32 in 1 OS from AtariMax

[tjlazer]

 

I'm not sure how possible it is. why?

 

Simply just for testing purposes.

[tjlazer]

AAAarrrggghhhh! Anyone else have problems trying to program these damn Flash chips? AM29F040? I am having trouble with one of mine. I went to erase it and it enly went to 13% and said finished. Erase chip produces "erase completed byDQ7" instantly. Then when I try to program, it gets to 49% and stops. I spend 2 hours trying to get this to work. I ended up ording 3 new chips from eBay to test next. I did see there is a way to do it with an Atarimax 8Mbit cart by doubing the ROM and programming the two chips on the cart. Might try that next....

[+Nezgar]

AAAarrrggghhhh! Anyone else have problems trying to program these damn Flash chips? AM29F040? I am having trouble with one of mine.

Mine are the same as yours. I have only 'read' the chip that came with the unit, and have only programmed the extra I got from eBay. What programmer are you using? With my TL866, the eBay chip was reporting a problem with the chip ID, that I didn't get with the original. Who knows with eBay parts if they're counterfeit... but I was able to erase and program it with the ID check disabled using the image read from the original chip without a problem.

 

Edit: Adding screenshot from my MiniPro software, showing IC selection, and ChipID after reading the AM29F040B-150JC that came with my 32-in-1.

[_The Doctor__]

generally if it fails in the stand alone programmer it will fail in the cart as well...using the cart with the flasher will do the erase / verify and then will fail. oddly sometimes the chip will fail in one and not the other, allowing a full erase on the other afterwards...

 

What I discovered... one of my standard programmers voltages was no longer clean (looked like some static on the scope).. this caused glitchy writes... so the Atari programmed the chip fine in the Atarimax cart... I replaced the power cord on the standard programmer and it was fine after that..

 

the other standard programmer failure was timing.... an erase worked, but a write with and erase failed... I was pushing data too fast and had to slow down the serial port speed...

again the atarimax card worked fine...

 

it is possible the chip is bad...

 

what programmer are you using? how is it powered? Try ID check disabled on generic chips...

 

can't hurt to use the Atari....

The only times times I had an issue using Atarimax carts were dirty card fingers and a bad flash chip (you could see a crack in the chip case...

 

DQ7 status as well as DQ6 and DQ2 status bit methods are fine for the 040b

I vaguely remember some talk of 040 chips and a different version of the same being an issue.

 

I say go for it and flash a game to the chip and see if it erases verifies and writes without issue using the MaxFlash cart.

Edited November 21, 2018 by _The Doctor__

[_The Doctor__]

http://robotics.ee.uwa.edu.au/eyebot/doc/DataSheets/29F040.pdf

I think the data sheet shows to 150

where as the earlier 040b shows to 90

http://www.mouser.com/ds/2/100/spansion%20inc_am29f040b_eol_21445e8-933839.pdf

a later data sheet shows an update

http://instrumentation.obs.carnegiescience.edu/ccd/parts/AM29F040B.pdf

 

Just for completeness sake. You can can double check your programmers settings against the sheet.

Edited November 21, 2018 by _The Doctor__

[+Nezgar]

AAAarrrggghhhh! Anyone else have problems trying to program these damn Flash chips? AM29F040? I am having trouble with one of mine. I went to erase it and it enly went to 13% and said finished. Erase chip produces "erase completed byDQ7" instantly. Then when I try to program, it gets to 49% and stops. I spend 2 hours trying to get this to work. I ended up ording 3 new chips from eBay to test next. I did see there is a way to do it with an Atarimax 8Mbit cart by doubing the ROM and programming the two chips on the cart. Might try that next....

 

Another thought... these chips are divided into 8 'sectors' of 64KB each, which can individually be erased, or write/programming protected. At the bottom of the screenshot in my last post, you can see this status along the bottom right row. 64K is 12.5% of 512K, maybe your sector 2 is protected/locked? (or others)

[_The Doctor__]

That would be awefully strange. because 12 percent would correlate with sector 1 locked (not 2), as that is 12.5 percent rounded up to 13...... but he is also stating 49 percent... unless writing sectors from 7 counting to down to 0 and locked 1 thru 3.

are we erasing opposite direction of our write?

Edited November 21, 2018 by _The Doctor__

[tjlazer]

Willem PCB6.0 and also the older PCB3. Was using USB power, but then tried with a 9V A/C. I was successful back in 2007 with the PCB3 when I reprogrammed my 32-in-1, but I tried it with my new PCB6.0 first, and then with the older PCB3. Might of messed something up as I as messing with jumpers, etc. Can you damage these chips?

[Guest]

Mine are the same as yours. I have only 'read' the chip that came with the unit, and have only programmed the extra I got from eBay. What programmer are you using? With my TL866, the eBay chip was reporting a problem with the chip ID, that I didn't get with the original. Who knows with eBay parts if they're counterfeit... but I was able to erase and program it with the ID check disabled using the image read from the original chip without a problem.

