28Cxxx EEPROM Programmer

[Universal2600]

Hello I've been tinkering with this and I'm stuck, I'm trying to get this to program an at29c010a 128k chip from atmel, I added the shift registers so I'd have 17 bits of addressing, I can get it to read out data, using the nano/ch340 Chinese sweatchips, changed the code to use long since int is a 16 bit variable and I need 17, but it will not write anything, which could just be a w/e pulse mismatch but I'm using the arduino serial terminal to send commands/test the circuit, but when try to write it just hangs and the duino won't respond unless you reset( no light flash unless I send a read command)

[Dr Jefyll]

Welcome, Universal2600

One thing you could do is treat the at29c010a 128k chip as a 64k chip by tying address input A16 high or low. Then the remaining sixteen address inputs would come from the programmer.

Two things arise from this. As a temporary, workaround solution it would allow you to program the chip (although rather inconveniently) in two separate passes. There'd be no need for hardware or software to support a 17-bit address, so these options could be omitted or temporarily bypassed.

If you're unable to make this work then it probably indicates a defective at29c010a or some other underlying issue. In other words bypassing your 17-bit mod's allows you to isolate certain parts of the problem -- which of course is a fundamental troubleshooting technique.

HTH -- good luck!

Jeff

[theodor613]

Just wanted to say thanks to rpress and marc to helping with the schematic on this programmer. I've made some boards based on the optimized version that rpress did.I also got it looking close to what unaclocker had. Haven't soldered anything yet but will this weekend.
Will report back when soldered and tested.

Here are some pictures in the mean time:
http://i1337.photobucket.com/albums/o67 ... cdhbpi.jpg
http://i1337.photobucket.com/albums/o67 ... 6ofnum.jpg
http://i1337.photobucket.com/albums/o67 ... lbroxs.jpg
http://i1337.photobucket.com/albums/o67 ... uokytb.jpg

Sorry for the links but the images are big.
In the last photo, i did not account for the zif socket size as i ordered the socket after i had submitted the gerbers. But the regular dip socket will be fine for me as i have a ic puller.

[dwight]

I use a 28C64 on my 6530 to 6532 adapter. I don't use a programmer,
I just use the 6502 on my KIM-1 and program it in place.
I'd think it would be a big deal to even used a C64 or less without
too much effort.
I'd use my EPROM programmer but the 28C64 is one of those
surface mount ones ( I think it is called a sot ).
The programming isn't too complicated.
I copy from an EPROM but I could just as easily used a serial input.
Dwight

[BigEd]

(Welcome theodor613! It's best if you do attach images, but if they are big, better not to inline them. The forum software will automatically show a reasonably-sized preview.)

[theodor613]

Ok, I will keep that in mind when I post up the boards after soldering the components.

[dwight]

I think the jest of my post was lost.
These chips were really intended to be programmed in place. No need to
make a special programmer.
Even with the chicken-and-egg problem, most any machine,
that runs, with an extra, empty, EPROM socket, can be easily be made
into a FLASH programmer with minimal write enable logic.
Dwight

[theodor613]

I am using this programmer just to program the 27sf512 chips.

I have everything soldered on one board. Everything goes smoothly(had a finicky USB wire connecting to my Leonardo and uno) but i am getting a failure for the verifying part.

[dwight]

You might check the data sheet for this chip. Unlike the 28C512, these need 12V applied to the correct pins to program. The 28Cxx parts have a built in charge pump circuit to create their programming voltage. 27sfxx do not program the same as the 28Cxx parts.
Dwight

[dwight]

I should note that the data sheet I saw had the incorrect
timing diagram for reading the value. One should use the
levels stated in the table and not the diagram.
Dwight

[theodor613]

Dwight, there is a charge pump circuit on the boards. Rpress has created the circuit to program the 27xXxX chips. It has worked before on their boards. I've double tripled check my traces and everything goes where it should. Next step would be my soldering. The caps and solder for them were exactly the same size. It might not have a proper connection. Gonna solder 2 more boards, one with hot air soldering and one with my reflow oven.

dwight wrote:

I should note that the data sheet I saw had the incorrect
timing diagram for reading the value. One should use the
levels stated in the table and not the diagram.
Dwight

I've message rpress about the reading issue.

[dwight]

As long as the Vpp is right, it should work.
It does need you to erase them before programming.
You can only flip bits one way by writing, unlike the
28C parts.
When you verify, does it show you the value read?
Did you check capacitor polarity?
Can you post the schematic?
Dwight

[theodor613]

Schematic is on page 8 and fourth post down. Polarity on the capacitors are correct. It is failing because it won't erase. So even though it says it wrote the data, it shows as nothing has changed.

[BigEd]

(You can link to a specific post using the little icon to the left of Posted: - in this case, here.)

[dwight]

I see no schematic there?
In any case, if I understand right you are using an arduino.
It sounds like you need to write some board test programs
to measure and verify things without a flash chip in the socket.
Dwight