XF551 Floppy Issues

[kbj]

Hey all,

 

I have an XF551 floppy and it won't read anything / format.  Does it's usual head dance when switched on (moves out a fraction and then back in).  When I format a disk in SpartaDOSx, it seems to do the format 4 times - using a DS DD disk but trying to format it as SS 40 single sided.  It will step out 39 steps, move back to the start and do that 4 times.  It will then tell me that it can't format the disk.  Does this with every disk I've tried (even those that format fine in a 1050).

 

I'm not sure about the track zero sensor - I think it works, but it looks to me as if it's been tampered with by someone else.. are there instructions somewhere that show the steps to calibrating it?  Service manual or software diags?

 

The clock on the baord is nice and clean at 8.333Mhz.. scoping the write/read pins on the 1772 show that data is being written when formatting and I think it looks fine too.

 

It's a Mitsumi mech if that helps.

 

Can anyone else shed any light?  

 

Much appreciated as always!

 

KbJ

 

[TGB1718]

Head clean

RPM check, should be same as 1050 rotational speed

 

If the track 0 sensor is out, it should still be able to format, but not be able to read/write other disks formatted on other drives.

 

Don't know if the Atari Diagnostics disk will work on an XF551, if it does, then easy to check rotations speed and align the head.

1050 Disk Diagnostics (1983)(Atari).atr

 

Obviously you need to be in a position to load the disk, so put the .atr on the 1050 and boot from it.

Edited September 9, 2021 by TGB1718
Update

[kbj]

Hi @TGB1718

 

Thanks for the infos.. I have cleaned the heads twice now with 99% IPA.. and they look spotless to me.

 

Ive tried the 1050 diagnostics RPM test but it reports 3289RPM - and from what I've read, there's no way to change the RPM on an XF551 as it uses a PLL or equivalent to set the speed automatically?  There's no pot on the board anywhere.. I don't think that the speed measurement from the 1050 diags is correct..

 

Using the RPM test from Sparta won't even start.. it just hangs..

 

Similarly, using the 1050 track zero test, it seems to work but you get to a point where adjustments are so small that you can never get it to say OK - it will just do the FORWARD/BACKWARD each time with the smallest of movements of the sensor.. not sure if that's works with the XF551 either..

 

Using SpartaDOSx from a SIDE3 cart, and formatting a disk (and have tried well over 10 disks now that format fine in a 1050) still goes through the motions of formatting the disk but right at the end when the head moves back to TR0, it will produce an SIO beep on the Atari and give a formatting error.

 

Trying to do a DIR on any disk gives a 144 error.

 

One thing I did wonder about.. With the stock OS on the Drive (v7.7) it does all of the above.  Sticking either Bob's fixed OS ROM or the Hyper ROM won't even get the drive to start a format.  Might it be the case that the SIO routines in the SIDE3 cart with SpartaDOSx are using Fast SIO?  I've tried the SIOSET comand and it does identify drive 1 being (and using) the XF551 settings.  I did try SIOSET 1 NORMAL but this made no differnce.

 

Maddening..

 

KbJ

[manterola]

I would say it is the stepper motor, and/or its position. I fixed one xf-551 by exchanging the motor from a broken mitsumi/newtronics mech.

 

[_The Doctor__]

The very first thing you do with and XF551 is to re-solder and beef up the SIO pins and traces,

the next thing is to make sure the head shield hasn't fallen off and is hiding/interfering with free movement.

Next up is checking the ribbon cable, then checking the stepper motor and alignment...

after that it's track zero...

after that index hole / track zero....

also consider checking the board is easy enough if you have a mech to swap...

if your board is socketed re seat prom...

 

lastly bulk erase any disk you wish to format, and always check each format/read from a fresh power up

 

Edited September 9, 2021 by _The Doctor__

[+kheller2]

The XFF551 can’t use the 1050 diags, nor does it spin at the same speed. Is it just formatting that doesn’t work or can you read disks in it formatted on the 1050?

 

The XF uses stock 360K floppy mechs of the period.  If you have access to an old PC you could swap mechs for testing. 

[+Nezgar]

3 hours ago, kbj said:

I have an XF551 floppy and it won't read anything / format.  Does it's usual head dance when switched on (moves out a fraction and then back in).  When I format a disk in SpartaDOSx

Since you say it can't read anything, it makes sense that after the 40 steps, the next phase is a verify, which is to read what it just formatted, and then should step back 40 steps - but a failure here will jump all the way back to track 0, further supporting that it can't read anything.

