Video Jail Bars: Is There a Root Cause?

[Ricky Spanish]

20 minutes ago, ACML said:

I believe the jailbar phenomenon started when we started to use LCD monitors.  CRT tubes don't seem to have the issue, hence why you are not seeing it on your CRT.  For some reason modern flat screen (non CRT) monitors are needed to see the jailbars.

 

Right, so, it seems memory can clean up some artifacts on old CRT's, but probably not on non-CRT tv's. Interesting.

[+mytek]

14 hours ago, Ricky Spanish said:

Right, so, it seems memory can clean up some artifacts on old CRT's, but probably not on non-CRT tv's. Interesting.

Yes most people would agree that CRTs were more or less immune to the video noise issues that became apparent with the use of LCD TV/monitors. Although I'm not so sure that memory really played much of a role in cleaning up video when a CRT was specifically being used.

 

Memory in itself, or should I say different brands of DRAM, won't change any jail bars (e.g., jail bars that are specifically caused by memory circuits) that may be originating from where they are used in a given circuit. This would be the case starting with the XL series on through the XEs. So I would suspect that changing to different brands of DRAM in those systems would likely make zero difference. However the 400/800 series used plug-in memory cards, and where those got produced by 3rd parties there would most certainly be a change in the way the PCB circuit paths were laid out, the power decoupling and filtering being used, and perhaps the proximity to video circuits - although that last part is a stretch. And in recent years there have been a couple of newer entries that use SRAM instead of DRAM which would certainly lower the noise threshold, especially where it comes to the lack of memory refresh being needed.

[Ricky Spanish]

isn't DRAM power constantly switched on/off,  whereas SRAM requires constant flow of power ? Could the constant on/off cause more interference ?

[Beeblebrox]

Just tacking on the end of this excellent thread. 

 

Have an XE (130xe) which has been upgraded to 320k using this method, which I've used many times in the past:

 

https://www.exxosforum.co.uk/forum/viewtopic.php?t=732

 

Also have a ugv in there hooked up to the din5 jack 4050 removed and luma, chroma and cvbs ferrites lifted. 

 

I am still getting wide jail bar interference on my lg flatron lcd. 

 

Almost identical to this image I grabbed off the net, but mine is more noticeable:

 

 

Incidentally the 3 x ground wires the ugv has I have tried grounding in different places.. Makes no difference. 

 

This machine also has the following Simple XE tweaks:

 

 

Even before I did the S. V. XE tweaks, or added in the ugv and ram upgrade, it had very bad wide jail bars.

 

Adding back the 4050 makes no difference. Also, adding back the XE sheilding plates makes no difference. 

 

I am given to understand these jailbars can be attributed to ram noise(?) and or power supply related, and to a greater extent in some a8s, especially the XE line, grounding has a big part in it. (The XE line doesn't have great grounding planes afaik). 

 

I was hoping that, in bypassing the luma/chroma/cvbs circuit, and taking the video from the ugv, it would help eleminate or reduce these considerably. 

 

FYI I am running off composite and a USB 5v 2amp power source. I use the same for my 600xl.amd 800xl with very good pictures. 

 

I will try powering this XE with a brick mains Atari psu to see if it makes a difference. 

 

 

 

 

Edited April 15 by Beeblebrox

[Ricky Spanish]

Since getting a LCD with s-video, there's been a vast improvement in the jail bars over HDMI and LCD.

[Beeblebrox]

11 minutes ago, Ricky Spanish said:

Since getting a LCD with s-video, there's been a vast improvement in the jail bars over HDMI and LCD.

Thanks. Yes, I agree that svideo is my default and the best, (it's shat I run my a8s off), but this machine is going to another owner who'll be using av composite as svideo isn't an option, so I wanted to make sure the av composite signal was optimised and these bars reduced as much as possible.

[Beeblebrox]

Managed to take some pics, follow my earlier post. Damn. These are bad. Again on composite. 

 

BTW this was with a brick Atari psu. Makes no difference if on this or usb power. So it's likely a ground issue and/or ram noise presumably?  Not sure what I can do but I need to have an acceptable picture on composite. 

 

 

[_The Doctor__]

Check if your machine has proper Freddie decoupling cap there were routing errors for connecting on some PCB's

adding some of the components back for the crystal clock even with an oscillator present can help also.

Edited April 16 by _The Doctor__

[+Stephen]

1 hour ago, Beeblebrox said:

Managed to take some pics, follow my earlier post. Damn. These are bad. Again on composite. 

 

BTW this was with a brick Atari psu. Makes no difference if on this or usb power. So it's likely a ground issue and/or ram noise presumably?  Not sure what I can do but I need to have an acceptable picture on composite. 

 

 

Sorry if this has already been discussed, but are you sure you have an adequately shielded monitor cable?

[Beeblebrox]

4 hours ago, Stephen said:

Sorry if this has already been discussed, but are you sure you have an adequately shielded monitor cable?

Thanks for the question. I use the same av composite cable for my other stock xe's and they are more or less fine. Nowhere near as bad as this XE. These lines are coming from this machine. I know XE video out quality isn't the best already, but this is really bad. 

[Beeblebrox]

5 hours ago, _The Doctor__ said:

Check if your machine has proper Freddie decoupling cap there were routing errors for connecting on some PCB's

adding some of the components back for the crystal clock even with an oscillator present can help also.

Thanks, will check Freddie.