Anyone know how to burn a single standard OS ROM?

[+Nezgar]

Actually, just remembering the issue I had last month after socketing a couple machines - I had leftover flux on the board that apparently has just enough conductivity to mess things up. Never showed up in continuity testing. Make sure flux is all cleaned off with isopropyl alcohol or electronics cleaner and qtip/toothbrush.

[xrbrevin]

erk that sounds a bit dodgy!

 

anyhoo, it was the MMU - the bad one from the 65xe caused the same brown screen when plugged into the 600xl and conversely, the 65xe works normally with the MMU from the 600xl!

 

interesting to find out that sys check II reports OS ROM failure when only the MMU has failed

 

i read that a pal 16L8 can be made into an MMU but my minipro TL866 only lists GAL devices.... i have a PAL16L8 and it appears to burn when i set it to GAL16V8 but only if i untick verify! sadly the machine just does the brown screen fail

 

are you aware of a workaround to burn the PAL chip?

Edited June 22, 2018 by xrbrevin

[xrbrevin]

The TMS27C128 is working after a heavy UV and re-burn - it has the hi speed SIO / reverse option / hard reset OS ROM running nicely now

[flashjazzcat]

interesting to find out that sys check II reports OS ROM failure when only the MMU has failed

Not surprising when you consider that the MMU controls PORTB bit 0 (OS ROM enable) and OS ROM chip select.

[xrbrevin]

its not a big deal, just an observation. i never knew there was a link between the OS and MMU until now.

 

update for my post above - the TMS 27C128 has just perished again. the computer boots straight to self memory test and now both ROM banks are red!

time to bin that chip methinks..

[1050]

Depends, might be normal behavior for a fast eprom

and thus it's still good.

 

For OS work, I've found that one wants to use a

200ns or slower eprom, faster ones will act wonky

when they do fine as a cartridge rom. Can't fathom

why either, just reporting how it is here.

[xrbrevin]

thanks for the info, they can be fickle these things..

thus far the Atmel chip is working ok in a 130xe so I will get some more, they are only £1 each new on the fleabay and they are electrically erasable instead of UV

 

i like the HISIO with cold reset and reverse option basic but also like to keep the stock XL/XE too so its a good solution

 

ive given up on making an MMU because the i couldnt get the .jed files to work - the GAL seemed to program ok but the computer did not boot. I found an MMU from a spares supplier for reasonable money instead

Edited June 25, 2018 by xrbrevin

[ACML]

Depends, might be normal behavior for a fast eprom

and thus it's still good.

 

For OS work, I've found that one wants to use a

200ns or slower eprom, faster ones will act wonky

when they do fine as a cartridge rom. Can't fathom

why either, just reporting how it is here.

Interesting. I have been using 120ns EPROMS and that might explain the pickiness I've found with 1200XL OS mods. Back in the day, the EPROMS were usually 250 or 300 ns.

[+Nezgar]

I just remembered about this drop-in 4-in-1 OS adapter from dropcheck. Burn 4 OS's into a 27C512 (or compatible EEPROM i guess too) and switch between the four:

https://www.bitsofthepast.com/?product=4in1-os-adapter

 

Nicer than soldering directly to chip pins for the multi-OS selection, and easier to re-program the chip again later too.

[AMenard]

How to burn a single standard os rom:

 

1- light fire

2- throw single standard os rom in it

 

One standard os rom burned coming up!

[+DrVenkman]

Bumping this thread because it sort of tangentially discusses a few things I'm trying to figure out without addressing them directly, and most of the people who are likely to know the answer were taking part in this one ...

 

So, background: I just got a TL866-II Plus to program (modern) stuff I need for some some other projects, and in learning how to use it, I decided to try to dump some more or less common Atari ROMs just to see how to use the thing, as well as testing a bunch of 4000-series and 74-series logic chips I have. I was able to successfully dump a 4K Tandom EPROM that I pulled from a dead 1050, and then dump a standard 16K OS ROM which came from an early production 130XE that I cannibalized for parts a couple years ago. When I pointed Altirra at my OS ROM dump, it ID'd it as an XL/XE rev 2 ROM, which at least confirms to device can read a good ROM.  

 

Then for giggles I decided to see if I could dump some 1200XL ROM chips I have but here's where I hit a snag: the mask ROMs I have are the usual sets of 2 x 8K DIP24 chips. There's nothing in the TL866 database for 8K DIP24 ROMs. I can lie and tell the TL866 it's a 4K ROM and it dutifully spits out 4K of code, including a legible snippet here or there (e.g., "BOOT ERROR" somewhere in there. 

 

Assuming I wanted to get a copy of these ROMs instead of just using .BINs available all over the place online, is there some kind of adapter board I need to McGyver up? Or is it just not possible to read an 8K DIP24 chip? 

[ACML]

I have a TL866, but haven't tried to read a 1200XL ROM.  You could trick it, but it would take two reads.  Read the first 4K then tie the high address bit high to Vcc and read the second block of 4K.  Append the two files to a single 8K.  BITD I did the reverse for burning on a Thompson Proburner.  I had 27256 and the Proburner could only do 27128.  So I burned the first 16K, tied the high address pin (26) high to Vcc and burned the second 16K block.  It was a kludge, but it worked.

