PROM programmer suggestions?

[Kchula-Rrit]

Not sure if this has been asked in a while, but does anyone have a suggestion for a good PROM/EPROM programmer that does not cost more than a used car?

Would like one that runs under Linux, but most I have seen want Windows, and I grew tired of the Microsoft "upgrade" merry-go-round years ago.

I looked at a Conitech (not sure of the spelling) unit ages ago, but could not justify buying one at the time.

K-R.

[atrax27407]

Check out the T56 at: TL866 High Performance Universal Programmer (autoelectric.cn)

 

 

Here is their Ebay store:  XGecu Official Store | eBay Stores

Edited October 15, 2023 by atrax27407

[matthew180]

@Kchula-Rrit

 

Good luck, let us know what you find, or if you give up and build your own and turn it into something you make and sell to the retro community.

 

Seriously.

 

Supporting older chips started being a problem over two decades (or more) ago, and your list of requirements will never be satisfied by a commercially available inexpensive device.  There is no money in retro computers or supporting ICs from 20 or more years ago.

 

The XGecu and TLxxx series of devices from autoelectric (linked in previous post) work well and are inexpensive, and they even have a pretty impressive list of supported chips.  However, they do not support MANY older ICs found in coin-op and retro systems, and their software is Windows only (which you will never get away from when it comes to unique hardware devices like this).  I own the TL866 and I have an old Win10 machine that I keep around just for running various software for some of the hardware "things" I have (TL866, camera for my stereo microscope, specialized architecture CAD software, printers, plotters, etc.)

 

Even with the TL866, I was working on an Apple2 disk controller a few weeks ago, and all the PROMs on that board are not supported by ANY version of ANY commercial hardware programmer I could find.  I think many of the ROMs in the Apple2 fall into that category as well, but I gave up after trying to find a way to read the PROMs.

 

Device programmers have always been expensive, so be prepared to spend a LOT of time, in addition to one or more of the following:

 

1. spend a lot of money.

2. build your own custom circuit for the chip you are dealing with.

3. a lot of 1 & 2 which evolves into 4.

4. buy something close to what you need and build adapters.

5. decide this is too expensive of a hobby when you start trying to fix them or mess with the 40+ year old electronics.

 

Another hobbyist decided to build their own universal retro chip tester:

 

https://8bit-museum.de/sonstiges/hardware-projekte/hardware-projekte-chip-tester-english/

 

No software needed, it is stand alone.  But it focuses on reading and testing, not programming.  It probably supports all the retro chips you will never need (although I don't see GROM support in the list), and there are adapters that allow programming a subset of the EPROMs that modern programmers do not support.

 

[atrax27407]

The TL866CS will support a surprising number of old EPROMs, some of which are not mentioned in the Support List. For example, the TMS2532 is nominally a 25V chip and is used by a host of TI cards. The Great Gazoo discovered that it uses the same die as the TMS2532A (21V) and will program very nicely at 21V in a TL866. All that is required is a simple 3 wire adapter and programming the chip as a TMS2732A. I have been able to program several hundred of them over the last 10 years or so. Most of the older 25V chips have newer 21 or 12.5 versions that will program (and are supported) by TL866.

 

Another thing to remember is that, even if the chip is not specifically mentioned in the Support List, there is usually a similar chip with identical pinouts that is supported. You can simply check the datasheet and program the chip as the variant that IS supported. I have found very few (in fact almost none) chips that I couldn't find a "work-around" to allow programming by using a simple adapter ot a similar algorithm.

 

The T56 will program up to 25V chips so it is likely that it will do most everything that you need it to do for our "venerble" TI99. In fact, if I didn't already have a couple of TL866s, I would definitely go for the T56. After all, it supports 30,000+ chips.

[Kchula-Rrit]

I am currently looking at a Batronix BX40, available from Newark in the US for $243.

