Datron CPU board PAL data (1271, 1281...)

[pelule]

Todd Micaleff created a great site for Datron 1281 DMM repair and calibration.
      https://xdevs.com/fix/d1281/
This site allso collects a lot of detailed design information - as usual for the xdevs teams.
Several proud 1281 onwers (like me) are supporting by their findings - but the info belwo may also of interest for others.

Background:
I faced a fault on one of my 1281 - I could located the damaged on digital assembly - a 8 MHz oscillator IC not oscillating.
That fault had some negative impact also to other digital parts on the board - but that's a different story.

The digital assembly (CPU board, 68k based) used at the 1281 is used in several Datron gears like 1271, 1281 and 4920.

During lifetime it had several revisions - I own two 1281 with different CPU boards, both with same Firmware (v3.12).
- the older one is using 2 PALs and many TTL ICs to handle the different adressings (Memory, I/O...).
   PAL1:      Uxxx, 14L8 [No 290161-1, Date 25-MAY-88], I have no "old" doc to verify the part reference
   PAL2:      Uyyy, 14L8 [No 290162-1, Date 29-MAY-88], I have no "old" doc to verify the part reference
- the newer is using 3 PALs to reduce part count.
   PAL1:      U107, 14L8 [No 400902-10, Date 24-MAY-89]
   PAL2:      U111, 14L8 [No 400904-10, Date 24-MAY-89]
   PAL3:      U110, 20L8 [No 400903-10, Date 24-MAY-89]

I don't own a PAL/GAL Programmer which is able to read 14L8 or 20L8. that products.
Both, 14L8 and 20L8, are PALs without latching - so read out is no big deal.

So I reactivated my last century DIY PC-Printer I/O expander (my CtrlLPT - yes the SW is based on MS-DOS).


/PeLuLe

[texaspyro]

PALs have a security fuse that can be blown that make reading the data out with a device programmer useless.

I built a device for reverse engineering combinatorial PALs (those without registers).  I've also used it on some simpler registered PALs.

It first tries to identify the PAL type by probing the pins for high/low/input/tristate combinations (or you can force the PAL type).  It then goes through all combinations of the input capable pins and builds a truth table of the outputs that is then fed to ESPRESSO to reduce the truth table to logic equations.

It uses a MegaDonkey Atmel 2561 based touchscreen micro-controller.  It's been years since I last used it. I built it for one specific task, but it has come in handy a few times.

[pelule]

Hi texaspyro

you absolutely correct - thats why I used the CtrlLPT Board (the same approch you described).
Beside documenting the PAL decoders my main target is to get the decoded adress / fidn the correct memory locations.
/PeLuLe

[rigrunner]

The read-out of the other PALs will follow, if ther is any interest at the community...

If you can read the others whilst you have the CtrlLPT running it will be worthwhile. One day a pal will die and whoever needs to burn another will be very pleased that you read the contents

[ManateeMafia]

Pelule,

Thanks for sharing all the information. It seems you are relentless to get every last detail of these great meters archived. Keep up the great work!   

I just received a 1271 that will be added to xDevs but it will be some time before I get to it.

[pelule]

Spend the free time during Xmax to finalize the Datron PAL readout & decoding.
What I did for each Datron PAL, I have access to:

• readout the original PAL (truth table file & CSV file for later comparison with Excel)
• manual decoding the logic equations and create a PLD for WinCUPL
• compiling PLD file with WinCUPL to a JEDEC file
• programming a GAL device with MiniPro TL866 (v6.60)
• readout the programmed GAL (truth table file & CSV file for later comparison with Excel)
• compare PAL-CVS and GAL-CVS with Excel

I have access to the following Datron PAL-ICs:

• 1281 - CPU brd with 2 PALs "290161-1 25-5-88" and "290162-1 25-5-88"
• 1281 - CPU brd with 3 PALs 400902-10 24-MAY-89", "400903-10 24-MAY-89" and "400904-10 25-MAY-89"
• 4950 - CPU brd with 3 PALs 400902-1 29-MAR-96", "400903-1 22-APR-96" and "400904-1 19-APR-96"

All JEDEC & PLD files and yet available information (schematics, picture) are attached as zipped files.

/PeLuLe

[pelule]

3rd ZIP-File

[martinator]

What does this mean?
 • manual decoding the logic equations and create a PLD for WinCUPL

You took the jed file and made equations from it?

Palasm can make equations from a Pal jed file but you already knew this? (doesn't always work)

[pelule]

No. That would be a bit silly as I have a tool "JED2EQN" for.
As I have the PAL ICs only (read out protected)
I logged the thruth table, made manually equations out of it, put equations into a PLD file and compiled it to a JEDEC file.

/PeLuLe

[MadTux]

How did you generate the logic equations? Binary counter with logic analyzer?

I've a similar problem with an old TIG welder that has 5PAL PLDs (4x 16L8 and 1x20L8) inside and I'm planning to dump the data before they go bad eventually. I thought about using a 20bit counter (5x 74HC193) and logging the data on HP16500C/16557D. But maybe there is a better way of doing it. There is a combinatorial PAL dumper project here:
 https://web.archive.org/web/20140318183438/http://cgfm2.emuviews.com/pal/index.html
If someone owns that thing, I'd be highly interested in borrowing it for a while.

[pelule]

How did you generate the logic equations? Binary counter with logic analyzer?

If you scroll this page up, you find the short description of my "old" hardware, using the printer port to read out the truth table.
You many find that information useful also:
http://www.vcfed.org/forum/entry.php?315-Cloning-a-PAL-HAL-%28Part-2%29&goto=prev
/PeLuLe