Gauging interest in Atari based Eprom Burner

[Dropcheck]

I would like to get one in November. and I have 6520 PIA chips that I have removed off old Broken Atari Computers.

 

Steve

Http://www.realdos.net

 

I hope to keep a running inventory in stock. It might not be many, and from time to time might be a delay in ordering parts and pcbs. Shouldn't be a problem. Just visit my site when you are ready and order.

[Dropcheck]

Update:

 

2nd run is nearing completion. Testing will commence in a few days. Shipment should be by the end of Oct or sooner.

[chevymad]

Received the PIA's from chinabay today. They're Conexant R6520APs. Don't have the burner yet to test but they look new.

[Dropcheck]

Received the PIA's from chinabay today. They're Conexant R6520APs. Don't have the burner yet to test but they look new.

 

Hi

 

Your board went out yesterday. Let me know how the chips work for you.

[chevymad]

Thanks, I'm looking forward to trying it out!

[chevymad]

Came in the mail today. Looks like my pia's work. Though I don't have a good working eprom to burn and i'm trying to learn how to work this with my sdrive and ultimate. I do have a 2532 that I use in my car and an adapter for it to work as a 2732. Using that I can verify the chip is either empty or full, but I haven't figured out how to load the bin file from my pc to the atari yet. This chip also got a bit beat up and i've repaired a few legs.. so it may just be the chip too as it didn't burn correctly with the minipro either last time.

 

Guess now I need a project and some new eproms.

[chevymad]

Allright. I just need to figure out how to get my .bin file from my pc to the atari. I burned the 2532 on my minipro and then read it with the atari burner. I then erased the chip and burned it with the atari. So it works! Even with the 2532 and a stack of adapters.

[Dropcheck]

Allright. I just need to figure out how to get my .bin file from my pc to the atari. I burned the 2532 on my minipro and then read it with the atari burner. I then erased the chip and burned it with the atari. So it works! Even with the 2532 and a stack of adapters.

 

 

Sounds like a job for RespeQt or AtariMax's SIO2PC software and hardware bundle.

 

I'd be interested in the 2532 to 2732 adapter. Now the burner will do 2732A. 2732 is a different beast. Be very care of the manufacture too. Upto and including some 2764s manufactures did not always use the same pinouts or internal construction. Programming voltages were higher too. Currently the burner is tested to 21V, which is the two programming voltages available. If the 2532 requires 25, you are out of luck.

Edited October 22, 2016 by Dropcheck

[chevymad]

This is a 2532 and it will burn ok @ 21v. It's not quite right for the minipro either. This may be why I have occasional problems burning it. I have to use the adapter in my minipro as well. I found that info here on this board and a pacman website, there are 2 methods to the adapter. One is just crossing 2 pins, the other has a couple diodes. Right now I'm just using the cross method.

[chevymad]

Here's one of the threads where it's discussed. http://atariage.com/forums/topic/230688-tl866cs-minipro-eprommer/page-3and a quote of one of the posts.

 

You can program TMS2532s on the TL866CS. You have to put together a home-brew adapter. Get two 24-pin sockets and bend out pins 21, 20, and 18 on the one you will be using for the top socket. Piggyback the two sockets.

Top socket pin 21 (Vpp) connected to bottom socket pin 20.

Top socket pin 20 (-CE) connected to bottom socket 18

Top socket pin 18 (A11) connected to bottom socket pin 21

Add a ZIF socket on top of the adapter if you wish..

Select TI TMS2732A from the menu and program your chip. It works fine even though the voltage is 21V instead of 25V . I haven't had one to fail yet!!!

[chevymad]

Ok.. one success. Multiple failures with the 2532. Reads fine but won't always burn. Probably the voltage difference. I put my pc bin file into an ATR and it seems to want to use that file now. I'm getting code $64. Is that due to a burning issue or a file read issue? And lastly will a 2532A replace a 2532jl with no problems? The car forum that shows how to burn this chip says specifically not to use a 2532A.

 

None of this is actual problems with the burner. Just me trying to do something it might be able to do(or not) and refamiliarize myself with using an Atari at the same time.

 

Thanks for making the hardware Dropcheck!

[Dropcheck]

Important FYI:

 

It's been brought to my attention that there is a failure potential of the reimaged RD burner under a certain switch setting. T1 apparently heats up dramatically when the Atari is first booted and before the software finishes loading, if the switch settings are actually set for the 2732A type.

