exxos's Atari, Amiga & retro forum

stephen_usher wrote: 19 May 2021 22:38 Oh. Don't buy Gorilla Epoxy resin, it's rubbish.

Instead of going off and becoming rock hard and hard to get off surfaces it becomes a rubbery thing which just peels off. I'm going to have to get some proper stuff to re-glue the F1 case.

Araldite is my weapon of choice, Rapid in particular.

Yeah, I'll have to find a Halford's. All B&Q had was Gorilla everything.

Over lunch I managed to make a proper floppy converter cable and test it out.


Well, it worked as well as the iffy one on Sunday, i.e. with the Gotek it passes the BIOS diagnostics, detects that there's a disk in the drive but can't read the data. as you can see here:





The data's getting to the pin on the floppy disk controller so either something strange is happening or the FDC is faulty. A replacement FDC is in the post at the moment and should arrive any day now so when that arrives I can swap it out and test again.

Oh, that's interesting!

I've found a high density floppy drive which mostly works (I think the /DSKCHG is confusing the /READY signal) and when I put a HD DOS floppy in there it gives error 99, which means non-bootable disk. If I dd(1) a 720K F1 bootable floppy image onto the disk it tries to boot but fails, of course, because it has the wrong number of sectors.

OK. Doubling the data rate in FlashFloppy on the Gotek didn't help.

P.S. I have FF config issues. I'll have to look later.

OK. I'm trying to configure FlashFloppy with a custom image format of 80 tracks, 9 sectors, two heads but with double the data rate, i.e. 500 rather than 250 and FF just isn't seeing the images. Gah!

OK, with the help of Keir Fraser I got the setting correct.

The FDC is indeed stuck in high density mode somehow. Modifying the data rate on the Gotek has allowed me to boot a virtual floppy!


The problem I have now is that the schematics I have don't actually show any circuitry to control the data rate as far as I can tell. I'll have to investigate further.

The specs I have and the motherboard differ greatly around the disk density switching circuit. In the schematic chip at location 3F generates the DENSITY line (Pin 2). However, on the revision I have this chip doesn't exist and pin 2 is connected directly to ground.

It seems somthing is also pulling Pin 17 of the WD2797A floppy controller high. This is the 8"/5.25" floppy compatibility switch. It seems that 8" disks ran at double the data rate of 5.25" disks, so this is what's causing the double data rate. However, the schematics don't show pin 17 being connected to anything at all.

This part of the motherboard seems very different to the Rev.C board described in the schematic.

I may try just lifting PIN 17 and seeing what happens.

I lifted pin 17 and nothing changed. It seems that it isn't actually connected. I could short it to see what happens though. Anyway, the second FDC has arrived.

Later I should have some more components arrive, enough to bring out the internal RGBI-C video to a ZX Spectrum+ 128K compatible socket so I can use the machine with a SCART monitor. The Spectrum+ 128K lead handles the intensity line so I should get the full 16 colour palette. I'll use this in preference to the standard Apricot monitor out port as that has only H-Sync/V-Sync rather than composite sync on the internal header.

By the way, this is how the Apricot F2 I got from Facebook Marketplace was packed. This is after being tidied up and the F2 put back together (mostly) but no packing materials have been removed.

That packing is shocking!!