 

fingers crossed you make some progress with this.

but I have a hunch the proprietary nature of some of these 32-in-1 will see temperamental reactions to some PLCC chips.

Edited November 21, 2018 by Guest

[+tf_hh]

AAAarrrggghhhh! Anyone else have problems trying to program these damn Flash chips? AM29F040? I am having trouble with one of mine. I went to erase it and it enly went to 13% and said finished. Erase chip produces "erase completed byDQ7" instantly. Then when I try to program, it gets to 49% and stops. I spend 2 hours trying to get this to work. I ended up ording 3 new chips from eBay to test next. I did see there is a way to do it with an Atarimax 8Mbit cart by doubing the ROM and programming the two chips on the cart. Might try that next....

 

Just one hint... do you try it with the right version of the 29F040? AMD offers to variants... AM29F040 and AM29F040B - the "B" is very important. The both versions differ in sector sizes, protection schemes and some other small details. IMHO the "B" version is more common in retro-stuff, but check it.

[+Nezgar]

My chips are:

• AM29F040B-120JC V - Acquired via eBay. Erases and programs OK in TL866 when ID check is ignored. 'Mostly' works in the Atari, but unstable, and I could see some pixels 'shimmering' on Gr.0 text, which was my hint at a memory timing issue. I presume this is because 120ns access time is faster than the original 150ns, and is too fast for the Atari...
• AM29F040B-150JC - Original chip from AtariMax. Works perfectly after erasing & programming custom ROM image with TL866.

I used the AtariMax APE Warp+ OS Builder tool to take my first stab at making my own custom 32-in-1 ROM with the following highlights:

• 1200XL Rev 11 is no longer incorrectly named Rev 5
• Edited all the XL/XE revisions for less confusing titles.. ie "XL/XE Rev.B,RB" instead of "XL/XE R-B,RB", and "PAL" instead of "P"
• Option P. XL/XE OS Rev.2 with Highspeed SIO patch V1.30 by Matthias Reichl, Reverse BASIC, shift-reset coldstart, fast math pack, corrected checksums for self test ROM check - from this post: http://atariage.com/forums/topic/206880-130xe-reverse-option-key-for-basic/page-2?do=findComment&comment=3944284
• Option 2. Shoestring's Atari 8bit RAM tester v1.5 - as per this thread: http://atariage.com/forums/topic/282022-shoestrings-atari-8bit-ram-tester/?p=4103968

And some others to fill the slots:

• Option Y. OldRunner (FAQ says by Irata Verlag, 1985) - not really familiar with this one
• Option Z. CSS XL Fix
• Option 0. Omniview XE (80 Column Eleanor: handler)
• Option 3. Ramrod XL - newell fast math routines
• Option 4. Omnimon XL (ROM resident monitor)

Maybe next I can figure out where the background colour byte is in the config OS to patch, and maybe even patch the font "just because"

 

My ROM image is attached.

32-in-1 Nezgar1.zip

[tjlazer]

 

Just one hint... do you try it with the right version of the 29F040? AMD offers to variants... AM29F040 and AM29F040B - the "B" is very important. The both versions differ in sector sizes, protection schemes and some other small details. IMHO the "B" version is more common in retro-stuff, but check it.

Yeah it’s the B chip. It’s actually from a Sun-Mark mega cart that I am having this problem on, and now I’m scared to try to reprogram my 32 in 1 OS chip.

 

The chips I bought is are AM29F040-90JC V

Edited November 21, 2018 by tjlazer

[_The Doctor__]

Willem PCB6.0 and also the older PCB3. Was using USB power, but then tried with a 9V A/C. I was successful back in 2007 with the PCB3 when I reprogrammed my 32-in-1, but I tried it with my new PCB6.0 first, and then with the older PCB3. Might of messed something up as I as messing with jumpers, etc. Can you damage these chips?

Yes you can, but I've only seen that when quite over voltage(greater than more than half a volt for a bit of time).. or if and adapter is backwards, pins off and power high. I will admit I've done it and scared myself. Almost always it was okay.. I have only seen one become temperature sensitive as I have one (I still have to replace that chip) in that is like that... works great warm.. not so much cold.

 

like all chips, correct orientation, nomenclature, and speed matters. I did post the data sheets for both the 040 and 040b. Two are final sheets the earlier sheets can be slightly different.

 

China has been known to provide mislabeled chips from time to time just like some of those dodgy sd cards or any number of other chips.

 

since you have reprogrammed the chip before it more than likely will come down to settings.

 

just took a look at the chip in this one, 040B -120JC

Edited November 21, 2018 by _The Doctor__

[_The Doctor__]

does your 32 in 1 have the chips like in my picture?

if yes then you can use maxflash cart method,

or once you sort out you programmer settings,

it should work fine.

