2732 EPROMs not working on Ralf David/Dropcheck Burner

[+Nezgar]

I ordered this re-imaged EPROM burner from bitsofthepast many months ago, and finally opened it up and remembered it didn't come with PIA's, so had to wait again for a pair of PIA's from ebay.

 

'Ralf David’s XL/XE Eprom Burner Bundle: https://www.bitsofthepast.com/?product=ralf-davids-xlxe-eprom-burner-bundle

 

I was inspired to finally give it a go when I had to re-program some EPROMs for mini-Speedy's I received from tf_hh which needed the newer NTSC compatible patched ROM. I burned 4 TI 27C256's with no problem and magically had working Speedy drives on an NTSC machine. I was quite proud of myself for burning my first EPROM myself

 

I was also able to successfully read a 16KB XL Mask Rom to file too..

 

Anyhow, to the question. One of the other things I would like to do is burn 2732's to make US Doubler, and stock 1050 OS ROM's... However, any time I attempt to read or write one of these, it seems to only process 256 bytes of the 4096. I confirmed it is actually reading the first 256 bytes correctly from a previously programmed USDoubler EPROM. (TMS 2732AJL-45). a SpartaDOS X compare of a known good USDoubler 4096byte ROM file and a file 'read' from the chip shows the first 256 bytes match. The first 256 bytes also seem to be all that it burns because it does not function in a drive after burning but does read the first 256 bytes back OK.

 

New chips i'm trying are: NMC27C32BQ, ST M2732A-2F1. Both behave the same.

 

I'm using the supplied 2732 Adapter, chip notch to the top, notch towards '3M' on the adapter. Adapter '3M' towards top when inserted in main board. Of course, I'm also using the 'updated' version of the software linked form bitsofthepast.com page that includes 2732A support...

 

Maybe these chips are not compatible with the board, but I would have hoped I would at least be able to read an EPROM currently working in a 1050.

 

So yeah, curious if anyone else has tried this. Cheers!

 

Edit: I hope it doesn't matter, but I installed 2 MC68B21 PIA's - the other option was 6520AP's

Edited January 20, 2018 by Nezgar

[_The Doctor__]

it does matter you need the N series I'll pm you

Edited January 20, 2018 by _The Doctor__

[Guest]

hi Doctor,

 

could you confirm which chips can be used for the 1050 (OS or USD) and which can be used for the 800xl

 

thanks

[Dropcheck]

I ordered this re-imaged EPROM burner from bitsofthepast many months ago, and finally opened it up and remembered it didn't come with PIA's, so had to wait again for a pair of PIA's from ebay.

 

'Ralf David’s XL/XE Eprom Burner Bundle: https://www.bitsofthepast.com/?product=ralf-davids-xlxe-eprom-burner-bundle

 

I was inspired to finally give it a go when I had to re-program some EPROMs for mini-Speedy's I received from tf_hh which needed the newer NTSC compatible patched ROM. I burned 4 TI 27C256's with no problem and magically had working Speedy drives on an NTSC machine. I was quite proud of myself for burning my first EPROM myself

 

I was also able to successfully read a 16KB XL Mask Rom to file too..

 

Anyhow, to the question. One of the other things I would like to do is burn 2732's to make US Doubler, and stock 1050 OS ROM's... However, any time I attempt to read or write one of these, it seems to only process 256 bytes of the 4096. I confirmed it is actually reading the first 256 bytes correctly from a previously programmed USDoubler EPROM. (TMS 2732AJL-45). a SpartaDOS X compare of a known good USDoubler 4096byte ROM file and a file 'read' from the chip shows the first 256 bytes match. The first 256 bytes also seem to be all that it burns because it does not function in a drive after burning but does read the first 256 bytes back OK.

 

New chips i'm trying are: NMC27C32BQ, ST M2732A-2F1. Both behave the same.

 

I'm using the supplied 2732 Adapter, chip notch to the top, notch towards '3M' on the adapter. Adapter '3M' towards top when inserted in main board. Of course, I'm also using the 'updated' version of the software linked form bitsofthepast.com page that includes 2732A support...

 

Maybe these chips are not compatible with the board, but I would have hoped I would at least be able to read an EPROM currently working in a 1050.

 

So yeah, curious if anyone else has tried this. Cheers!

 

Edit: I hope it doesn't matter, but I installed 2 MC68B21 PIA's - the other option was 6520AP's

 

Several things could be causing the issue.

 

I found in my testing that although you can select the switch settings for the correct eprom type using the T key, you must also manually enter the burn address for the 27C32 or 2732A. That is set by pressing the P key. Enter four zeros for the start address and then enter '0FFF' for the end address. This may also apply when reading the same size masked eproms.

