General Stuff(tm)

[stephen_usher]

I'm increasingly convinced that the problem is with the floppy drive. Specifically the microcontroller chip on the circuits board. This is the same major item which failed on the F1's drive (other than the motor board) and on this one looks like it's a refurbished replacement (lots of flux residue around the chip). So, maybe it's a common failure for these early drives, even when they were new.

[rubber_jonnie]

I'm increasingly convinced that the problem is with the floppy drive. Specifically the microcontroller chip on the circuits board. This is the same major item which failed on the F1's drive (other than the motor board) and on this one looks like it's a refurbished replacement (lots of flux residue around the chip). So, maybe it's a common failure for these early drives, even when they were new.

Is it feasible to test the floppy on the F1, or the F1's floppy on the portable?

[stephen_usher]

The F1's floppy is a dead-un. If I use freeze-spray on the controller chip continuously it will initialise and be seen by the machine. The motor board is completely dead.

The FP's floppy mostly works but if started from cold will half the time work (read data) and half the time not. If I allow it to warm up a bit and then power cycle then it works correctly. If it's in a non-working boot state then it needs the drive power cycling for it to start working again.

I verified that it's not the controller as the external Gotek can be read when the original floppy isn't working correctly.

[rubber_jonnie]

The F1's floppy is a dead-un. If I use freeze-spray on the controller chip continuously it will initialise and be seen by the machine. The motor board is completely dead.

The FP's floppy mostly works but if started from cold will half the time work (read data) and half the time not. If I allow it to warm up a bit and then power cycle then it works correctly. If it's in a non-working boot state then it needs the drive power cycling for it to start working again.

I verified that it's not the controller as the external Gotek can be read when the original floppy isn't working correctly.

What a pain!

[stephen_usher]

Last night I managed to get the Apricot Field Test memory test to work at last on the FP. The strange thing is that if BUFFERS= is set in CONFIG.SYS then the memory test fails at around 230K and the machine crashes. If I remove the line from CONFIG.SYS then the memory test passes. This is not the case for the the other F Series machines and it's weird that MSDOS would put the buffers in the middle of program space. Oh well.

Anyway, it looks like the hardware is OK.

[rubber_jonnie]

Last night I managed to get the Apricot Field Test memory test to work at last on the FP. The strange thing is that if BUFFERS= is set in CONFIG.SYS then the memory test fails at around 230K and the machine crashes. If I remove the line from CONFIG.SYS then the memory test passes. This is not the case for the the other F Series machines and it's weird that MSDOS would put the buffers in the middle of program space. Oh well.

Anyway, it looks like the hardware is OK.

All this fruit based talk is making me hungry 😁

Good news at least regarding the hardware though.

[stephen_usher]

And so, it's back to setting things up ready to extract data from old DAT tapes.

Initially I set up the SPARCstation IPC but then thought that it may actually be better with a newer, faster machine with a whole 100Mb ethernet connection. So, this evening I brought the SS20 back from storage.

Well, it wouldn't boot properly with the internal SCSI CDROM drive installed. *sigh* So opened the machine up and part of the plastic on the outside of the case just fell apart. It's become that brittle that the slightest touch is causing it to break now. Unfortunately, this seems to be normal for that era of machine. It was a sort of PVC-like plastic that's losing all its pliability and strength.

Anyway, the machine is up and running now and I've recompiled the 'vmsbackup' software for Solaris 9. I'll now have to await the arrival of the other tapes tomorrow (probably).

Oh, and whilst I was doing that I decided to try a last-ditched effort to get the custom chip in the Sun 3/80 working by baking it. Some chips come alive again if "annealed" in an oven, so 200C for 2 hours. Once it's cooled down I'll see if it's killed or cured it. It was mostly dead anyway.

[stephen_usher]

After the baking process one of the signals came back, but two didn't. The machine's dead.

[rubber_jonnie]

After the baking process one of the signals came back, but two didn't. The machine's dead.

Interesting, I have watched a couple of videos on this process and nearly tried it with the VIC chip from my VIC-20 as it was marginal, but just didn't get there in the end. I could still try it as it goes, but it's not high on the to do list.

Shame it didn't work for you though.

[stephen_usher]

Today's fun is installing a TMS2RGB board to my original Memotech MTX500, which I had to fix three years ago after being dead since 1988.

So, the first task was to clean up the bodge wires on the bottom of the board, which was a mod to allow the use of normal EEPROMs rather than the original mask ROMs. After that it was off with the modulator! (This is part of a cunning plan... to be continued...)

IMG_2808.jpg

The TMS2RGB board sits on the underneath of the TMS99xx VDP socket. Unfortunately, whoever fitted the socket at Memotech wanted the pins to be as flush with the PCB as possible and had cut them down, meaning fitting the new board would be tricky...

IMG_2810.jpg

IMG_2809.jpg

To connect the TMS2RGB board to the rest of the world it needs a ribbon cable soldered onto the board and taken elsewhere where a plug needs to be fitted so that an external cable can be hooked up. I got this micro-8 way DIN socket and cable. All I needed to do after soldering the board on is fit to the ribbon cable and fix the socket in an appropriate place.

Of course, with the TMS2RGB being on the bottom of the motherboard, routing the cable might be a problem.

Anyway. After refreshing the socket pin solder, sucking up the excess so that the pins stood proud of the PCB at least a little bit I managed to get the board on. (That was a bit of a struggle.)

IMG_2812.jpg

At this point I hooked up the machine to power and video and turned on... no smoke... Blue screen with "Ready" and a flashing cursor. The machine still worked! :-)

Back to the fabrication. Solder the wires onto the socket and then route them around to the board and solder them on. Then attach the socket to the top of the board with superglue and solder. (I would have preferred just solder but it wasn't strong enough.)

IMG_2813.jpg

IMG_2814.jpg

Can you guess what my cunning plan is yet?...

Oh yes, I'm using the opening which was used for the modulator for the new RGB socket! :-)

IMG_2815.jpg

IMG_2816.jpg

After putting it all back together the modification to the Memotech is complete. Now I have to make the other end of the cable before I can properly test it. However, signals are coming out of the wires at the end of the cable, which is encouraging.

IMG_2817.jpg