The Batronix Web-site says:

• Supports "very old NMOS EPROMS that need >15V." I specifically looked through their supported-devices list for 25V parts (2716, 2532, 2764, etc).  They conveniently listed the voltage on a lot of the parts.
• Works with Windows and Linux.  For Windows it wants NET 3.51 which, I think is Win7 (a friend gave me a Win7 machine some years ago.)  For my Linux purposes it wants Ubuntu 14 or higher (I use Debian, which the site says will also work if I load Mono; will have to try it)

Unfortunately, I blew my computer budget for the month with Shift838 and ArcadeShopper, so it will probably have to be a Christmas present.

 

[matthew180]

Oh, that Batronix looks nice!  Thanks for sharing that brand.  A BX48 just might be my next purchase if I find enough supported retro chips in their documentation.  Just supporting non-Windows OSes is a very big surprise and adds helps tip the decision in their direction.

[+Ksarul]

I use a BK Precision 844USB for most of the truly ancient chips, as it supports just about everything from 2508 to 2564, as well as 2708 to 2764 chips. I've even gotten it to work fine with 512k x 8 Fujitsu 27C4000 chips using a 27C4001 adapter--a seriously heavy lift, as none of the rest of the dozen or so programmers I have can do that. Downside is that it is relatively expensive new ($500+)--and recently discontinued, so the only options are now used.

[+FarmerPotato]

I hit a wall when I realized 2516 and TMS2532 aren't supported in the T56 --was not able to get a good read, much less write on a fresh one. TI and Intel had different pinouts. The 2516 are in a bunch of TM990 boards. 
 

 

I figure to just wire up my own reader, then use an adapter to a 64K EEPROM. 
 

I want to try out a 28C16 EEPROM. One issue is the CE and OE  being swapped vs TMS2516.  It may or may not tolerate that in use. 
 

 

[atrax27407]

Here are a couple work-arounds for the TL866. Brain sells the premade 2532->2732 adapters on his site.

TMS2532 Adapter.pdf Adapter for 2732 in a TMS2532 Socket.pdf

Replacing a 2716 with a 2732 EPROM.pdf

Edited October 23, 2023 by atrax27407

[atrax27407]

Also, since the TMS2532 "clones" were licensed by TI, they will also program flawlessly with the TMS2532 adapter at 21V. The Motorola MCM2532C and Hitachi HN462532G both work without any problems. 

[atrax27407]

There is also an adapter that allows the TL866 to do 27C322/27C160/27C800/27C400 chips. It is available on eBay in both finished and as a bare board when I last checked. Of course, there are all sorts of adapters for PLCCs and TSOPs as well.        

Edited October 24, 2023 by atrax27407

[Kchula-Rrit]

On 10/22/2023 at 11:33 AM, matthew180 said:

@Kchula-Rrit

 

Good luck, let us know what you find, or if you give up and build your own and turn it into something you make and sell to the retro community.

...

 

After negotiating with the finance department, I went ahead and ordered the Batronix BX40 (manufacturer's info).  The heart pills can wait for another month... just joking!

 Newark has them in stock.  A few days ago I found some AT28C64B 8Kx8 and AT28C256 EEPROMs in my piles of parts.  We'll see how they program.

Hopefully, I can use them in my cartridge project.

K-R.

[matthew180]

Cool.  I emailed the support at Batronix, and they replied within a day.  They said they were not really willing or able to support older ICs, which is not unexpected, but also not great to hear for retro computing.  I'll be curious to know what your experience is like with it.  It still might be worth it to replace the TL866, if only for non-Windows software.

[Kchula-Rrit]

Thanks for letting me know.  I wonder if the supported-devices list I saw on the Batronix web-site was an old one.  I do not see me using things like 1702s, but hope it works with 2764s and such.

The info on the Newark site said the programmer works with Win98, which would have been a large bonus for me, as I still have a couple of them that I still use.  However, Batronix says their Windows software wants NET 3.51, which, I think, is Vista at minimum.  Sigh.

K-R.

[Kchula-Rrit]

The Batronix finally arrived a few days ago; have to load the software to try it out.  Here is a photo to show its size.  The 40-pin Zip-DIP gives an idea of its size.  I have been told I have little womany hands. 8-)

The first PROM programmer I used was about the size of a bread box, and used a 9600bps serial link to load the data.  I feel like Rip van Winkle!