 

You can even take that amd 040b -150jc from the sun mark and put it in a maxflash and see if you can program it. ( provided your sunmark data is backed up somewhere of course )

[+Nezgar]

just took a look at the chip in this one, 040B -120JC

 

Interesting your -120JC V chip model is the same as the one I was having some instability with... if that's stable for you, then 120ns probably isn't the issue, but moreso the fact I got it from china eBay (ie counterfiet/defective)

[Guest]

thanks to those who've actually contributed/helped in this thread - at least they outweigh the aggressors.

wondering if this place needs a relaunch as AtariRage

[phaeron]

I reversed Nezgar's dump and there isn't actually much that the 32-in-1 does that needs to be emulated. Besides its flash being split into 32 x 16KB ROM slots, the only other functionality added is sensing for the Select button on power-on and the ability to read and write the selected OS index. All you need to emulate the menu is to extract the last 16KB of the image and map it into any emulator as a standard XL/XE OS ROM. The menu slot OS is just a modified OS-B with the menu code replacing the math pack and spliced into the cartridge init code, so it should be compatible with just about any emulator. The only issue is that the menu will report an error due to not being able to read the current OS selection.

 

Unlike the Ultimate1MB where the BIOS takes control first and can configure the hardware before booting the OS, the 32-in-1 doesn't run the menu OS first, it boots directly into the selected OS in hardware. There are only three additional connections to the 32-in-1 besides the OS ROM socket: reset line, Select button, and PORTB bit 7. The Select button is thus sensed in hardware, though Option is sensed in software for the APE control menu. PB7 is used for an unusual signaling mechanism where the current OS selection is transmitted serially at very low speed over the self-test bit (!). On write, the selected OS is changed and the computer is automatically reset. There unfortunately isn't much room to enhance the menu OS as the hardware capabilities are minimal.

 

I find it unlikely that flash compatibility would be an issue as there isn't anything in the menu OS that is flash specific, and flash control mode probably can't even be entered to query or write to the flash as there is no R/W line going to the 32-in-1. OTOH, there is a possibility for a CPLD compatibility issue with the serial protocol, though there isn't much in the protocol to change. Regarding flashjazzcat's ROM dump, it appears to be corrupt due to a bad A8 line connection as every page is duplicated in the dump -- thus the confusion about the missing menu entries.

 

[Kyle22]

I had a 32 in 1, it did not work at all in a 1200XL I had. It worked in another machine, but I wouldn't recommend it to anyone.

The U1M is so much more flexible and reliable.

 

32 in 1 sucks.

[_The Doctor__]

That's 3 votes for corrupt image

[BillC]

I don't have an AtariMax 32-in-1 upgrade so was unaware that it used "Ramrod XL" as a menu entry.

 

The Newell Ramrod XL is the PCB that allows switching between 3 different ROM/EPROM chips, I have an original that I ordered directly form Newell BITD, the name for the standard Newell XL OS ROM was "OSNXL".

 

The Ramrod XL usually came with an OSNXL chip which may be why the wrong name has been used, I paid extra to get the Omnimon XL ROM instead.

Edited November 22, 2018 by BillC

[_The Doctor__]

Be better if you put the 'Corvus Integrator Rom' on this please.... hen's teeth here...

[+Nezgar]

I don't have an AtariMax 32-in-1 upgrade so was unaware that it used "Ramrod XL" as a menu entry.

 

It doesn't come with that entry - the last few OS's were ones I picked myself when I built my own custom ROM.

 

Previous posts #9, #35, and #39 show various OS selections seen pre-programmed from AtariMax.

 

The Newell Ramrod XL is the PCB that allows switching between 3 different ROM/EPROM chips, I have an original that I ordered directly form Newell BITD, the name for the standard Newell XL OS ROM was "OSNXL".

 

The Ramrod XL usually came with an OSNXL chip which may be why the wrong name has been used, I paid extra to get the Omnimon XL ROM instead.

 

You're probably right it should be called OSNXL.. I named it that since that was what the ROMs filename I picked. Also didnt help that "RAMROD" was the label my dad used for that one of the homebrew 4 switchable OS's he installed in my 130XE BITD...

[Kyle22]

I have been looking for The Integrator ROM forever!

 

I think it was David Small that did it.

 

Anyone know where to find it?

 

It was an 800 replacement OS that talks to the Corvus hard drive over the joy ports. It allows BOOTING from Corvus instead of needing to boot a floppy first.

 

 

Edit: Its proper name is ADS Integrator.

Edited November 22, 2018 by Kyle22

[tjlazer]

Anyone have the RamRod XL and A8Diag ROM? I found a RAMROD.ROM, not sure if that is the same one.

[_The Doctor__]

They are in nezgars 32 in 1 image.