 

These two eproms you list are programmed with different voltages. The 27C32 is usually programmed with 12.5V and the 2732A is programmed with 21V. Usually the programming voltage is stamped on the eprom. If the 27C32 is listing 25V, then it cannot be programmed using the burner.

 

Also the programming speed may need to be changed from FAST to SAFE(SAVE) especially for the 2732A.

 

Did you do an empty test first? If the eprom is not erasing completely, then of course it won't burn correctly either.

 

Based on the PIA part # you gave, you should have the correct PIAs installed. If you were able to burn other eproms, then the burner is working. However settings and the eproms themselves are now in control as to whether you can burn/read them. Sadly as these eproms age, they begin to break down, become unreliable and finally stop working altogether. Remember they are not being made any more by reputable chip manufactures. The 'new' 'refurbished' eproms you get off of Ebay and the like could in fact be mislabeled counterfeits.

 

Also one final thing, Which version is this burner?

Edited January 20, 2018 by Dropcheck

[_The Doctor__]

OS ROM chip in the Atari XL/XE series requires a compatible 27128 for the most part

also be aware on some of these new old stock and reclaimed eproms a proper erase can take an hour or two!

 

looks like lenore had the data ready to go, I was pulling up data sheets... she's already posted the info

 

If my noggin is pulling thoughts properly

ST M2732A-2F1 worked just fine as a USD prom...

I didn't like the low write voltage on the other prom... but that's just me

Edited January 20, 2018 by _The Doctor__

[Guest]

cheers, really helpful

[+Nezgar]

Thanks for the info!

 

I will try manually entering the burn address 0FFF, I didn't pay attention to that before. Didn't have to for the 27C128 or 27C256.

 

Both the 2732A and 27C32 did pass blank check, after my 2 programming attempts and 5 mins in the eraser, but might only be checking 1st 256 bytes)

 

Previously when programming the speedy 27C256's i had a couple errors using FAST mode, and had better success with SAVE mode, so I have generally been preferring that mode.

 

Yes the 2732A required 25V, but I was able to adjust the 21V pot to 25V and I think it performed the same (first 256 bytes successful) I have check my notes again when back at the Atari.

 

I don't see an explicit version on this board, maybe just the year?. "XL/XE Burner" "c 1989 Ralf David" and "Reimaged by Dropcheck 2016 Sanctioned by ABBUC"

 

Side note:

 

I have 1 12.5V 27C128 I've been trying to use that tests blank (all FF) after 5 mins in the eraser, but always fails on a burn with dozens of error locations that won't set bits to 0. I presume EPROM is bad, I presume extra time in the eraser won't help with this.

[Dropcheck]

Thanks for the info!

 

I will try manually entering the burn address 0FFF, I didn't pay attention to that before. Didn't have to for the 27C128 or 27C256.

 

Both the 2732A and 27C32 did pass blank check, after my 2 programming attempts and 5 mins in the eraser, but might only be checking 1st 256 bytes)

 

Previously when programming the speedy 27C256's i had a couple errors using FAST mode, and had better success with SAVE mode, so I have generally been preferring that mode.

 

Yes the 2732A required 25V, but I was able to adjust the 21V pot to 25V and I think it performed the same (first 256 bytes successful) I have check my notes again when back at the Atari.

 

I don't see an explicit version on this board, maybe just the year?. "XL/XE Burner" "c 1989 Ralf David" and "Reimaged by Dropcheck 2016 Sanctioned by ABBUC"

 

Side note:

 

I have 1 12.5V 27C128 I've been trying to use that tests blank (all FF) after 5 mins in the eraser, but always fails on a burn with dozens of error locations that won't set bits to 0. I presume EPROM is bad, I presume extra time in the eraser won't help with this.

 

The 2732A is 21V programming voltage. A 2732 requires 25V and sometimes more. They are different eproms.

 

I don't know of any legit 2732A with a programming voltage of more than 21V.

 

I usually give at least 15-30 min in the eraser. More than that and the eprom is on it's way out of life.

Edited January 20, 2018 by Dropcheck

[guus.assmann]

Hallo,

Also very important for erasing Eproms, make sure the pins are all electrically connected.

If not, the Eprom may be destroyed and this show in failure to program.

It will show erased but won't accept any programming.

BR/

Guus

[+Nezgar]

Success! I programmed two of the ST M2732A's (21.0V) with the 4KB USDoubler code.

 

But man, these are way more finicky than the larger ones. Basically 2 additional steps I needed to do to succeed:

- Make sure to choose <P> Destination Params and manually enter $0000,0FFF even though that's whats already showing up top.