 

Simple solution is to always put both switch settings in the full up position before shutting down or starting up the Atari computer with the board connected.

 

So far I have not heard about or experienced an actual failure, but there is no reason to tempt fate.

 

 

On a happier note:

 

tf_hh has potentially found a fix for the MC68B21 PIA problem. He has done a prototype fix and I've sent for an adapter board from Itead to see if I can retrofit existing boards that have been sold without PIAs. Those boards will come in today I believe and I'll assemble a test to see if it is going to work. If it does, then all future boards will come with the fix already applied. All current boards sold without the PIAs will come with the adapter installed.

 

I am hoping that this will also open up using the R65C21C2 PIA as well. The more the merrier.

[GoodByteXL]

Important FYI:

 

It's been brought to my attention that there is a failure potential of the reimaged RD burner under a certain switch setting. T1 apparently heats up dramatically when the Atari is first booted and before the software finishes loading, if the switch settings are actually set for the 2732A type.

 

Simple solution is to always put both switch settings in the full up position before shutting down or starting up the Atari computer with the board connected.

 

So far I have not heard about or experienced an actual failure, but there is no reason to tempt fate.

 

 

On a happier note:

 

tf_hh has potentially found a fix for the MC68B21 PIA problem. He has done a prototype fix and I've sent for an adapter board from Itead to see if I can retrofit existing boards that have been sold without PIAs. Those boards will come in today I believe and I'll assemble a test to see if it is going to work. If it does, then all future boards will come with the fix already applied. All current boards sold without the PIAs will come with the adapter installed.

 

I am hoping that this will also open up using the R65C21C2 PIA as well. The more the merrier.

 

And also a need to know:

 

Additionally, there is a confirmed glitch in the software version 2.5: When doing read/verify of a 2732 EPROM just 256 bytes are processed and not the full 4096 bytes.

[Dropcheck]

 

And also a need to know:

 

Additionally, there is a confirmed glitch in the software version 2.5: When doing read/verify of a 2732 EPROM just 256 bytes are processed and not the full 4096 bytes.

 

 

Details are important. A general blanket statement is not justified. What is the specific manufacturer and exact part # of the eprom you have attempted a real burn, verify and read with? I want to be able to see if I can duplicate the issue.

 

I have personally verified that two manufactures of eproms do indeed burn, verify and read the full 4096 bytes. They are NMC27C32AB I believe made by National Semiconductor and D2732A-3 made again I believe by Intel. One is a CMOS and only requires 12.75V to program and the other is 21V for programming. If there was a fundamental flaw in the settings, software or adapter I would not have been able to really burn, verify and read with any eprom.

 

No one can say that all eproms will burn/verify/read without failure. In fact I will guarantee some will fail. This is a brute force programmer and as such if there is the slightest manufacture's variance in specs for a particular eprom type you run the risk of a burn/verify/read failure. There are no new eproms being made that I am aware of, particularly the 2732s. So what we are using is at best never used, but aged and at worst previously used and erased sometimes many times and still aged. Eproms have a life span of not only erasures and reprogramming but also simple age and can become flakey or fail sooner than that even, regardless if the eprom previously burn/verfied/read before..

[Dropcheck]

On a happier note:

 

tf_hh has potentially found a fix for the MC68B21 PIA problem. He has done a prototype fix and I've sent for an adapter board from Itead to see if I can retrofit existing boards that have been sold without PIAs. Those boards will come in today I believe and I'll assemble a test to see if it is going to work. If it does, then all future boards will come with the fix already applied. All current boards sold without the PIAs will come with the adapter installed.

 

I am hoping that this will also open up using the R65C21C2 PIA as well. The more the merrier.

The boards did come in today. I soldered one of them together and completed a full test run of all eproms the board is supposed to program using a pair of MC68B21 PIAs. The programming went off without a hitch. Wa....hoo!

 

Next I want to test with the pair of R65C21C2 chips I have to see if they too will now work. That work should be completed in the morning.

[GoodByteXL]

Details are important.

...

No one can say that all eproms will burn/verify/read without failure. In fact I will guarantee some will fail. This is a brute force programmer and as such if there is the slightest manufacture's variance in specs for a particular eprom type you run the risk of a burn/verify/read failure. There are no new eproms being made that I am aware of, particularly the 2732s. So what we are using is at best never used, but aged and at worst previously used and erased sometimes many times and still aged. Eproms have a life span of not only erasures and reprogramming but also simple age and can become flakey or fail sooner than that even, regardless if the eprom previously burn/verfied/read before..

