XF551 Troubleshooting

[1050]

Throughout all this I've been lurking with not much

help to add. But the last post makes it plain to me

what might be going on. It's the same guy cleaning the

heads on both defective marginal working drives...

 

WD-40 works for a head cleaning solution but it requires

5 minutes of time by the clock. Most people have access

to it and if cleaned up afterward then becomes a more

universal solvent to use than the 90% IPA that historically

is the poison of choice. A goodly number of bottles in

the store will be found to contain 50% IPA though and

some stores don't carry the 90% version at all. You'll

have to buy at some other store then and it can be

troublesome to find in that case.

 

Using 50% instead of 90% isn't going to work ever, it

just isn't the same cleaning potential. And it still

takes time as the most important ingredient as well.

Perhaps half the time of WD-40 so 2.5 minutes of wetted

movement across the head. Movement need not be violent

or vigorous, it can be slow and lazy with the eye on

the second hand. A longer 150 seconds doesn't exist

and when that becomes a full 300 with WD-40 it's all

I can do myself. But this agony is the most important

ingredient needed in proper cleaning of the heads.

 

Every symptom you have screams dirty heads to me.

Having two drives with the same symptoms isn't a

normal situation unless you are not cleaning the

heads properly and only then it makes perfect sense.

And then it screams even louder at me again.

[telengard]

I'll definitely go over the heads again, I had done the usual scrub w/ 99% IPA. The behavior seems very specific in that it barely works, and that same behavior happens with both mechs.

 

Also, tested out the regulators this morning, rock solid 5v and 12V, including under load.

 

I did notice what looks like a transistor also attached to the big metal enclosure for heat dissipation, what is that for?

 

Also, what did you mean by "it's the same guy"?

 

thanks again everyone for the help, at this point I have to be narrowing in on the issue as a lot of things have been eliminated as potential issues.

[+kheller2]

Schematics if you need them: http://atariage.com/forums/topic/284363-xf551-schematics/?p=4142496

[telengard]

Thanks for that!

 

I had some time today so I dug out a bunch of disks and it seems a few select disks have no issue at all, load with no errors. Most have lots of BOOT ERROR, etc. I will test these with other drives, as I doubt I have that many corrupt floppies. And it also doesn't explain why I can't format (have tried multiple disks).

[_The Doctor__]

what are you trying to format at/ with?

[telengard]

Various floppies through a few different DOS verions (Atari 2.5, MyDos, etc). It travels what seems like to every track, comes back to the beginning, I hear 1 beep, and then it repeats. With some DOS it will error out, some it just keeps trying.

 

The part that is odd to me is that some disks load with seemingly no errors (there aren't many that do), and the others get lots of errors. I had swapped mechs and saw the same behavior, but I guess it is possible *both* of them are out of alignment?

 

I still need to test RPMs as well... I'll do that tonight

[+bob1200xl]

You might consider poor clamping force. The jacket gets contaminated on some disks which makes it difficult to spin the disk evenly. The disk will work in one drive (with good clamping force) and fail on another. If you can't verify RPM, it may be clamping. Disks with no center reinforcement are particularly bad. You see this on 1050s, mostly.

 

You should scope your power. 12volts will read 12volts even with 4 volts of hum on it. Try reading the AC voltage on your power regs. Better be zero...

 

I would be careful swapping drives on an XF551. The drives they use draw very little power. Throwing an old 5 1/4 in there may overload the power supplies. It won't hurt them, but it won't make happy faces.

 

I don't have much luck with XF551s, either.

 

Bob

[telengard]

Hmm, that makes a lot of sense. Is there a way to improve the clamping force? I do plan on verifying the RPMs soon. And I did look at the 5 and 12v with my o-scope. Very clean, I didn't see much ripple, etc if any.

[telengard]

Was able to check out the RPMs today, given how slow the drive works when it does work, this *kinda* makes sense. But I would have figured that there was a range of RPMs that the drive would work (around 300rpm or so). It's also possible the program isn't correctly reading the RPMs. I've attached a screenshot. Some disks wouldn't register at all. One of the disks that loaded fine, but very slowly did give some readings of 36 and 37.

 

 

Edited December 24, 2018 by telengard

[+Nezgar]

You mentioned you're using it with a 130XE - have you tried it with any other models, like an 800XL? The reason I ask is that NTSC 130XE's have speed limiting capacitors on the SIO bus which have been known to cause problems with high speed SIO, moreso than other models, to even cause issues with normal speed SIO with XF-551. The drive's stock NTSC ROM bit timing is very close to out of tolerance for NTSC machines, and those caps can cause the computer to miss bits:

 

http://atariage.com/forums/topic/138304-cant-flip-disk-on-xf551/?p=1671218

 

So you could remove the speed limiting caps from the 130XE, or replace the drives ROM with hyper XF that has SIO bit timing that NTSC machines can tolerate better...

[telengard]

I’ve also tested with an 800XL and seeing the same behavior.

 

Does the RPM output seem legit? I had tried a different mech and got the same behavior, it seems improbable 2 mechs could both have super slow RPMs. The control board seems to be the issue and I‘ve already redone the solder joints of the SIO ports and checked the regulators. What’s left on the board?

 

Also, I cannot find the pot to tweak the motor speed to save my life, where is it? I don’t plan on touching it unless it needs adjustment, just curious to where it is.

[+Nezgar]

300RPM is equal to 5 revs per second. 37 RPM is way off, I think being skewed by the read errors. Can you eyeball if it looks like it's turning ~5 times per second or not? (37 would mean like 1 rev per second)

 

I'm not sure where the speed pot is, maybe someone else here can answer that.

