exxos's Atari, Amiga & retro forum

stephen_usher wrote: 14 Sep 2022 20:13 My goodness! I'd forgotten just how slow tape drives are!

Indeed. I'm currently using one as a host for a SCSI CD-ROM drive. (Must find a SCSI CD caddy at some point). :)

I finally got around to working upon the Apricot FP and its memory problem.

I've desoldered all the RAM and it all tests fine in the tester. This suggests an address line issue further up the tree, but I can't verify this without replacing all the RAM and seeing if the issue goes away or not.


So, this afternoon I'll clean up the board, fit sockets and then different RAM chips. I can then test.

Because this is a multiplexed address bus the upper and lower addresses are held in buffer chips, so those would be the next suspects.

Unfortunately, due to the design I can't test the system "live" as I can't fit the display if the board is on the bench.

P.S. It's not buffers on the FP, it's a TMS4500 memory controller. :-/
P.P.S. Actually some 373s feed into the TMS4500.

Will keep my fingers crossed for a good result :crossed1:

It's not the RAM. A complete memory transplant and the machine behaves the same way. So, it's either the TMS4500 memory controller chip or one of the 74LS373 which latch the address on the bus.

It's a very strange address at which the machine crashes, 0x3A000. I'm guessing something's looping back to zero and its overwriting the interrupt table.

I'll see if I have any 373s in stock. If I do then I'll swap IC25 which latches the top bits of the address.

Swapped the top 373 and no change, and the old chip tests OK in the EPROM programmer too. It must be the TMS4500. Boo!

At least it's socketed.

I'm now wondering if it's a power supply issue.

The PSU is only outputting ~4.8V and by the time it reaches the RAM area it's down to 4.6V. The DATA-IN on the RAM is only getting up to 3.5-3.6V for a high, which is good enough for TTL but it's a bit iffy.

I've replaced the capacitors in the output area of the PSU from what I remember. Being a "cheap" Aztec supply it doesn't have adjustment pots only fixed resistors for the output regulation.

Anyway, I don't have time to look into it this week as there are final preparations for the show next weekend.

Well, that's the Cyber Legends set-up for the Memotech verified:

• Memotech MTX512 with MFX board fitted.
• VGA cable.
• Dell 17" VGA LCD Monitor with Sound Bar and "kettle" lead.
• Phono cable with adapter for stereo jack.
• Zip Stick joystick.

OK, back to the Apricot FP...

When lightly loaded (just the floppy) the +5V is at 5.01V. With one of the boards in there it's ~4.9V. With both boards it's 4.64V. But, if I remove some load by disconnecting the floppy drive the voltage doesn't go up it goes down, to 4.4V!

Anyway, I ran my memory tests again and whatever I've done has changed the failure address from 3:2A00 to 3:3900. I've just moved the memory into sockets (same chips), replaced the 74LS373 driving the upper address lines and control lines and reseated the CPU and memory controller chips.

I'm sure the low power must be having some effect.

stephen_usher wrote: 18 Sep 2022 14:53 Well, that's the Cyber Legends set-up for the Memotech verified:
...

I dont think I can ever see too many pictures of that beauty. Cannot wait to see it for real. :)

The TMS4500A-15 arrived today. I may have time this evening to swap the chips to see if it makes any difference.