View Full Version : Non-standard Citizen 26-pin floppy drives

February 10th, 2017, 03:39 PM
I am the proud owner of a new Toshiba T3100SX, which is almost (but not quite) not working.

One of the things which is almost but not quite not working is the floppy drive, which is behaving as if one head is broken: if I format a disk with it, every sector with head 0 works, while every sector with head 1 doesn't.

First question: does this ring a bell? Does it sound fixable?

The drive itself is a Citizen unit with a 26-pin connector. This appears to be ancient and obsolete but relatively standard. The maintenance manual has the pinout, which matches this:


Second question: if I just wire it up to a modern 3.5" drive, connecting the obvious signal lines up, is it likely to work? I assume I need to either ground DRVSA and raise DRVSB on the modern drive, or possibly the other way round; I can never keep inverted signals straight in my head. So, has anyone done this who might have a wiring diagram I can copy so that I can avoid having to think about it?

February 11th, 2017, 10:02 AM
I did something similar to my T1600 when the floppy drive died. Soldered up a cable to use a "normal" floppy drive, and it worked. But unfortunately that was over 10 years ago so I don't remember any details.

February 11th, 2017, 01:24 PM
Well, it's good to know that it's at least possible.

I cleaned the disk heads with cotton buds dipped in isopropyl alcohol --- the smallest bottle I could buy was 100mL; I have quite a lot left --- and it didn't seem to make much difference. I *did* manage to find a sector editor, and bullied it onto the machine.

The results are quite interesting; head 0 is absolutely fine. For head 1, the first 15 tracks or so are fine, which is how I'm getting files on and off it, but beyond there the machine has progressively more and more trouble reading sectors --- first requiring several retries, but then by about track 40 there are permanent CRC errors.

But it won't *format* head 1 at any cost. It's an instant failure, even on track 0.

This doesn't sound like an alignment issue, because if it were then the machine would be able to format and read and write its own disks. It's not the stepper motor being out of alignment, because if it were head 0 wouldn't work either. It's not BIOS setup, because then nothing would work (also the floppy drive setup is non-configurable on this machine). It's not an electrical problem with the head amps or something because then it wouldn't work on the low tracks but not work on the high tracks...

...so what *is* it?

It's so nearly working that I don't really want to try a dodgy rewiring of my only 26-pin ribbon cable until I know for sure that the drive's irreparable!

February 11th, 2017, 03:29 PM
For reference, I found a couple of threads from other people doing exactly this; it comes with a handy cross-reference chart, although they never reported back whether it worked or not...


The 26-pin layout appears to be a standard, although usually showing up on flatflex rather than IDC like I have (although belatedly I realise that the defunct T1200 floppy drive I have with the flatflex connector is going to be the same standard). There are ludicrously overpriced adaptors:


...and emulators, which I'm sure would be ludicrously overpowered if I could find any pricing information:


(although the idea of a floppy drive emulator with built in ethernet does seem a little scary!)

February 13th, 2017, 12:34 AM
...and we have a winner!

Turns out that there are a bunch of old synthesisers which use 26-pin floppies to load samples, so there are lots of people who are really attached to their hardware who are desperate to use modern drives. You can get adapters for about 20 currency units off ebay, or make your own:


So I bodged up a cable based on the spec from the second link above above, and it almost, but not quite worked --- the trick is that my Toshiba T3100SX actually uses the READY line in the floppy connector, while the Yamaha doesn't; but modern floppy drives don't actually *produce* a READY line so there was nothing to hook it up to. Luckily I found that my drive has a jumper setting that makes it generate READY on pin 34 instead of DSKCHNG. Setting this made it Just Work. A modern beige floppy drive looks a bit startling in the dark grey T3100SX case but it fits fine and works fine.

Also, a handy tip: rewiring connectors like this is easy; I prised the 26-pin IDC connector off the end of the old ribbon cable, split up the strands of one end of the modern 34-way ribbon cable, and pushed each individual wire onto the connector. Looks a bit tacky and you need to trim any unused strands so that they don't brush against the PCB and short out, but it can be done entirely by hand, doesn't need soldering, and is completely reversible.

Hopefully this is useful to anyone else in the same position --- these 26-pin drives are getting increasingly rare.

May 19th, 2017, 05:49 PM
I have the same issue as OP.

Would prefer to create custom cable, should have all of the parts on hand.

I do need more specific understanding of wiring power from original cable, as clean as possible, given the VCC5V from three pins, and all of the grounds. If OP has an image of their custom cable I would like to see.

May 20th, 2017, 11:08 AM
I don't have it handy right now, so I can't give you a photo, but there's not much to see --- it's just a 34-pin ribbon cable with a 26-pin plug on one end and a mess of randomly rearranged wires.

Regarding power: I didn't bother trying to extract power from the 26-pin end (at the laptop). I had the make a custom PSU board for this machine anyway, so the drive's powered directly from there via a standard molex cable.

May 20th, 2017, 03:06 PM
Thanks for response. Given the small space inside I may go with converter board. But may build reverse cable to test drive in other system. I suppose since T5200 has expansion slots I could try using floppy controller from multi I/O board as well.

October 17th, 2017, 06:53 AM
An addendum to the above: reading up on floppy disk drives I found this article which describes the floppy disk interface in copious detail:


It claims:

Not all drives provide a signal when they are ready and not all controllers make use of it. This signal
is active low, so if your controller needs it and if the drive does not provide it, then you can short the
signal with the adjacent GND pin. It is said that this may cause a few retries, whereas it works
perfectly fine for others. But if your drive provides a ready signal, then use it. Note: The Amstrad
PCW models require this signal. Further, they require its absence for recognizing that there is no
disk in the drive. An artificial RDY signal must be provided if the drive does not offer one, which
modern 3.5" drives donít, although other ones do.

Therefore I may have been able to get away with grounding READY and using an ordinary drive. Hopefully that's useful for anyone else reading this.

October 17th, 2017, 08:20 AM
... I may have been able to get away with grounding READY and using an ordinary drive. Hopefully that's useful for anyone else reading this.
Many 3.5" drives do generate a /READY signal but don't provide a switch or pluggable jumper to select it; as a matter of fact I just happen to have a couple of 3.5HD drives beside me and they both have a soldered DC/RY jumper. Might be worth while to trace pin 34 back to see if there is a jumper somewhere along the way, and if so run a jumper from /RY to pin 8.


October 17th, 2017, 09:46 AM
I'll add that grounding the READY/ line doesn't work for all software. There is stuff out there that determines a disk change by waiting for READY/ to go inactive and then active again. You may need to use a switch on that line.