resurrect my 'REVD' motherboard form its current 'red screen' condition and because that board does not have an VBXE I plugged it in to the 'Monitor' lead to test it. As I say I did not manage to repair the board. However, when I finally gave in, for whatever reason I did not reconnect the 'Monitor' lead to my working setup.

 

And... When I switched it on the patterning was magically absent! However when I did reconnect the lead the patterning returned... Wow... If this is going to be what has been behind the problem all along I am going to me really furious with myself!!!

 

If it is then I would guess it is more a shortcoming of the TV than the Atari... But... I don't know. Is this from an electronics standpoint possibly the cause?