The TRAK track buffer?

[burkhart87]

Anyone know of a schematic for these Trak drives?

 I think I have an undocumented ROM, Ver. 1.10.  

.....and look at pin 20 of the 8031;

.....and this lacks (empty socket) the second RAM chip found in the other drive in this thread.

 

This was my first drive, 1983, never had any problems until recently.  AFAIK, it was purchased new by my parents and has never been opened until now.  So it's functioned for 40 years with that bent pin. 

[burkhart87]

Here is the Trak V. 1.10 ROM.

trak AT v1.10.bin

[burkhart87]

The symptom: my AT-D2 was not reading or booting properly.  It would start the boot process (e.g. a DOS disk) but stop at sector 14.  Or, with the TRAK as D2 I could boot a DOS from a working drive then read the D2 directory, but the directory information would often be garbled or incomplete.  Anyway, the problem was the 8031, a replacement restored the drive.

 

I started tracing/testing the power circuit as a first step and started a hand-drawn partial schematic.  Since no schematic exists, probably I should make a copy for sharing and posterity.  I have no idea how to do that though.  Any suggestions would help.

 

The ribbon cable connecting the main board with the front panel board broke as soon as I looked at it.  This seems to happen to everyone, so beware when you open one of these.

 

I cannot get any of the the other ROMS (1.11, 1.12, 1.13) to work with my drive.  Upon powering, they cause both single and double density LEDs to light up and the drive does not post to the SIO.  My mainboard is also the earliest I've seen (Rev. B, compare with KHeller's pic in this thread), with a different power supply/relay junction and less of what I think is a ground rail.  Maybe this 1.10 ROM needs this Rev.B board.  I checked the M2732's I flashed with the later ROMs.  They look OK at first glance when compared to the ROM files themselves; i.e. overall file length, first and last 10 or so hex values, "copyright TRAK Microcomputer" ASCII found at the end of some of the ROMs.  So I don't think its a case of flashing bad ROMs.

Edited September 26, 2023 by burkhart87

[_The Doctor__]

a bad what?

[burkhart87]

19 minutes ago, _The Doctor__ said:

a bad what?

Sorry, hit the wrong key and it posted before I was done typing.

 

[+MrFish]

1 hour ago, burkhart87 said:

Here is the Trak V. 1.10 ROM.

Thanks for posting the ROM.

 

1 hour ago, burkhart87 said:

Maybe this 1.10 ROM needs this Rev.B board.

Possibly.

 

[+MrFish]

1 hour ago, burkhart87 said:

Here is the Trak V. 1.10 ROM.

trak AT v1.10.bin 4 kB · 0 downloads

 

I got a CRC32 of: 8CD1672E

 

You should probably verify that this matches with your original file.

 

You can use a hash tool like CHK to find out.

 

[burkhart87]

1 hour ago, MrFish said:

 

I got a CRC32 of: 8CD1672E

 

You should probably verify that this matches with your original file.

 

You can use a hash tool like CHK to find out.

 

I get 8CD1672E. And the software told me the file was an "Atari 8-bit cartridge image"  🙂

[burkhart87]

2 hours ago, MrFish said:

Thanks for posting the ROM.

 

Possibly.

 

I wrote the V 1.10 file from my drive (i.e. the one posted above) to a blanked M2732A and the drive works with the newly-written V 1.10 ROM.  Phew.

 

Next I'll try the same but with Mr. Martian's TRAK ROM:

 

[+Larry]

Mr. Martian did us a lot of good with his UltraSpeed roms for Indus, Rana (IIRC), and Trak.  Wish he would drop in and say "hello."

[+MrFish]

3 hours ago, burkhart87 said:

And the software told me the file was an "Atari 8-bit cartridge image"

It's just getting that information from the system based on the file extension and whatever program you have tied to the extension -- which is probably Altirra (based on the description you quoted).

 

[burkhart87]

Mr. Martian's TRAK ROM works in my drive, even better than the original V 1.10 at consistently reading DD formatted disks from other drives. 😃   I think I'll just leave it in the drive.

 

I erased and re-flashed M2732A's with V 1.11, 1.12 and 1.13.  Same results as before, the drive did not post.  

 

FWIW, it seems TRAK drives with Ver. B boards do not work with the later ROMs.  

Edited September 27, 2023 by burkhart87

[Mathy]

Hello Larry

 

On 9/26/2023 at 11:56 PM, Larry said:

Wish he would drop in and say "hello."

 

Maybe he'll show up if we do this: @MrMartian  🙂

 

Sincerely

 

Mathy

 

[MrMartian]

Or, a friend lets me know about this and nudges me to check in.....

[MrMartian]

On 9/27/2023 at 12:59 PM, burkhart87 said:

Mr. Martian's TRAK ROM works in my drive, even better than the original V 1.10 at consistently reading DD formatted disks from other drives. 😃   I think I'll just leave it in the drive.

 

I erased and re-flashed M2732A's with V 1.11, 1.12 and 1.13.  Same results as before, the drive did not post.  

 

FWIW, it seems TRAK drives with Ver. B boards do not work with the later ROMs.  

 

Plus, it gives you the ability to do 1050 Enhanced Density on the Trak...

Edited November 2, 2023 by MrMartian

[burkhart87]

On 11/2/2023 at 7:27 PM, MrMartian said:

 

Plus, it gives you the ability to do 1050 Enhanced Density on the Trak...

I may try that just to excersize an Atari 8-bit historical oddity.  All these years, programs, games, disks, etc. and I've never needed or used enhanced density.  😄

[burkhart87]

On 1/23/2017 at 7:24 AM, Larry said:

 

 

I've seen several threads here that describe the "memory tricks" used to store sector data for R/W. And those remind me of another one -- the TRAK with its track buffer. (I had an AT-D2 that worked, and think I still have the board from one.) IIRC, it had a 2K memory module for the track buffer. No problem in SD, but on the surface, it is a bit shy of enough to hold the full track's data in the buffer, so it must have held at least one sector's data somewhere else in the system. (?)

