Modding a TEAC FD-55 for double-step (80 to 40 track conversion)

[OCT]

As the thread went nowhere in German so far (and the author of the original modding instructions has just notified me that he has discontinued work on the CPC), this one goes out to the world at large (with a nod to R.E.M.):

Besides a comprehensive manual from 1983 (!) and a RY (Ready) jumper to make them compatible with CPCs, the TEAC FD-55 line of 5.25 inch floppy disk drives have the advantage of of being switchable from 80 to 40 tracks by cutting one of its resistors (link in German).

However, as discussed at http://cpcwiki.eu/forum/index.php?topic=608 (in German), there seem to be variations among revisions of the circuit board (unsurprisingly for a line the has run that long).

I have made a hi-res picture of the version I ended up with (an FD-55GFR 193-U (not exactly like http://www.teac.com/DSPD/pdf/5fd0050a.pdf either) that regrettably came with the PCB unscrewed, i.e. Track 0 misaligned, and would very much appreciate any information on how the mod should be carried out on this one, or further details/images on how it was successfully done on other TEAC FD-55s or their East German Robotron K5601/CM5643 clones:

(Click to enlarge)

[Gryzor]

Can I ask - what's the practical use of this after all?

[OCT]

The advantage is that it allows you to read 40-track and 80-track disks in the same drive, and switch between them as needed for the respective floppy disk.

On most other drives this would require much more elaborate step-doubling circuitry.

Few of us could afford 80-track 5.25" media and drives back in the day; the 40-track ones (similar in capacity to the 3" ones used on the CPC rather than the Joyce) are still the most common to find with CPC software on them (and the drives only have a specified lifetime of 5 years, i.e. will require replacement every now and then though the 40-track variant has not been build anymore for 2 decades), but not readable on 80-track drives which use 96 rather than 48 tracks per inch, i.e. advance the R/W head only half as far per step: They would need to step from physical track 0 to 2 rather than 1, etc. (harder to work around the resulting mismatch between physical and logical track numbers in software such as ParaDOS, if a simple hardware switch will do).

Of course, no media should be written in double-step on an 80-track drive if it ever needs to be readable on 40-track ones again, but the double-step mod allows at least for the easy reading/retrieval of data from times immemorial.

[Gryzor]

Ah - sorry, I thought it permanently converted the drive, that's why I couldn't see the use... thanks for the explanation

[OCT]

The mods are discussed at some length in http://z80cpu.eu/archive/news/comp.os.cpm/threads/df15c71978c2bdec.html;
see the contributions by Jürgen Strobel, Apr 25 2005 7:17 pm; John Elliott, Apr 25 2005 8:04 pm; and nos, Apr 27 2005 5:25 pm (specifically mentioning the GFR variant) in that thread; cf. also the reference to the FD55-GFV-17 in http://www.classiccmp.org/pipermail/cctalk/1999-March/122818.html indicating that 1.2M ("HD") drives could indeed be converted in the same way.
So the knowledge how to do the trick seems to have been pretty wide-spread until rather recently; even to the extent that there is no further information on how to locate the equivalent of "R19" in slightly modified revisions.
Since more than a few CPC owners must have done this mod already (and it's going to stay a repetitive chore), of course I am hoping to get the details (and share whatever information we can find) here, with the scope expanded to the common denominator of English.
(It seems counterintuitive only at first that this may simply be one of the resistors in line with the stepper motor coils, as depending upon their design, these may indeed make a wider, i.e. double step when the next coil is not powered - but I wouldn't want to solder things apart based on guesswork alone, as I only have my CPC within reach every few weeks these days, and 5.25" drives with READY are hard to come by, with only a few up for auction on eBay each month...)

[OCT]

Harold Bower in http://www.mail-archive.com/linux-hardware@senator-bedfellow.mit.edu/msg02692.html (28 Apr 2000 08:47:45 GMT) further explained that:

If you can find some old 720K drives by TEAC, you will find a jumper that actually forces a double-step the drive.  It appears that they used a common stepper and electronics between 40 and 80 track drives (at least the ones I had in the mid 80s), and the technology was carried forward.

[OCT]

The Atari folks also compiled some info (in German) back in 1986, on earlier models that are long gone (5 years TTL) of course:
http://www.stcarchiv.de/stc1986/04_teac.php
http://www.stcarchiv.de/stc1986/09_fremdfloppy.php

[Cholo]

Wow, was a bit surpriced when i saw this post. Last week i ordered a Teac FD55BR-102U DS/DD 5.25" floppy (total random pick on ebay) and today i got word from the PO that the parcel with the drive has arrived. A bit scary but all this information will come in handy 

[OCT]

Nothing scary about it, rather good to know it has a chance of working on a CPC - but please do make detailed photos as we might have a chance of figuring out what the notorious R15/R19 (10 or 100 ohms, who knows...) and their equivalents on the FD-55's other revisions are.

[Cholo]

Okies, Got the drive yesterday and quick comparison with your pic and at first they looked the same (same jumpers and such). Still, R15-19 does indeed seem to be located quite differently tho. Ill get some pics up as quick as possible.

[Cholo]

Okies here are some pictures:
http://img6.imageshack.us/img6/538/fd5501.jpg
http://img100.imageshack.us/img100/1561/fd5502.jpg
Hope they are readable (best of 200 blurry pictures).