 

[BillC]

5 hours ago, DrVenkman said:

Then for giggles I decided to see if I could dump some 1200XL ROM chips I have but here's where I hit a snag: the mask ROMs I have are the usual sets of 2 x 8K DIP24 chips. There's nothing in the TL866 database for 8K DIP24 ROMs.

I believe the 8k mask ROM chips Atari used for the 1200XL OS and the 16k cartridges are 2364 chips.

https://ia803003.us.archive.org/20/items/mos_2364_rom_feb_1980/mos_2364_rom_feb_1980.pdf

 

If this is the case it should be possible to dump their contents using another Atari, a compatible 16k cartridge, and the binary save function of DOS. Since these ROMs can have different chip enable settings it may be necessary to try dumping each chip in both of the cartridge board sockets.

 

 

 

 

 

[Ralphy Rocket]

3 minutes ago, BillC said:

I believe the 8k mask ROM chips Atari used for the 1200XL OS and the 16k cartridges are 2364 chips.

https://ia803003.us.archive.org/20/items/mos_2364_rom_feb_1980/mos_2364_rom_feb_1980.pdf

 

If this is the case it should be possible to dump their contents using another Atari, a compatible 16k cartridge, and the binary save function of DOS. Since these ROMs can have different chip enable settings it may be necessary to try dumping each chip in both of the cartridge board sockets.

 

 

 

 

 

True!! I remember using a basic program to make a 1200xl OS to dump to disk..if the 1200xl is working of course...

 

[+Nezgar]

5 hours ago, DrVenkman said:

was able to successfully dump a 4K Tandom EPROM

Tandon

 

1050 EPROMs are 2732, which is directly supported by the TL866. Jumper settings on the motherboard can switch the pinout and invert a chip select pin for 2332 configuration, which you can probably read in a TL866.

 

5 hours ago, DrVenkman said:

I decided to see if I could dump some 1200XL ROM chips I have but here's where I hit a snag: the mask ROMs I have are the usual sets of 2 x 8K DIP24 chips. There's nothing in the TL866 database for 8K DIP24 ROMs. I can lie and tell the TL866 it's a 4K ROM and it dutifully spits out 4K of code, including a legible snippet here or there (e.g., "BOOT ERROR" somewhere in there. 

The 1200XL ROM's, and the BASIC ROM in all the rest of the XL/XE's are indeed 2364. This is a common 8K chip found on Atari cartridge mask ROM's. In Atari's 16KB carts, there were two 2364 mask ROM's. Both sockets on the PCB were wired identically except for Pin 20 would be mapped to S4 or S5. Atari 8K carts with 2 4K 2332's would be manufactured with opposite function on one of the chip enables (can't remember pin 20/21 combo)

 

6 hours ago, DrVenkman said:

Assuming I wanted to get a copy of these ROMs instead of just using .BINs available all over the place online, is there some kind of adapter board I need to McGyver up? Or is it just not possible to read an 8K DIP24 chip? 

There are lots of adapters to to convert 2764's to 2364's, but not so much the other way around (although should be easy, it should be as simple as remapping 3 pins). Nice work getting the TL866 to dump at least 4K of it though! That means you can could theoretically dump the "other" half of the ROM by the McGyver method by manually applying the opposite signal of whatever the TL866 is normally sending to the highest address line A12 / pin 21.

 

Adapters aren't unheard of for the TL866... The hobby community appears strong with this one. ie I got an adapter to remap pins to program 2532's in 2732 mode, and 27c160/322/400/800's one bank at a time which are normally larger than the TL866 supports. Used it to program a 4x512KB Amiga OS EPROM for a friend.

 

I've found the easiest way to dump 2364's so far without a dedicated adapter is to simply put them into a socketed 8K or 16K Atari cartridge PCB (one intended for mask ROM's, not EPROMs) ie BASIC or Asteroids, and then dump it out using DOS save memory command, as @BillC already mentioned. If it's not an actual game ROM, the Atari won't recognize the "magic bytes" and won't try to automatically execute it. I've dumped 810 2316, Percom 2516, 1050 2332, and 850 2332's this way. Save the whole 16KB cartridge area, and then crop it down in a hex editor after, if you're not sure where it will land in the address space.

[BillC]

7 hours ago, DrVenkman said:

Then for giggles I decided to see if I could dump some 1200XL ROM chips I have but here's where I hit a snag: the mask ROMs I have are the usual sets of 2 x 8K DIP24 chips.

 

1 hour ago, Ralphy Rocket said:

True!! I remember using a basic program to make a 1200xl OS to dump to disk..if the 1200xl is working of course...

Ralphy's method is much simpler, just have the 1200XL read its own OS and save to disk. The thread linked below has 2 methods to do this, a simple BASIC program that does it byte by byte and a ML program. Both methods would result in a single 16kB file, it would be easy to modify the BASIC code if desired.

 

[+DrVenkman]

Thanks, guys! Lots to learn and digest here so I appreciate the pointers!

 

As to this:

3 hours ago, BillC said:

Ralphy's method is much simpler, just have the 1200XL read its own OS and save to disk

Well, this particular 1200XL now has a U1MB in it and is modded to allow for a single-chip 16K 800XL ROM. I’d have to dump another one of my 1200XL’s ROM chip to do this, rather than just figuring out a way to read this pair.  But in any case, I do appreciate your knowledge and suggestions.