[1050]

There isn't a speed pot on the XF-551 since it's speed

is governed by a phase locked loop to precisely 300

RPM

 

Read errors will ruin the RPM check code every time.

You will not be able to check RPM until you can read

reliably. Don't be surprised when it says 288 RPM either,

downgraded system clock is properly mimicking a

standard Atari disk drive.

[telengard]

Hmm, so that is most likely working OK. I'll try and eyeball it for rotations, but I'm not sure what to check next on the control board. Are there some common failure points on the board other than the SIO jack solder joints?

[telengard]

The drive is back on my bench and I decided to try the other mech I had again just to be sure. I have a feeling that it is not compatible. It fails in a similar way, but the timing is different. I did change the DS jumper to what I believe was 0 (and pretty much every other setting and the same behavior). Model is Panasonic JU-475-4CGJ.

 

For 360k mechs I also have a Teac FD-55BV and a Panasonic JU-455 I could try.

 

Is there a way to tell if a mech is compatible?

 

And related to this, are the symptoms I've described so far, very intermittent/slow read ability indicative of a head going/gone bad?

[BillC]

Throughout all this I've been lurking with not much

help to add. But the last post makes it plain to me

what might be going on. It's the same guy cleaning the

heads on both defective marginal working drives...

 

WD-40 works for a head cleaning solution but it requires

5 minutes of time by the clock. Most people have access

to it and if cleaned up afterward then becomes a more

universal solvent to use than the 90% IPA that historically

is the poison of choice. A goodly number of bottles in

the store will be found to contain 50% IPA though and

some stores don't carry the 90% version at all. You'll

have to buy at some other store then and it can be

troublesome to find in that case.

 

Using 50% instead of 90% isn't going to work ever, it

just isn't the same cleaning potential. And it still

takes time as the most important ingredient as well.

Perhaps half the time of WD-40 so 2.5 minutes of wetted

movement across the head. Movement need not be violent

or vigorous, it can be slow and lazy with the eye on

the second hand. A longer 150 seconds doesn't exist

and when that becomes a full 300 with WD-40 it's all

I can do myself. But this agony is the most important

ingredient needed in proper cleaning of the heads.

 

Every symptom you have screams dirty heads to me.

Having two drives with the same symptoms isn't a

normal situation unless you are not cleaning the

heads properly and only then it makes perfect sense.

And then it screams even louder at me again.

I don't know about elsewhere, but I have found that many pharmacies in Canada carry both types. I have purchased 99% IPA from 3 different ones over the years, including in a town of less than 5000 people.

[1050]

Model is Panasonic JU-475-4CGJ.

1.2 meg drive there, why are you trying it?

 

Is there a way to tell if a mech is compatible?

Yes, using a PC's BIOS that is capable of setting for

either 1.2 meg or 360K types, set it for 360 type and

boot. If you get floppy fail:40 error message then you

have the wrong type connected. On the PC drive ID needs

to be drive #1, on the XF-551 drive #0.

 

And related to this, are the symptoms I've described so far, very intermittent/slow read ability indicative of a head going/gone bad?

Wrong media might be a problem and I've never tried

the 1.2 meg drives in a XF-551 since it's bashing of

the head against the wall for no good purpose. Stick to

360K only drives and you might get better results.

 

Complete testing as a PC drive first might be the

best approach to discover bad habits and other failings

of a drive there where proper results are commonly

given as a test report. I don't have a DOS mode

program to recommend for this task as I just move

onto a working drive instead. Windows will not be

able to do very well with a 5.25 drive since they

were passe at the time and no consideration given

to windows drivers for 5.25 flawless access. Instead

you will get countless requests to format the disk

again and again and none will ever succeed in getting

rid of the format pop up in windows.

 

Nature of the beast - 9x machine, stay in true boot to

DOS mode for real results. Don't have it? Then you can't

do it.

[telengard]

Thanks for the info, I had the JU-475-4CGJ written down in my inventory as a 360k drive, seems I was wrong there whenever I had written that down.

 

I don't think I've ever run across an intermittently working head. The drive works, it's just slow to load or just gets an error every other "beep" and even then seems to load.

 

I'll dig out another 360k mech, I believe I have a few more PC based ones... will 1050 mechs work? I have a bunch of those. My guess is no since they aren't "PC" mechs.

[1050]

Your guess is correct, 1050 mechs do not work, sorry.

 

Part of the read problem with 475 panasonic might

be the half width head reading data.

 

Some have used 1.2 meg drives, but I've never tried.

Can't say I can fathom how they got it done either.

[hueyjones70]

I have two xf551 mechanical that are not working. I know the pcb is ok, I tried it with another mech. One gives a beep about every two seconds and eventually gives a boot error. The other makes sounds as if it is going to boot for about 3 or 4 seconds then it starts beeping about every two seconds and eventually gives a boot error. I have cleaned the heads but not as thoroughly as this article recommends. I will clean them better tomorrow. Is there any other obvious thing I should check.

[1050]

A good bootable disk in the drive is paramount to a

successful boot up. No HD disk, no blank disks, etc.

 

Best place to check a XF-551 mech is on a working PC

(9x) that is set up for a 360K mech and it works as a

MS-DOS disk drive. If it functions there, then the

move to the Atari is almost easy.

 

9x is shorthand for winders 95 98 or ME. Boot to DOS

mode only please - winders will take a dump on it

otherwise.

[telengard]

I did get to the bottom of this. Replacing the mech fixed it. I'm surprised it could work intermittently like that, but maybe it was just the head.