I asked about how to program the ATF22V10C on reddit a couple of weeks ago, and was told no less than 3 times by someone named "Linker3000" that I should come over here. So, here I am.

With his help I found some information that gave me false hope that I knew how to program the chip. I wouldn't discover until after ordering them that the homebrew programmer I'd studied the source code of was known not to work with the ATF22V10C. However, the differences weren't hard to figure out. The ATF22V10C just wants the address bits in the opposite order, the data bits have to be read on the opposite phase of the clock signal, one of the fuses is impossible to verify as being programmed correctly, and that power-down feature is enabled by default and so I had to figure out how to disable it.

So I made a web page about what I figured out. Since I note there are several threads here about programming this chip specifically, and more about programmable logic in general, I thought you all might want to have a look at it: http://www.ecstaticlyrics.com/electronics/ATF22V10C/

I'd be quite amused if anyone could give me any more information about the chip, but after searching the internet for days, I'm fairly convinced that I now know more about how it works than anyone else does, aside from those bastards at Atmel who refuse to publish any information about it.

Welcome, Moosfet, and thanks for sharing your information. You're not the only one with an interest in homebrew programmers, and the ATF22V10C isn't the only eligible chip.

To me, the roll-your-own approach is natural and has obvious appear -- not only to avoid the purchase price and shipping cost/delay of a "store-bought" programmer, but also because many of us actually relish undertaking a problem they can solve themselves, testing and refining their skills in the process.

The lack of published programming info is disappointing but not a bit surprising. For one thing, it costs money to publish doc. More to the point, manufacturers want to avoid anything that might compromise the reliable operation of their products in the field. Homebrew programmers are a missing link in their chain of quality control, and probably that's why they suppress info pertaining to programming procedures. But that doesn't stop us, in a spirit of friendly competition, from un-suppressing that information!

-- Jeff

_________________
In 1988 my 65C02 got six new registers and 44 new full-speed instructions!
https://laughtonelectronics.com/Arcana/ ... mmary.html

Welcome, Moosfet, and thanks for a thorough investigation and useful write-up!

Your write-up's colorful vocabulary made me smile, too.

Mike B.

Ooh! Thank you.

I know what's going to be in my next electronic parts order. (-:

Very nice, ATMEL is way too protective regarding this information and I could't as well never think how they sell there chips. Especially as also the list of compatible programmers they provide is utterly outdated and lists only rather expensive vendors. It would be interesting if this information also applys to the ATF2500.

Do let me know if you have any problems following my instructions. While I've proofread them several times, the number of times I "corrected" my description of pins 4 through 9 before finally describing them correctly (at least I hope so) is kind of sad. I seem to have been much less confused about everything else though, so I'm pretty sure I have all of the details correct now.

Also, I recommend working on the read functionality at first, so that you can make sure you have your bit clocking routines functional before burning through any of the chip's rather limited lifetime of 100 erase cycles. The identification string that is in the chip by default can serve as an aid to recognizing when you're reading the chip's contents correctly. Also, if you keep clocking bits after you read the 132 bits from a row, the next six bits are the row address that you read from, just with the bits inverted, and so looking for that can help you to see that the chip is receiving your address bits correctly and that your reading of the chip's output bits isn't misaligned.

The chip seems to be resilient to bad programming. I ordered ten chips assuming I might break a few in the process of trying to sort out how to program it, but the chip I did all of my testing on still works, and that's after I decided that since my programmer was already working, I might as well risk breaking one of the chips by doing seemingly incorrect things just to see what might happen, e.g. longer/shorter strobes when erasing and writing, writing to unused row addresses, etc. So as long as you don't accidentally erase the chip more than 100 times, or apply the programming voltage to the wrong pin, I expect just about anyone can get a programmer working without having to buy more than one chip for testing.

At the very least, the pin-out is going to be different, but I suspect a lot of the process would be similar. I've also built programmers for the AT89S52 and for the AVR series of chips (ATtiny48, ATmega328, etc.), and despite being significantly different platforms, Atmel gave both a rather similar programming procedure, in that each requires four-byte "instructions" to load data into the chip and then tell it what programming operation you wish to perform. So it's easy to imagine that all of their programmable logic also shares a rather similar programming procedure, especially if it is from the same era. However, if it is different, e.g. using the four-byte "instructions," then sorting out how it works by guesswork may prove to be quite time-consuming. (Atmel provides programming information for the AT89S52 and the AVR series, so I didn't have to figure that stuff out on my own.)

The easiest way to figure it out would be to find someone who has a programmer for the chips and borrow it so that you can look at what it is doing with a logic analyzer. That would easily show you which pins do what. Then you can feed the programmer some custom JEDEC files to reveal where each bit should be programmed into the chip's memory, and also you can compile some simple circuits in WinCUPL to sort out which bits in the JEDEC file are assigned to which circuits in the chip.