Yes, and I sent you details in e-mails and described what is happening here and made a video to show you, what is not working. What more can I do to help to kill bugs?

 

But all you responded with was a simple denial of my statements instead of testing it by yourself (at least you didn't write in your e-mails that you had tested what I described). I am neither a programmer nor a hardware developer, just a long time experienced A8 user. And my goal was to help to wipe off the obvious bugs from that project. But I guess my approach was understood as pointless criticism.

 

In the meantime there is professional proof that I stumbled across bugs. With the help of tf_hh they are identified and described in the ABBUC forum. I recommend every user of that burner to study it thoroughly before trying to burn a 2732 EPROM. If you follow the guide lines there it works like a charm.

 

If not following the hints the minimum problem to happen is that it is not working with 2732 types whereas 2764-27512 are programmed fine. The max problem that could occur is a damage of the burner.

 

To state it very clearly:

My goal was and still is to help to you to bring this burner as a re-imaged project back to the community. And despite lacking a lot of skills I put quite some effort into it, dug out the original manufacturer sources from the ABBUC archive, sent as much material as I could about the old burner I use and tested all the items from the old user guides to make sure it is most accurate before I started to develop an English version of it.

 

As a thank you I received a complementary burner with two 2732 adapters from you and promised to test it with the goal to describe in the user guide which are the needs to know to successfully burn 2732 types with the new version of the burner and the original version using the second adapter. Thanks to tf_hh, who described the necessary work around to successfully burn, read and verify 2732 EPROMs with the re-imaged version, this is now possible and I will adapt it to the user guide in English and German. Last step will be to test it with the original burner and your adapter coming days.

 

So let's blast the hardware and software bugs and make it to a perfect project from bits of the past.

[Timothy Kline]

Off-topic, but rather than start a new topic...

 

Looks like things are shaking again in Dropcheck's Atari-OK Corral... another earthquake. O.o

 

--Tim

[Dropcheck]

Yes, and I sent you details in e-mails and described what is happening here and made a video to show you, what is not working. What more can I do to help to kill bugs?

 

But all you responded with was a simple denial of my statements instead of testing it by yourself (at least you didn't write in your e-mails that you had tested what I described). I am neither a programmer nor a hardware developer, just a long time experienced A8 user. And my goal was to help to wipe off the obvious bugs from that project. But I guess my approach was understood as pointless criticism.

 

In the meantime there is professional proof that I stumbled across bugs. With the help of tf_hh they are identified and described in the ABBUC forum. I recommend every user of that burner to study it thoroughly before trying to burn a 2732 EPROM. If you follow the guide lines there it works like a charm.

 

If not following the hints the minimum problem to happen is that it is not working with 2732 types whereas 2764-27512 are programmed fine. The max problem that could occur is a damage of the burner.

 

To state it very clearly:

My goal was and still is to help to you to bring this burner as a re-imaged project back to the community. And despite lacking a lot of skills I put quite some effort into it, dug out the original manufacturer sources from the ABBUC archive, sent as much material as I could about the old burner I use and tested all the items from the old user guides to make sure it is most accurate before I started to develop an English version of it.

 

As a thank you I received a complementary burner with two 2732 adapters from you and promised to test it with the goal to describe in the user guide which are the needs to know to successfully burn 2732 types with the new version of the burner and the original version using the second adapter. Thanks to tf_hh, who described the necessary work around to successfully burn, read and verify 2732 EPROMs with the re-imaged version, this is now possible and I will adapt it to the user guide in English and German. Last step will be to test it with the original burner and your adapter coming days.

 

So let's blast the hardware and software bugs and make it to a perfect project from bits of the past.

 

I verify every board I send out that it will correctly burn/verify/read the eproms sizes I say it will handle. GoodbyteXL was sent a working mainboard and adapter. He also volunteered himself to update the manual with the correct information for the new eprom burning feature.

 

This is what GoodbyteXL is not telling you.

 

On Oct 29th he notified me that he had recieved a bunch of 2732A eproms and that he was having problems successfully burning them. His complaint was that only 256 bytes were being written. I replied that evening my time and asked him some questions and one of them was if he had tried manually entering the starting and ending parameter. I wanted to see if there was a problem with the parameters(length of bytes to burn) The original software only sets three eprom sizes automatically, 27128, 27256 and 27512. To correctly burn the 2764 the manual instructs you to manually enter the starting and ending parameters. Since the changes I made to the software deal only with the graphic display, not with the actual burning routines the same manual entry for starting and ending parameters for 2732A eproms has to occur. I added the ending parameter for the 2732A to the display in case you did not know it before hand.