And here is one where i marked R15 and R19:
http://img297.imageshack.us/img297/6861/fd5503.jpg

The labal at the back:
http://img245.imageshack.us/img245/7720/backvx.jpg

I noticed that we have the same jumpers set, except i see that you have RY and DC connected.

[OCT]

Many thanks for the photos!

here is one where i marked R15 and R19:
http://img297.imageshack.us/img297/6861/fd5503.jpg
I noticed that we have the same jumpers set, except i see that you have RY and DC connected.

Yes, I indicated that I'd taken the picture before setting RY as required for the CPC, as I had to run the drive on a PC first to try and re-adjust the circuit board (which was quite literally "missing that one final screw" that IMHO should have held the Track 0 detector in place).

Unless the drive recognizes from pin 2 that a CPC cannot possibly read/write High Density (unlike 3.5" AFAIK a 5.25" drive itself cannot mechanically tell the two types of disk apart), further jumper settings may be required elsewhere to set the density to 300kbps/300rpm (Double Density) rather than 500kbps/360rpm (HD): http://www.oldskool.org/disk2fdi/525HDMOD.htm#Teac
(details in brackets are subject to debate http://z80cpu.eu/archive/news/comp.os.cpm/threads/f485ff79b9988caa.html).

Is any of R15 and R19 of the required impedance (10 or 100 ohms, i.e. brown-black-black or brown-black-brown) for the double-stepping mod (upon which hopefully someone will elaborate further)? None of them seems to be near a crystal, or controlling the stepper motor.

http://www.teac.com/DSPD/support/floppy_drives/floppy_drives.htm should come in handy to tell the various flavours of the FD-55GFR apart, and http://www.bitsavers.org/pdf/teac/FD55BV_Spec.pdf as well as http://www.bitsavers.org/pdf/teac/FD55_Spec.pdf (FD-55A through E) provide details on yet another revision.

[Cholo]

Im guessing its a quite different setup on my drive as i can see that R19 is Green-Green-Red and R15 is Light brown-gray(dark brown/black)-Black i think.

Noticed someone selling another Teac FD-55GFR 142-U here:
http://cgi.ebay.co.uk/TEAC-FD-55GFR-142-U-DS-HD-1-2Mb-Floppy-Disk-Drive_W0QQitemZ330394737346QQcmdZViewItemQQptZUK_VintageComputing_RL?hash=item4ced0f9ac2

[OCT]

In this case I wouldn't follow G.o.S.T.'s advice (all FD-55s said to be using the same board) either - cutting any of these will probably do no good and we should certainly try and find out what exactly the mod was supposed to do (disabling one of the stepper coils?!).
Someone around here or in the CP/M forums/newsgroups has got to know...
...or maybe at http://www.robotrontechnik.de/html/forum/thwb/board.php?boardid=1 as remakes for both the drives (K5601/CM5643) and the CPC itself (KC compact) had been produced in East Germany.

[OCT]

Im guessing its a quite different setup on my drive as i can see that R19 is Green-Green-Red and R15 is Light brown-gray(dark brown/black)-Black i think.

Noticed someone selling another Teac FD-55GFR 142-U

If he gets the asking price, I should hurry to resell mine.

For the record, given the confusion even among the experts at http://www.classiccmp.org/pipermail/cctalk/2005-March/189289.html regarding density and rotation speed, I tried every combination of the I and LG jumpers on my FD-55GFR 193-U and found out that on this revision at least, both have to be set in addition to RY (from DC, i.e. each a change to the drive's HD-capable defaults for PCs), as in the photo below.

Since questions about double-stepping are still pending all over the place, I have used ParaDOS to "emulate the mod", i.e. double-step the drive in software for the time being, according to what I figured out at http://cpcwiki.eu/forum/index.php/topic,624.msg7155.html#msg7155 - which makes the drive happily read entire 40-track disks (rather than just the directory) as much as 20 years old (unless&until it hits the hardware limit of 40/80 tracks, forged in solid steel towards the hub on these drives - and that's a big "but" for anyone who's used their disks to the fullest that earlier mechanisms would permit, i.e. at least 42 tracks).

[OCT]

...or maybe at http://www.robotrontechnik.de/html/forum/thwb/board.php?boardid=1 as remakes for both the drives (K5601/CM5643) and the CPC itself (KC compact) had been produced in East Germany.

And indeed they did.
@Cholo: If your drive has a TEAC 2206 controller, you're in luck; DL has just provided the low-down in pictures at http://www.robotrontechnik.de/html/forum/thwb/showtopic.php?threadid=5732#50860 for you.
Will have to wait for similar details on the TEAC 2207 IC found on my drive, with a different pin count, and hence pinout...

[Cholo]

Quite a different model i got i think. Neither got the 2206 or 2207 IC .. its called 13443226-00 BA9581K 850 104A. Then again the 226 might mean 2206 perhaps. Still no resistor (like R19) near the same area as in those pics. So ill better not try any crazy randomness 

[OCT]

OIC you had an external image of it at http://img6.imageshack.us/img6/538/fd5501.jpg - from the pin count (80 rather than 100) it looks like yet a different beast than the one on my drive.
Let's hope Robotron lasted long enough for TEAC to tell them their secrets on the later controllers as well, or that someone will be able to provide info on which is their TRK pin.

[TFM]

Hi there! This thread has a bunch of valuable informatins. We should make a wiki site out of it.

[Gryzor]

+1. Can anyone make a summary of all this?

[xesrjb]

I have to check this...

xesrjb