PDA

View Full Version : Commodore 4040 Woes



KevinO
December 6th, 2016, 07:40 PM
Tonight I fired up my 4016, which is working today because of help from folks here. I decided I was going to play some old PET games. I currently have ONE disk full of PET games, and it's got stuff on both sides. Or rather HAD stuff on both sides. I completely forgot about an oddity on this system. If you boot the drive (or the computer, I think) with a disk inserted, the disk is erased. Every time. So now my disk is blank.

So I've got two levels of questions. First, is there something I might do to repair this 4040 drive, and someplace I can start?

Or second, do I just give up on this drive altogether? I'm a huge C64 enthusiast, but I never use a physical disk drive anymore. I always use SD card solutions. So what about the Pet? Mike had the PETDisk way back when, but it sounds like there's other options now. Can someone outline what's available and give some recommendations? Thanks.

Ruud
December 7th, 2016, 03:36 AM
What is your definition of blank? If you erase a track that isn't track 18, the directory track, it should still be possible to read the directory. I can imagine that due to an hardware error the track under the head is erased the moment you power the drive on. If the last action before you powered the drive off was reading the directory, the head is still above the directory track and it could be eased then.

You can test everything by moving the head to one of the outer tracks when the drive is powered off, insert a formatted disk, power the drive on and read the directory. If the directory is gone, then I will be suprised.
If it could be read indeed. Power the drive off and on again and read the directory. If it is gone now, there is a hardware problem. What kind of a problem, I honestly have no idea right now.

Good luck!

Groetjes, Ruud

Kakemoms
December 7th, 2016, 09:37 AM
Very strange! Are you sure it has been erased? Maybe you should check it in another drive. The XX40 drives tend to get communication problems over time due to a poor implementation of the IEEE interface HW. It can lead the poor read performance so that loading fails, e.g. 6522 or other HW related problem gives biterrors. If it reads in another drive I would check the 4040 PCB board.

dave_m
December 7th, 2016, 12:29 PM
What is your definition of blank? If you erase a track that isn't track 18, the directory track, it should still be possible to read the directory.

Hi Ruud,
I had a Model 4020 drive in the late 70's and a 4040 for the last several years. Both will corrupt a diskette such that the directory is garbage if there was a power cycle with the diskette seated. In the old days we all got used to removing the floppy after LOADs and SAVEs. Also with the old 4020s at least we had to remember to Initialize the floppy when first inserted or risk corruption due to old directory tables (BAM?) in the 4020 memory.

Early Commodore users were soon well trained after being 'bit' by their PETs. <smile>

KevinO
December 7th, 2016, 02:31 PM
Actually, Ruud was right. I shouldn't have used the word 'erased', but I had assumed that the disk was damaged in some way as to make it unreadable, but when I manually re-positioned the head, it suddenly became readable again. I have very little familiarity with this drive, but it seems like it's losing which track its on. I've got very little familiarity with this drive, but I clearly need to dig into it further and try to figure out what is really happening.

There's also some kind of bad connection, because my first directory listing gave me the listing mixed with garbage characters. When I reseated the connectors from the drive mech to the PCB, that seems to have gone away.

dave_m
December 7th, 2016, 04:16 PM
Actually, Ruud was right. I shouldn't have used the word 'erased', but I had assumed that the disk was damaged in some way as to make it unreadable, but when I manually re-positioned the head, it suddenly became readable again.

Hi Kevin,
I don't quite understand. Are you saying that simply removing the floppy, and reinserting and performing an Initialize or a Directory read did not work to clear the problem (assuming a non-corrupted floppy)? And you had to physically move the head?

KevinO
December 7th, 2016, 05:43 PM
Hi Kevin,
I don't quite understand. Are you saying that simply removing the floppy, and reinserting and performing an Initialize or a Directory read did not work to clear the problem (assuming a non-corrupted floppy)? And you had to physically move the head?

I wouldn't say that I've proven this exactly, but that was what happened. Ridiculous? Maybe I need to see if I can repeat the experiment!

I've had something similar with some other drives. I can't remember for sure, but it seems to me like it was a 1541-II that would "get stupid", and stop reading ANY disks, but moving the head by hand seemed to shake it out of that behavior, and then it would be fine for a bit. Maybe this is completely different....

dave_m
December 7th, 2016, 11:19 PM
Or second, do I just give up on this drive altogether? I'm a huge C64 enthusiast, but I never use a physical disk drive anymore. I always use SD card solutions. So what about the Pet? Mike had the PETDisk way back when, but it sounds like there's other options now. Can someone outline what's available and give some recommendations? Thanks.

Here is some good info on the different types of SD card replacements for PET IEEE drives. See matrix at bottom of the page.

PET SD Card solutions (http://www.primrosebank.net/computers/pet/projects/pet_petsd_alt.htm)

The one thing I would like to know is which one of these (if any) support dual drive emulation? This is important for use in the software applications for the SuperPet which expects the system disk in drive 0: and data files in drive 1: with both being IEEE device 8.

Holmes
December 8th, 2016, 04:17 PM
Here's a thread where the creator of PetSD chimes in on the dual drive support:

http://www.vcfed.org/forum/showthread.php?27208-petSD/page2

His response is in post #15. Seems to be possible by creating two partitions on the sd card.

AndyG
December 8th, 2016, 11:14 PM
As a reference, I use the PET MicroSD card from Dave Curran (Tynemouth Software) ..... been working great for me. I use the internal version after a forgetful moment with the externally mounted variant :cool:

He is based in the UK and It doesn't support dual drive though

Kakemoms
December 8th, 2016, 11:15 PM
I wouldn't say that I've proven this exactly, but that was what happened. Ridiculous? Maybe I need to see if I can repeat the experiment!

