Forgottenmyname - Progress

[exxos]

I did a quick simulation based on the 3904 & 337..

Basically the Atari 3904 outputs a tiny higher voltage, but not by much.. but could be enough to cause issues...

2.335V vs 2.397V.

I guess you could swap the BC337 out for the 2N3904 from a the donor ST board.. see if that helps...

s1.jpg (77.52 KiB) Viewed 3824 times

[PhilC]

I might make a new cable with hsync and hsync. Would everything else be the same minus the comp sync obviously

[troed]

I might make a new cable with hsync and hsync. Would everything else be the same minus the comp sync obviously

Measure them at the monitor port first But yeah, if they're ok there then you should have a stable image and maybe csync for some reason is off.

[exxos]

I might make a new cable with hsync and hsync. Would everything else be the same minus the comp sync obviously

You should get near 5V from the GLUE on the sync then.. composite voltage might be a fraction to low.. I've seen it a lot before with sync faults on ST's...

[exxos]

It makes me wonder to replace the composite transistor with a AND gate instead.. then also buffer the V&H sync though the same IC.. would need 3 gates and they normally have 4 . But would help protect and buffer the syncs and up the voltage on them... assuming thats the problem..

The syncs have diodes on them, assume for protection from external stuff blowing up the GLUE etc, but possible the diodes are not even needed if a external buffer is used.

[PhilC]

replaced the 14 pin plug on my scart lead and the tv syncs just fine now with the composite sync, which means another thing ticked off for us.

Still getting a yellow screen though but no longer flickering.

Just the rising voltage problem on the data lines etc now

[exxos]

Just the rising voltage problem on the data lines etc now

I'm not really following what you mean there... voltage on the buses should rise up to 5V unless the CPU sets it to 0V... voltage could be anything in between cycles, can appear like a 2V triangle etc.

[PhilC]

Ok, maybe not a good description,

Loookng at it on 2us range on the scope on the CPU address lines, I get a good 10 seconds of pulses, around 3v and then it goes logic high to 5v and then won't do anything unless i do a cold start.

fyi /BG starts high, then pulses then goes low after 5 seconds
/BGACK stays high
/BR stays low
/BERR is high with a very quick pulse of low every 2 seconds then is high after address bus activity ceased.

[exxos]

I think BR is probably confusing the GLUE.. got 2 fixes for that now though

[PhilC]

Ok, fixes done.

1 step forward is that i have a stable screen, although its yellow and im back to the 2mhz pulses on /AS and the data lines