At the moment I have these chips in my stash:

Flash (?)

MX27C1000PC-15  (had what looked like a BIOS sticker on it)

N28F010-120 (Intel, these are actually labelled flash)

AT28C64B-15PU   8K x 8

AT28C256-15PU  32K x 8

EPROMs:

M27C256B-20 (ST micro)

MM2708Q (National Semi)

AM27C040-70   512K x 8

K-R.

Edited November 16, 2023 by Kchula-Rrit
27C040 is an EPROM

[atrax27407]

The TL866 will do almost all of the ones you listed. The ones that aren't have a pin-compatible algorithm that will work by programming it "as xxxx".

[atrax27407]

The only really problematic EPROM is the TI TMS2716 which has three supply voltages and is NOT JEDEC compliant and not compatible with anything.. However, the TMS2516 IS JEDEC compliant and compatible with all other 2716s.

Edited November 16, 2023 by atrax27407

[Kchula-Rrit]

On 11/16/2023 at 10:01 AM, atrax27407 said:

The only really problematic EPROM is the TI TMS2716 which has three supply voltages and is NOT JEDEC compliant and not compatible with anything.. However, the TMS2516 IS JEDEC compliant and compatible with all other 2716s.

I remember a co-worker advising me against using TI 2716s in the company prototypes back in 1982-83 and using 2516s instead, if I used TI chips.  The first time I used a 2716 I inserted it backwards in a test fixture, and marveled at the little pilot light clearly visible in the little window.  When I took it out of the socket it was really hot, and did not work when I plugged it in correctly.  So much for the "pilot light."

Another Rip van Winkle moment!

K-R.

[Kchula-Rrit]

Here is my first experience with the Batronix.

First thing, it had no external power jack, so it need to run from a powered hub.  First frustration, I had to hunt down a hub, then find a power brick to power it.

I gave up trying to get it to work with Linux after some hours of fiddling.  The Web site said that Linux requires Mono (Linux version of Microsoft NET) to work, so I installed it first.  Then I downloaded the Batronix program, called "prog-express," and tried to load it.  My system is Debian 11 ("Bullseye"), so I used the Ubuntu version.  dpkg reported a whole list of unmet dependencies, all related to Mono.  It wanted "mono2", which I could not find on the Mono Web-site.  No amount of swearing fixed the situation/ 8-)  I removed prog-express and all the mono stuff I had already loaded.

After downloading the latest Windows version (3.8.9-4) I tried to install it.  Now it says it wants M$-NET v4.  My system has v3.5 installed, so I removed prog-express.  Then I tried the oldest one on the Batronix Web-site, 3.6.0-1.  It loaded, but complained that the firmware in the programmer was too new to work!  The programmer was v2.38, and prog-express would not talk to anything newer than v2.35.  Arrgh!

After downloading successively newer versions I got everything to work with 3.8.8-1.

But the adventure was not over yet.  I plugged a 27C256 into the programmer and tried to read it.  The data was all 0xFF, so I figured that the PROM had erased itself over the years it had sat in my parts box.  I changed a few bytes and tried burning the PROM.  Exited the program and re-started it to read the PROM.  It came up all 0xFF again.  Then I noticed that I had the chip plugged in backwards, so I turned it around.  Every ZIP-DIP I had used before had Pin 1 where the lever is located.  This one is the other way round.  Anyway, after turning the chip around it still came up all 0xFF!  Then I noticed the markings on the case, which show the chip needing to be seated next to the lever; I had placed it at the other end, with Pin 1 up against the corner opposite the lever.

After I placed the chip next to the lever, as marked on the case, it finally worked!  The chip produced readable data!

Now, to try writing to a chip.  I grabbed a AM27C040 that had arrived a couple of days ago and read it.  That worked; the data all came up 0xFF, like I expected for a new chip.  Then I again changed a few bytes and and burned it to the chip.  It worked!

Thanks for reading.

K-R.