MegaSTE Repair Continued - Video Noise

[displaced]

Hi!

Thought I'd start another thread specifically for diagnosing the video interference my newly-working MSTE exhibits.

I've had a chance to try out a few floppies - the drive seems fine (as in the disks that don't work on my other STs also don't work on this, and those that do, do). The only game I've got handy that likes the STE is "I, Ball" - reasonable game, but great music!

Having the machine run under a bit of load has brought up an issue that's not really visible when pootling around on the desktop. There's quite severe video interference, particularly visible on plain, dark backgrounds (such as in I, Ball).

I've read quite a few threads, here and elsewhere about video interference on various STs, so I've got a few things to try. But I noticed something a little interesting that someone with more experience might see as a pointer to the underlying cause.

As you can see in the photo (and if I manage to attach it, the video), there are lots of vertical, flickering lines. But what's odd is that there's a distinct change in the interference pattern between what would be the border area and the actual display area. It's not too clear in the photo, but in-person, the interference pattern shows a clear 'edge' where the desktop's borders would start.

Unsure if that would let me rule out, for example, the power supply as a cause or what, but thought it worth asking here in case you'd seen similar and found the cause!

The photo:

IMG_5745.jpg (166.02 KiB) Viewed 3658 times

...and the video:

MLJHYrw.gif (64.15 MiB) Viewed 1297 times

The machine's connected with an ubeswitch-style adapter to an OSSC with a pre-made, known-good VGA cable. I use the very same adapter, OSSC and cable with my STFM without problem.

All thoughts welcome!

[exxos]

Very odd. Can't say I have a mste or ossc, but the ghosting fix can be done on the mste which may help. Though odd things like that, I would probably just try a 47pf cap on all the RGB port pins to gnd and see if you can hit on a signal which makes the video better. Could be noise on HSYNC for example, hard to say.

[displaced]

Great - I’ll give that a go!

As an aside, do you have an opinion on a good set of reasonable quality electrolytic caps? I’ve been buying them ‘on spec’ whenever I’ve had a specific use for them, but it’d be handy to have a box of decent caps of useful specs.

Thanks!

[exxos]

Panasonic FR series are good. But the range isn't great. Its a matter of wading through all the data sheets until you find the best ones.

[displaced]

Thank you!

I found a 47pf amongst some caps I'd ordered when repairing my Archie. Unfortunately it made no difference to the signal when placed across C506 - absolutely no difference in the output at all.

I saw a post on atari-forum or atariage (which I can't blimmin' well find now) where these sorts of regular patterns were traced back to a noisy PSU. I'd lazily just checked the voltages with a multimeter, which were perfect -- but I've just watched the 5V on my oscilloscope.

Now, until a few weeks ago, I'd last used a scope back in 1995 and can just about remember whatever I'd needed to learn to pass GCSE Electronics. So, I've taken a couple of grabs from the scope which I hope you wouldn't mind helping me interpret.

First I calibrated both the scope and probe, then clipped to a ground and poked the each of the +5 lines on the PSU plug with the pointy bit (they all look the same).

Now, on the face of it, these look quite noisy to me, but I know that if you look close enough, everything looks noisy and I might be on the wrong track. What do you reckon based on these?

bmp_23_007.jpg (75.05 KiB) Viewed 3562 times

bmp_23_027.jpg (78.45 KiB) Viewed 3562 times

The entire waveform bounces around a little bit as well - shame this scope doesn't save video!

...and here's another shot of the interference which is very visible on the Frontier intro - shows the regularity and 'sparkliness' of it quite well.

IMG_5759.jpg (142.75 KiB) Viewed 3562 times

[exxos]

47pf on RGB lines... You need 47uf or something like that for the cap as stated on my site..

PSUs can cause noise, it doesn't take much .. But mste PSUs are generally good. I would also suggest you not use parity Simms and clean all sockets and ICS with ipa wipes.

[displaced]

Ah - thank you - sorry, I'd meant to say I'd give the RGB lines a try with caps at the weekend. And the uF/pF mixup is entirely because I'm stupidly doing this stuff late at night

I'll follow your info properly before posting another load of nonsense!

Thanks!

[exxos]

See here also. https://www.exxosforum.co.uk/atari/last/ ... tm#PSM5341

[displaced]

Thank you!

I added the proper caps to RGB lines without change. So then proceeded to repeat the removal, cleaning and re-seating of everything. Luckily the four 1MB SIMMs in my STFM’s Marpet upgrade are non-parity. Swapped those in, but no improvement.

Finally I hit on something that at least changed the pattern of interference. Removing and replacing the shifter changes the severity of the problem, but never actually clears it. At one point it was very nearly clear, but instead there was a severe pattern of flickering dash-shaped blocks at the extreme left of the display. Of course, I didn’t take a photo, and it hasn’t happened again!

I’ve cleaned the chip and socket thoroughly, and have adjusted all the socket pins to make good contact. Can’t see anything physically wrong.

Short of any ideas, I think I’ll live with this for the time being. Got a saved-search on eBay for a replacement C301712-001 in case one pops up. Even if it doesn’t fix it it’s nice to have a spare!

[exxos]

Have you tried pushing down on the IC ? Maybe the socket is badly soldered ?