 

GoodbyteXL's reply did not state he had tried manually entering the parameters. His main focus seemed to be that some hypothetical burn method that didn't actually use a real eprom would correctly model the burning method and it wasn't working either, He also did not give a specific list of what settings he was using. Just a general statement that of course he was using the correct settings.

 

On Oct 30th I sent another email and tried again to get a detailed description of what settings he was using. I also sent him a detailed list of the settings I was using and specifically stated to manually enter the starting and ending parameters and try a real burn. I specifically cautioned him not to attempt a high level 'fake' burn method as it most likely would not work with the new switch setting and adapter requirement. He did send me a video that showed the screen behavior of his complaint and continued to champion the fake burn method. He still did not indicate that he had attempted the manual parameter entry or what the results were.

 

In the meantime apparently he took his concerns to tf_hh where using GoodbyteXL's settings tf_hh was able to verify that his settings resulted in a failed burn/verify/read. That only proved that GoodbyteXL's settings could be duplicated, not that there was a problem with the burning of 2732A eproms.

 

During the process tf_hh apparently found an undocumented/unknowed problem with the one remaining open switch position, the one I had used to route the programming voltage to the correct pin for the 2732A eprom adapter. Since I had effectively cloned the original board that flaw is also on the original boards. It isn't fatal, but there is cause for concern. For both original and reimaged boards you want to set the switches either both up or both down prior to shutting down or starting up the computer with the board attached. Upon tf_hh notifying me I posted the update in the forum topic.

 

On Oct 31st, in my reply to tf_hh I again stated the detail list of the settings I was using, indicating that the starting and ending parameters had to be manually entered. I asked him to test those settings. Just as I had asked GoodbyteXL to do two days before. tf_hh was quickly able to verify that if you manually enter the parameters it does indeed burn/verify/read correctly and replied on Nov 1st. tf_hh also notified GoodbyteXL who confirms he is now able to successfully burn his eproms.

 

If GoodByteXL had done what I asked him to do on the 29th, he would have been able to successfully burn his eprom. Instead he came on the forum and made misleading statements on Nov 1st. I have wasted literately days troubleshooting this non-issue and all he had to do was follow my instructions.

 

My mandate from ABBUC was to clone the board and make it available for sale. I was given permission to add the option to burn a 2732A eprom via an adapter. I also have attempted to add the option to use a MC68B21 or R65C21P2 as an alternative PIA and offer that for sale at a reduced amount for those who have their own PIAs. But I was not given permission to fundamentally change the hardware or software. I cloned the hardware so well that the unknown/undocumented fourth switch position issue was also transfered to the new version. I supplied a new version of the software modified only enough to display the new option, what the new switch setting should be and what the ending parameter should be. If you don't want to use that, then the original software is available in the same downloaded compressed file.

 

Enough said.

[Dropcheck]

Last update in this thread:

 

I have finished testing the PIA adapter that will allow you to use other PIAs such as the MC68B21 and R65C21P2. Both additional PIAs work well. All currently sold boards were sent out Saturday.

 

I hope to have a normally carried inventory of new boards after the first of the year. So if you are interested check back on my website then to order.

[tuf]

I hate to bring this up but I'm having issues burning TMS-2732A-25JL eproms on my Rolf burner. BTW Dropcheck, this thing is awesome. I've burnt a ton of 27128 and 27256 without incident. I read up on the issues above and tried everything listed, even reviewing the ABBUC thread for additional details.

 

My process is this:

 

1) Start the burner/Atari with the switches in the UP position

2) X - change voltage to 21v to match the chips spec

3) F - input a 4k bin file name

4) P - manually enter the following: 0000,0FFF

5) Set switches to 2732 mode

6) E - run an empty test

7) B - burn, and get an error code just after the bin is loaded from the disk

 

If I switch from fast to 'save' I actually see something that looks like a successful burn (the address quickly advances at the bottom on the screen) but the chip tests blank afterwards and does not work.

 

If anyone sees that I've missed a step somewhere, I'd appreciate some advice.

 

Thanks!

[1050]

Always assume the eprom is bad until it programs

and works?

