General Stuff(tm)

[stephen_usher]

Seeing as I'm working from home and it's pretty quiet at work I swapped the last of the caps on the low voltage side of the Apricot FP PSU.

The originals had leaked slightly but they were obviously working well enough because after fitting the new ones there was no change in the performance.

I did look at the voltage balancing resistors but the arrangement was slightly different from the schematic for the 2A version of the PSU so I left it alone. It looked as though Apricot themselves had replace the resistors themselves, so maybe they wanted 4.7V?

[rubber_jonnie]

Seeing as I'm working from home and it's pretty quiet at work I swapped the last of the caps on the low voltage side of the Apricot FP PSU.

The originals had leaked slightly but they were obviously working well enough because after fitting the new ones there was no change in the performance.

I did look at the voltage balancing resistors but the arrangement was slightly different from the schematic for the 2A version of the PSU so I left it alone. It looked as though Apricot themselves had replace the resistors themselves, so maybe they wanted 4.7V?

Is that similar to the BBC PSU where a set of resistors are factory fitted to determine the output?

[stephen_usher]

Yes. The same company designed the two PSUs so they have a great deal in common.

[rubber_jonnie]

Yes. The same company designed the two PSUs so they have a great deal in common.

That makes sense then.

[stephen_usher]

Lashed up an Apricot FP to VGA cable to see what the colour monitor output was like.

Pretty pointless actually.

Although you can switch to it during boot and use it in MSDOS mode when programs start they generally either crash or they switch back to the LCD screen.

MSDOS text mode is OK, though the display is buggy in that it echos the first pixel of a new line on the end of the line above.

[stephen_usher]

I haz new memory controller! Let's see this evening if it fixes the memory issues.

[rubber_jonnie]

I haz new memory controller! Let's see this evening if it fixes the memory issues.

Oooh, interesting, looking forward to hearing the results.

[stephen_usher]

Well, it's not the memory controller. :-/

It's not the CPU either as I've tried swapping that. There's not a lot else on between the two on the address lines other than the three 373 latches. But why is it affecting only some addresses above 256K in the lower three bits? That's that I can't figure out. The lower three bits are handled by a separate 373 to the upper ones so how can one affect another?

Here are the two ends of the bus:

FP-CPU-Address-Bus.jpg

FP-Memory-Address-Bus.jpg

[stephen_usher]

I s'pose the quickest thing to do will be to swap out the high address 74LS373 (IC25) and if nothing changes then change the lowest three 41256 chips on the low bank of the memory. If I've got to extract the memory I might as well remove it all and socket the whole lot, plus the empty 16 positions. This will allow me to upgrade to the full 1MB.

P.S. Or I could swap out IC40 (74LS244) as I guess that it could be marginal and 244s are prone to failure.

[stephen_usher]

Freeze spray is a good tool.

Freeze the 373s, no change.
Freeze the 244s, no change.
Freeze the RAM chips... Loads of errors.

Once things have warmed up again I'll try freezing individual chips.