 

2 hours ago, TGB1718 said:

RPM check, should be same as 1050 rotational speed

1050 nominal is RPM 288RM. XF551 is 300RPM. Though you will be unable to perform an RPM speed test if it can't read anything, hence the erronous report of 3000+RPM...

 

One thing that would be good to know is with a disk already formatted good on a 1050, then formatted (and failed) in the XF551, put the disk back in the 1050 and see if the contents are still readable, or if they have been erased. This will tell us if the erase/write function is doing anything in the XF551.

 

I'm also thinking that the "easiest" next step would be to swap out a different mech into the unit, to see if the problem is with the controller board, or the mech, but I realize this may be difficult if one doesn't have a spare mech, and bare DD 5.25 mechs are hard to come by and pricey...

 

Edit: a couple other random thoughts: cleaning heads may be in order. also check the power outputs are at/close to 5.0V and 12.0V.

[kbj]

Hey - thanks for all of the suggestions!

 

This is what's already been done:

• Cleaned the heads with IPA - they look spotless
• Changed C14 on the mech board as someone suggested elsewhere
• Removed the solder from both SIO ports and resolderd
• Checked the rest of the PCB for dry joints  - none that I can see
• Metered out both ribbon cables from the mech to the PCB (All good)
• Metered out the other ribbon cable that goes between the 2 PCB's on the mech - ok
• Checked the voltages to the mech - 5.02v and 12.01v
• Compared the contents of the ROM to one from another user's dump of the same v7.7 OS - ok
• Checked the 2 sensors on the mech (one for doisk inserted and the other for the hole on the disk) - both appear to change when a disk is inserted
• Checked the master clock - 8.333Mhz 
• Checked the read/write pins on the 1772 when formatting - and those look good too I'm assuming.. there are pulses in between the steps (is that good?)

 

I did wonder if I could use any of the 1050 diag tests - thanks for confirming that I can't @kheller2..

 

I'll look at doing what you suggest @Nezgar

 

@_The Doctor__ Do you have instructions on how best to calibrate the track zero sensor? Was there ever a service manual produced for the XF551?

 

Thanks again everyone for all of the suggestions/help!  This is why AtariAge is the best!

 

Cheers,

 

KbJ

[kbj]

Just a quick update..

 

Formatted a disk on a 1050 (which I know works as I use it and it passes all the diag tests and will load commercial software).. then conencted the XF551 - it won't let me DIR the disk - get a 144 ERROR disk is write protected or bad sector error message.  Connect up the 1050 again and the DIR is good.

 

If I then take that disk and put it back in the XF551 and format it, it fails to format.

 

Stick it back in the 1050 and I get the 144 ERROR disk is write protected or bad sector error once more.  Looks like the XF551 is destroying the format on the disk when I try to format it..

 

Another thing I noted was that when I connect up the 1050 but don't insert a disk and boot Sparta from my SIDE3 cart, it takes a few seconds to get me to the D1: prompt.  Doing the same with the XF551, it takes about 30-40 secs to get to the D1: prompt with a lot of SIO beeps.. and the disk light is on all of the time and spindle motor is spinning.  It will go off shortly after the D1: prompt appears.  Not sure if this is normal as it's my first XF551.

 

At this stage, it's looking like the mech at fault and not the main PCB?

 

KbJ

Edited September 9, 2021 by kbj
Additional info

[+bob1200xl]

What are you trying to format - SD, DD? The XF551 gets all twisted up when you try to format in DD and you have booted with a SD diskette in the XF551.

 

Make your 1050 D1: and the XF551 D2:. Boot DOS 2.0 in D1: with a SD diskette in D2:. Format D2: in SD - should work.

 

Bob

 

[kbj]

Hey @bob1200xl

 

It doesn't seem to matter whether I boot with a disk in the drive or not.  It still refuses to format any disks (and I've been thorugh 2 whole boxes of disks now).  I've tried SD / DD and no good..

 

I just tried what you said - and no difference.  The XF551 will still not format a disk.  

 

KbJ

[sup8pdct]

On the mech itself, there is a flex cable between head and board. You may be able to see it through a hole in top shield. if it is broken, then mech is gone.

Only thing i can think of trying is to gently remove the cable and reseat, or move it around in the connector. Remove top shield if you need to.

 

James

[kbj]

Hey @sup8pdct

 

Yes, there are 2 flex cables - one from the head and one from the TR0 sensor.  Both are fine - I have removed and reseated them and checked them with a meter for continuity.