I've had something similar with some other drives. I can't remember for sure, but it seems to me like it was a 1541-II that would "get stupid", and stop reading ANY disks, but moving the head by hand seemed to shake it out of that behavior, and then it would be fine for a bit. Maybe this is completely different....

The "banging" of the read/write head (the noise you hear when you switch it on) to initialize these drives to position 0 (especially the 1541), tend to dislocate the positioning of the read/write head over time. In such a case, the drive will start with the head in position 0, but slightly misadjusted. E.g. its real positon is half way towards -1 or 1. By manually touching the write/read head you will dislocate it so that it may get closer to position 0 (when it initializes) and therefore manages to read some data. After a while it goes back to the original misadjusted position and reading/writing becomes difficult.

For such a drive, formatting and subsequently writing data to a disk will result in all the tracks being written in a misadjusted postion. You may have a case were taking out and reinserting the floppy will slightly dislodge the read/write head and either result in a non-readable situation or a partially readable situation. It will be mostly due to chance.

To correct this problem there exists software that can help in adjusting these drives (its available on the C64 at least). It involves physically adjusting the read/write head, so you probably want to talk to someone with experience in doing that.

A well adjusted 4040 drive should be able to read floppies that were written on a well-adjusted 1541. Please note that the 4040 can't write to a floppy that was formatted on a 1541 (due to a change in the index bytes).

dave_m
December 8th, 2016, 11:46 PM
Here's a thread where the creator of PetSD chimes in on the dual drive support:

http://www.vcfed.org/forum/showthread.php?27208-petSD/page2

His response is in post #15. Seems to be possible by creating two partitions on the sd card.

I wonder if its been tried with a SuperPet? It seems that one would have to fire up the PET in 6502 mode to issue the CP2 command to initialize the PetSD, and then switch over to the 6809 mode to run FORTRAN, etc.

There will be a reset at the switchover. I would hope the PetSD would stay initialized in dual floppy mode. Right now one needs a dual floppy drive to run the Waterloo SuperPet software. It would be wonderful if one of these SD card gadgets would work also.
-Dave

KevinO
December 9th, 2016, 04:10 AM
The "banging" of the read/write head (the noise you hear when you switch it on) to initialize these drives to position 0 (especially the 1541), tend to dislocate the positioning of the read/write head over time. In such a case, the drive will start with the head in position 0, but slightly misadjusted. E.g. its real positon is half way towards -1 or 1. By manually touching the write/read head you will dislocate it so that it may get closer to position 0 (when it initializes) and therefore manages to read some data. After a while it goes back to the original misadjusted position and reading/writing becomes difficult.

For such a drive, formatting and subsequently writing data to a disk will result in all the tracks being written in a misadjusted postion. You may have a case were taking out and reinserting the floppy will slightly dislodge the read/write head and either result in a non-readable situation or a partially readable situation. It will be mostly due to chance.

To correct this problem there exists software that can help in adjusting these drives (its available on the C64 at least). It involves physically adjusting the read/write head, so you probably want to talk to someone with experience in doing that.

A well adjusted 4040 drive should be able to read floppies that were written on a well-adjusted 1541. Please note that the 4040 can't write to a floppy that was formatted on a 1541 (due to a change in the index bytes).

Hmm....I didn't know that last fact. I remembered that the disks were different somehow, but I didn't know the details. My 2 disks were definitely written on the 1541, made from D64 images. I wonder if that is a factor in this strange behavior?

I'm no stranger to the 1541 or 1541 drive alignment, so we're good there, but this 4040 seems like a very different animal to me. I have not been able to reproduce my "erased disk" problem since, but I have found a couple other disks that I had made last time I had this unit up and running, and they are definitely unreadable either in the 4040 or the 1541. In fact, saying that jogs my memory as well. One of the disks was actually formatted on the 4040 and I had typed in a ROM testing program that we were using to troubleshoot one of my 8032s, and during one of the power cycles of the computer and/or drive, the contents went bye-bye.

While I'd still like to try to fix this drive, I also made the practical decision to buy the PET MicroSD from Tynemouth Software.

AdamAnt316
December 9th, 2016, 07:03 AM
My guess is that the erased disk was due to some sort of magnetic surge which was somehow caused when the drive was turned on. I know that the tape drive in the Coleco Adam had an issue like that, which was compounded by the manual telling users to insert the desired cassette in the drive before turning the computer on. Personally, I first noticed it with the internal drive on my Commodore 128D. Normal disks seemed OK, though GEOS disks eventually started to exhibit weird errors after being left in the drive. I also experienced it when trying to boot a MacWorks floppy in my Apple Lisa. Corrupted two or three of the boot disks that way before I figured it out. :headslap:
-Adam

dave_m
December 9th, 2016, 07:59 AM
My 2 disks were definitely written on the 1541, made from D64 images. I wonder if that is a factor in this strange behavior?

Kakemoms is very right. Absolutely a problem if you had a floppy formatted on a Model 1541 and later SAVED programs from the PET. I ran across this problem when I used a 1541 connected to a PC to download PET programs from the net. I then READ those floppies into the PET using the 4040 with no problems. However, if I later SAVED any programs from the PET into that floppy, I got a corrupted floppy. I learned to load the 1541 formatted program into the PET and immediately switched to a 4040 formatted floppy and then SAVED the program there for future use on the PET.

Andre Fachat wrote an excellent article on the problem. http://www.6502.org/users/andre/petindex/drives/writegap.html

dave_m
December 22nd, 2016, 11:17 AM
While I'd still like to try to fix this drive, I also made the practical decision to buy the PET MicroSD from Tynemouth Software.

Kevin,
Are you just worried about the power cycling corrupting floppies or is there something else?