- Block out ambient light from the EPROM window during programming! I found that with my LED room lights on, the programming would move much slower, and produce continuous errors every few hundred bytes. Just covering it with my finger when I discovered this showed an immediate improvement in speed.

 

Observation... The program generates a lot of colours in 2732 mode more than the other modes, like a decompress to start, sometimes a red or green background for the first 256 bytes, etc. Not sure what it means, but it works...

 

I then took the new EPROM and swapped it out of my currently working US Doubler drive and it worked! I then tried the 2nd EPROM and saw the flash of death emit from the EPROM window, and sure enough in my excitement I had plugged it in backwards. D'oh! so that chip is toast.

 

For the record regarding voltages, i the chips I've been trying - none of them are 25V as I had previously incorrectly thought. Just the NMC ones are 12.75V instead of 12V which require a small pot adjustment prior to burning,

 

NMC27C32B 12.75V

ST M2732A 21.0V

ST M27128A 12.0V

tms27c256 12.0V

Edited January 20, 2018 by Nezgar

[_The Doctor__]

yep I put them in sockets that have all the pins wired together and sit the in the uv for about 50 minutes... works just fine.... never an issue...

I always cover the windows during programming and leave them covered,

ST M2732A always a good choice , placing them in the socket backwards.. always a bad choice

um yeah parameters are important....

 

great to see your success!

Edited January 20, 2018 by _The Doctor__

[Dropcheck]

Success! I programmed two of the ST M2732A's (21.0V) with the 4KB USDoubler code.

 

But man, these are way more finicky than the larger ones. Basically 2 additional steps I needed to do to succeed:

- Make sure to choose <P> Destination Params and manually enter $0000,0FFF even though that's whats already showing up top.

- Block out ambient light from the EPROM window during programming! I found that with my LED room lights on, the programming would move much slower, and produce continuous errors every few hundred bytes. Just covering it with my finger when I discovered this showed an immediate improvement in speed.

 

Observation... The program generates a lot of colours in 2732 mode more than the other modes, like a decompress to start, sometimes a red or green background for the first 256 bytes, etc. Not sure what it means, but it works...

 

I then took the new EPROM and swapped it out of my currently working US Doubler drive and it worked! I then tried the 2nd EPROM and saw the flash of death emit from the EPROM window, and sure enough in my excitement I had plugged it in backwards. D'oh! so that chip is toast.

 

For the record regarding voltages, i the chips I've been trying - none of them are 25V as I had previously incorrectly thought. Just the NMC ones are 12.75V instead of 12V which require a small pot adjustment prior to burning,

 

NMC27C32B 12.75V

ST M2732A 21.0V

ST M27128A 12.0V

tms27c256 12.0V

 

I've toasted not a few eproms in my excitement as well.

 

But it's good that you were successful in programming them.

 

Yes the burner does get psychedelic on the smaller eproms. Not sure why. I think the need to manually enter the start/end address is a missed mov or sta in the program when the 2732A feature was added.

[+Nezgar]

Also very important for erasing Eproms, make sure the pins are all electrically connected.

If not, the Eprom may be destroyed and this show in failure to program.

It will show erased but won't accept any programming.

You mean short all the pins together during erasure? I don't think I've heard of anyone doing this before. I guess I could solder all the pins together on a socket.

 

Edit: oh sorry I think you just mean make sure good connectivity on all pins during programming.

Edited January 21, 2018 by Nezgar

[_The Doctor__]

Both!, Good connection practices during all steps as well as erasing old chips for extended time, the tied together pins protect in an electrical as well as slightly thermal way. All I know is I have had far less dead chips doing it this way for the past number of decades. it just works. He is offering his advice from experience, and I second that experience. Edited January 21, 2018 by _The Doctor__

[guus.assmann]

Hello Nezgar,

 

What I mean is that erasing may destroy an Eprom if the pins are insulated from each other.

So put the Eprom in a piece of conducting foam.

Alternatively, put some aluminum foil over a pieces of foam and stick the Eprom-pins in that.

Either way, the pins are electrically connected.

 

Cause: During discharge you put energy in the chip. (Light = energy)

Thus there can be (too larges) voltage differences inside the chip.

And that can be destructive.

I found this out when Eproms like the 2708 were about $150 a piece. (My boss was happy about this)

Modern Eproms are less sensitive, but still....

 

BR/

Guus

[_The Doctor__]

I laughingly called it the solar panel effect, where the chip cooks itself during erasure. It converts light to energy the chip gets warm, electricity is generated and it all goes south. I used sockets with all the pins tied together. He used tinfoil and conductive foam. It's basically the same solution... it's old school for the new school..

Edited January 21, 2018 by _The Doctor__