1050 tandon vs the other mech

[cwilbar]

7 hours ago, Nezgar said:

Yeah, in theory you'd see the amplitude of the target track decreased somewhat.

 

The ROM has to send two steps to the stepper to for a complete track step in a 40-track drive. 1 step per track for an 80 track drive. The stock ROM in a Tandon 1050 actually performs an extra step forward/back at the end of a forward seek (Half physical track), but not when stepping backwards. Atari's WST stock ROM does not do this.

 

It would be a consistent 1/4 track offset on a 40 track disk across the entire disk.

Is it known why Atari did thiis ?  It seems odd that it would do this since in theory a half track forward, a half track back will leave you right where you were.  Does it happen for all operations (read / write / format) or only format or something ?  Seems that even if the step rate was not set on a Tandon mech to the same as on a WST, that the half step forward/back being patched out would improve Tandon 1050 performance.

 

I prefer the construction of the Tandon mech myself, and since I think all but one 1050 I have/have had is Tandon based, I like that I can easily swap hardware to troubleshoot when a problem arises.

 

I have successfully used a WST ROM with a Tandon mech (in testing anyway).  I now believe it was a 1/4 track off (at least on tracks above 0).  Alignment was close enough that it appeared to work.

 

Anyone patch out the extra forward a half track, back a half track on the 1050 ROM ?

[+Nezgar]

On 6/29/2020 at 12:46 AM, _The Doctor__ said:

Think about that as the head moves back and forth essentially skipping a beat on each reversal (the correction back step going forward but none on going back)

Advance to next track = 3 forward steps, 1 reverse step. (Net effect = 2 forward steps)

Retract to previous track = 2 reverse steps

 

No cumulative effect. Only a slight increased time penalty on a forward seek.

[_The Doctor__]

isn't this one of the major reasons the ROMs are made to the mech, I am aware of the logic used to make it work. That is why it said I don't think using the WST ROM and Tandon ROM interchangeably was a good idea. Without correction it would continue to walk wouldn't it?

[cwilbar]

The distance each step moves would be the same, it's just that the shaft of the stepper would be 45 degrees offset due to the differences (as I understand it from the previously posted link on steppers).

 

So there is no difference except a fixed amount of deviation between them.

 

If I get my alignment testing figured out, I'm going to test this at some point... align a tandon with tandon firmware, then put in WST firmware, and see if the alignment is affected.  If I'm understanding everything correctly, it will be off.  Probably good enough if the drive was aligned fine to begin with, and all the media used in it was written with an aligned drive.  Its where differences in alignment in opposite directions, that small partial track difference would come into play enough to cause read/write errors.

 

 

[+Nezgar]

Sorry @_The Doctor__ I finally get what you mean!

 

Since we don't know exactly why the Tandon firmware was programmed to do the extra step forward/back, I agree it is important to ensure that function is preserved in any 1050 ROM driving Tandon mechs.

 

(Along with the aforementioned 45 degree stepper offset issue)

[ijor]

Interesting information indeed about the stepper.

 

I'm not sure, but I still believe the WST mechanism should be "backwards" compatible. Otherwise all the 1050 enhancements would not work reliably and the issue would have been known. If it is the other way around, if the WST ROM can be used reliably on a Tandon mechanism, this is less certain.

 

Anybody knows more or less when the first WST mechanism were brought to the market?

[+Nezgar]

4 minutes ago, ijor said:

I'm not sure, but I still believe the WST mechanism should be "backwards" compatible. Otherwise all the 1050 enhancements would not work reliably and the issue would have been known. If it is the other way around, if the WST ROM can be used reliably on a Tandon mechanism, this is less certain.

The thing is, it would mostly work and be transparent to the user. The drive would just increase it's gain when reading an affected disk from another drive to compensate. Healthy disks from other drives, and disks formatted in such a drive would generally have no issues...)

 

Personally, I had a WST/Hong Kong drive with a happy board which uses Tandon stepper encoding. I never noticed any issue in practice, and probably moreso due to most of my personal disks were formatted and used in the same drive. The only issue could have been when reading the disk in another drive, and even more so 30 years later... Maybe I would have had less bad sectors when archiving to ATR's if I used the original WST drive with happy to read them back...

 

12 minutes ago, ijor said:

Anybody knows more or less when the first WST mechanism were brought to the market?

Here's a good excerpt from @hunmanik's FAQ. The answer appears to be Nov 84 (until Feb 85).

The 1050 was made for Atari in Singapore by Atari-PCI Enterprises Pte. Ltd.
from May 1983 to December 1984 (very common), in Hong Kong by Atari-Wong Co.
from November 1984 to February 1985 (fairly rare), and again in Singapore by
Atari-PCI Enterprises Pte. Ltd. from October 1985 to December 1985 (also
fairly rare).

1050 drives made in Singapore use a Tandon drive mechanism, while 1050 drives
made in Hong Kong use a World Storage Technology (WST) drive mechanism.

Several different 1050 ROM (Atari part FC100541) versions have been documented
by Atari users, with no practical differences among them.  1983 units shipped
with Rev J, 1984 Singapore units shipped with Rev K, 1984-1985 Hong Kong units
shipped with an alternate version of Rev K, and 1985 Singapore units shipped
with Rev L.  See: http://atariage.com/forums/topic/156462-1050-roms/?p=3580891

1983 production units shipped with DOS 2.0S and in North America only, units
produced winter 1984 to winter 1985 shipped (worldwide) with DOS 3, and units
produced fall 1985 shipped with DOS 2.5.

 

[_The Doctor__]

yep, if you read carefully you see hong kong were world storage and had an alternative k rom... just piece it together.

in Hong Kong by Atari-Wong Co.
from November 1984 to February 1985 (fairly rare)

while 1050 drives
made in Hong Kong use a World Storage Technology (WST) drive mechanism.

 1984-1985 Hong Kong units
shipped with an alternate version of Rev K

I also remember some chatter about upgrades not working perfectly with WST drives, with possible solutions or alternative ROMs. I don't envy the task of trying to find old BBS bases, news groups, and online service talks about it (I think those are all mostly gone forever) even here on AtariAge with 4,551,231 or so posts to sift through this would be nightmarish...

Edited July 1, 2020 by _The Doctor__