 

Ta,

 

KbJ

[BillC]

9 hours ago, TGB1718 said:

RPM check, should be same as 1050 rotational speed

The 810/1050 mechanisms rotate at 288RPM while the XF551 mechanisms are 300RPM, Atari increased the read/write speed by 4% to compensate.

 

[Peri Noid]

According to my knowledge (confirmed with observations of 2 faulty drives) your drive has its head(s) dead. You can try and format a floppy in a single-sided format to check, if the lower head (0) still works - if yes, you can at least read standard SD/ED/DD disks. But I guess it's bad. You can measure the ribbons going from the head to the drive's motherboard, it is usually pretty easy to see, that a head is faulty (not always as with one of my drives but - usually is).

The way to fix it is to replace the mech with another one. Unfortunately, only the original Mitsumi D503 drives fit the case - they are 2cm shorter than regular DD drives. It is possible to push a longer drive but it requires physical modification of the power section and no longer looks original ? I did so in one of my XF551 using an Alps DD drive - I prefered it working than fully original.

Edited September 9, 2021 by Peri Noid

[Sugarland]

 

Yes you might have a dead head

[manterola]

8 hours ago, kheller2 said:

 

The XF uses stock 360K floppy mechs of the period.  If you have access to an old PC you could swap mechs for testing. 

I think this is a good path. The best thing would be to find the same mitsumi newtronics mech. Others 360Kb/300RPM mechs work, however all the ones I have tried (several panasonics and teac) are too long for the space available.

I also did some experiments with 1.2MB newtronics/mitsumi as they are easier to find, and have the needed (short) deep. I made it rotate to 300KB and I was able to format/read/write in the few tests I ran, but the steps were "short" so the disk were unreadable for the other standard Atari drives.

I remember reading about a modification that moves the big capacitors and there is some cutting of the heatsink, too  in order to provide space for a Teac mech. So that is another option if the original mitsumi mech cannot be repair or replaced.

In summary, if you can find the same mitsumi drive or other 360KB/300RPM mech, you would end up with a great device, unless you have the knowledge/instruments/time to repair/align, etc, the mech you have.

Edit: if you dont mind and xf-551 iun original condition you can also replace the mech by a 3.5" floppy mech, but you will also need to replace the eprom to put the right fiwmare for 3.5" mech

Edited September 9, 2021 by manterola
another idea....

[_The Doctor__]

the cable(s) going to the head(s) can be checked as well. can't hurt to touch up where the connect, since hack at a dead thing won't hurt. No one says you transplant some heads either.

Edited September 9, 2021 by _The Doctor__

[+bob1200xl]

Did this XF551 ever work?

 

Bob

[kbj]

Thanks for all of the replies!

 

Not so good if the heads are messed up.. I guess that measuring the ribbon cable, I can see if there's a resistance across some of the pins to see if the heads are truly dead?  It would be great if Atari had produced a service manual for the drive - and If mitsumi had the same for the drive mech.  I don't want to give up on it just yet (although I may be flogging a dead drive here)..

 

The 551 did work up until recently.. It was reading and writing fine, then just started giving those errors.

 

CHeers,

 

KbJ

Edited September 10, 2021 by kbj
Spelling and missing part

[TGB1718]

11 hours ago, BillC said:

The 810/1050 mechanisms rotate at 288RPM while the XF551 mechanisms are 300RPM, Atari increased the read/write speed by 4% to compensate.

Thanks @BillC I didn't know that

[kbj]

12 hours ago, Sugarland said:

 

Yes you might have a dead head

Jsut noticed this - priceless! ?

[1050]

Only because I can easily do this, I would install the mech in an old school windows 98 box and go with MS-DOS tools to do the testing of the mech there. So many more tools on that side of the fence for that kind of diagnosis. Sounds like somethings up and you'd be light years ahead to know where the problem really is. The dodgy XF-551 board or the mech alone.

[+Larry]

Two thoughts, if you can determine that it is a mech issue: someone in your part of the world may have spare mechs from 3-1/2" conversions.  The XF551 also makes a great 720K drive, using either a 720KB or 1.44 MB mechanism.  Probably the best currently available conversion is the HyperXF from Stefan Dorndorf (since it will do true UltraSpeed).  If you have a burner, you can burn your own rom, but here's a link if you don't:

 

https://www.8bitclassics.com/product/atari-xf551-disk-drive-rom-upgrade/

 

 

 

 

[Peri Noid]

With Hyper-XF 3.5" ROM (ver.B) you can use both 3.5" and 5.25" HD mechs and format the disks to 720KB. The problem is, such a drive is no longer compatible with other Atari disk formats. I have two 3.5" drives like this and I also played with a 5.25" one.