 

 

 

On 1/23/2017 at 7:58 AM, evilmoo said:

According to this: http://www.strotmann.de/~cas/Infothek/TRAK_AT/TRAK-AT.pdf

The 4K (not 2K as you remember) printer buffer was also used as a sector cache. It could also be upgrated to 16K.

On 1/23/2017 at 8:54 AM, Larry said:

 

 

Here is an article from Antic that also mentions the 2K buffer that could be upgraded to 4K, so there could have been a change along the way to upgrade the models to 4K (as the docs you linked indicated).

 

http://www.atarimagazines.com/v3n4/diskdrivesurvey.html

 

It would be cool if someone else who had used the TRAK could weigh in on it, also.

 

-Larry

Yes, some drives must have been shipped with a 2K buffer, i.e. one 2K RAM chip with the other socket empty, although the documentation available indicates all drives had 4K.  See the pics of my drive above, w/ one M58725P 2K chip and the other socket empty.  

All of the other pics I can find of Trak boards show 4K.  This got me wondering if the "later" ROMs in the 4K boards require the 4K, as none of the later ROMs (Vers. 1.11, 1.12 and 1.13) worked in my drive, which had the V 1.10 ROM.  So, I put another M58725P in my drive with the V 1.13 ROM, and presto, a working drive.

 

V 1.13 recognizes DD disks, then switches from SD to DD when booting, faster than the other ROMs, including Mr. Martians.  Other than that, I didn't notice much of a difference, but did not do any stringent testing.  

 

If I return the drive to 2K, no bueno as before.  Both density LEDs light up at the same time and the drive does not post.

I didn't try the V 1.11 or 1.12 ROMs with the 4K.  

[manterola]

According to an image of an addendum the switches were assigned to different functions depending on the firmware version.

Supposedly, you can disable/enable the printer port, and also disable/enable the turbo mode, using the switches.

Since SW1 and SW2 are used to select the Atari Drive number, I assume that SW3 and SW4 should be used in newer firmware for Printer port and Turbo mode, respectively. That means that you cannot longer disable the auto diagnostic (as mentioned in the manual).

I have not tested any of this, yet. So, I am not certain that this is true, and which is the version that implemented the mentioned changes.

[burkhart87]

 I couldn't find a pdf of the addendum anywhere, so I scanned mine.

Trak Addendum.pdf

 

It doesn't mention any of the switches controlling Turbo mode, only the diagnostics or printer port.

 

[+kheller2]

"If the Automatic Diagnostic routine encounters an error, it will flash the single/double density indicators. "   So, its encountering an error   Lack of RAM as you showed.

And apparently there is yet another addendum that talks about the turbo switch.  See previous posts in this thread of a quarter picture of it.

[manterola]

6 hours ago, burkhart87 said:

 I couldn't find a pdf of the addendum anywhere, so I scanned mine.

Trak Addendum.pdf 1.06 MB · 3 downloads

 

It doesn't mention any of the switches controlling Turbo mode, only the diagnostics or printer port.

 

This is very nice and interesting. Thanks for taking the time to scan your addendum, it is the only one I have seen so far (more about this in the next paragraph). Do you know the firmware for which this addendum you just uploaded is applicable to?

What happened is that I found a picture of the Trak Disk drive with addendum piece of paper folded twice, so you can read 1/4 of the actual addendum and from than you can deduce that the functioning of the switches changed to enable/disable Turbo Mode, and that the diagnostic setting in not controllable by the user anymore. So, thank you again for scanning an actual full page addendum.

This is the picture and post  I am talking about:

 

 

 

 

 

 

 

 

[manterola]

I did some test, and I can confirm what we thought from the 1/4 page addendum: switch 3 is to enable/disable the printer port, and switch 4 en/dis the "turbo mode". I found the turbo mode, not that fast, and without further information about it, I am just assuming that the slight observed increase in the beep-to-beep frequency is what I enabled/disabled.

 

[burkhart87]

On 11/13/2023 at 9:35 PM, manterola said:

Do you know the firmware for which this addendum you just uploaded is applicable to?

 

 

 

 

 

 

 

Yes, the addendum I posted came with my drive above (posts from Sept. 2023), Ver. 1.10 ROM.  This drive also has a unique (so far) board with some differences in the power supply and what I believe to be the ground rail when compared to the boards from "later" drives in this thread.

[burkhart87]

4 hours ago, manterola said:

I did some test, and I can confirm what we thought from the 1/4 page addendum: switch 3 is to enable/disable the printer port, and switch 4 en/dis the "turbo mode". I found the turbo mode, not that fast, and without further information about it, I am just assuming that the slight observed increase in the beep-to-beep frequency is what I enabled/disabled.

 

I think "Turbo mode" also needs the missing/vaporware "Turbo-charged" software to take advantage of the printer buffer (P. 5 of the manual).  P. 14 of the Trak manual talks about the software having to be purchased from "...your local Atari dealer." 😄  So it does not appear that this Turbo software was included on the Trak disk bundled with the drives (the software on these disks is also mostly still missing). 

 

 

The manual comprises some other BS.  As an example, the touted diagnostics are alleged to be capable of output of a "...status report on your screen or printer..." 😄, see P. 6.  In reality, the "diagnostics" are merely flashing LEDs that tell you its time to "...consult your local dealer" (addendum).

 

Don't misunderstand me, I love my Trak drive for nostalgic and performance reasons, but much of what the manual promises was not delivered in reality.