 

Essential leg work is to look

up TMS-2732A-25JL on a search engine and verify

that you have a 21 V programmable eprom.

Checks out that you do, but this is my first

step since there are 25 Volt eproms out there

under the 2732 flavor. In this case go by only

a Texas Instrument published document that also

has your device listed by exact number.

 

I would hook up a chineseum LED voltage display

to your 21 volt onboard source, verify that what

it says is accurate enough to use and see what

the voltage actually is during programming

attempt.

http://www.ebay.com/itm/Car-Auto-Vehicle-Boat-0-56-LED-Digital-Voltmeter-Battery-Charging-Indicator-Red/222613762471

 

Then there is the "advice" found in very rare

and hard to find programmer manuals and/or

datasheets that suggest you should try cheating

the voltage setting by half a volt, so adjust it

to 21.5.

 

And after that, reset to 21 Volt and toss the

eproms for bad ones and get another load in there

as that batch will never program - they are bad.

This happens a lot, any marks found on their

surface is a possible indication that they have been

kept in bags and tossed about like so much dirty

laundry. Bent pins another sign of wasted money down

the rat hole for the same reason - eproms are very

static sensitive, unlike most of our Atari chips which

can be handled a lot without too much worry, eproms

always need to be worried about. Trust them like you

would a known thief which is NONE at all until they

prove out.

 

Another sign that they may be bad is the requirement

for more than 20 minutes of erasure exposure.

Although the rare good one will require another 5

minutes, it is a sign that they are not quite right

in the head.

Edited September 21, 2017 by 1050

[Dropcheck]

I hate to bring this up but I'm having issues burning TMS-2732A-25JL eproms on my Rolf burner. BTW Dropcheck, this thing is awesome. I've burnt a ton of 27128 and 27256 without incident. I read up on the issues above and tried everything listed, even reviewing the ABBUC thread for additional details.

 

My process is this:

 

1) Start the burner/Atari with the switches in the UP position

2) X - change voltage to 21v to match the chips spec

3) F - input a 4k bin file name

4) P - manually enter the following: 0000,0FFF

5) Set switches to 2732 mode

6) E - run an empty test

7) B - burn, and get an error code just after the bin is loaded from the disk

 

If I switch from fast to 'save' I actually see something that looks like a successful burn (the address quickly advances at the bottom on the screen) but the chip tests blank afterwards and does not work.

 

If anyone sees that I've missed a step somewhere, I'd appreciate some advice.

 

Thanks!

 

Hi,

 

A couple of questions to gain more info:

 

Did you do a blank check first? Did it indicate the eprom was blank? How did you erase the eproms? How long did you run the eraser?

 

How many of the 2732A eproms have you tried to burn? Have you tried any other 2732A eproms from another manufacture?

 

 

Dropcheck

[chevymad]

Earlier in the thread there was some talk about Atari PIA chips being 2mhz and should work. Has this been confirmed by anyone? I have at least 1 bad Conexant PIA, and have yet to successfully burn a chip - for multiple reasons. But I'll still need to order at least one new PIA, hopefully avoiding China this time.

[+Nezgar]

Earlier in the thread there was some talk about Atari PIA chips being 2mhz and should work. Has this been confirmed by anyone? I have at least 1 bad Conexant PIA, and have yet to successfully burn a chip - for multiple reasons. But I'll still need to order at least one new PIA, hopefully avoiding China this time.

 

Well, the Atari runs at 1.79Mhz, so Atari would have chosen PIA's of at least 2mhz...

 

Just checked two 800XL's I have handy, and their PIA's are Synertek P6520A variants. No letter = 1Mhz, letter "A" = 2Mhz. (for Rockwell, C=Ceramic, P = Plastic) But googling a few other motherboards, I do see some 6520's without the "A" so not sure about those. Probably 2Mhz, but the replacement of the Atari Part number replaced what would have been the indicative manufactuer part #.

 

Dropcheck's page specified either "6520AP or MC68B21" so back in 2017 I just picked up 2 MC68B21's for mine from eBay china... They worked fine after I got a handle on how to use the software... It's really finnicky. And heed the warning to ensure no EPROM is present when the computer is powered up...

 

Same seller still has them listed: https://www.ebay.com/itm/271809726866

[chevymad]

Mine's a version 1 board so the mc68b21's won't work. However, I do see that she now says that "most" Atari PIA's should work. Best has them for $1.95 I think.. couple other things I'd like, so probably order a couple from them.