S4 appears to program but eprom wont work in cct

[Guest]

Hi,

I just erased 10 x 27C256 eproms, burnt an updated software version on them and now they appear dead when I plug them in.

They all verified the same as the master (and still do) and the checksums are all the same. The master eprom works in the circuit, but none of the copies.

I burnt 60 of these exact same types a couple weeks ago with absolutely no drama. The only thing that went wrong was a couple didn't ersase properly, which I picked up as I started the process, so re-erased them.. Every one worked without a hiccup.

I'd had to reload the 'Library' because I had installed a Serial library to do some small eeproms since the batch of 60 x 27C256B's.

The Dataman didn't want to accept the new library for some time but then did, and everything else appeared normal.

I have since reloaded the serial library again to check that it would change .. and then back to the one for the job. Library went in ok this tiime, but still no joy.

I can put in any of the copies I've made (in the programmer) and they still verify as the same as my original data, and the checksums are still all the same too ... yet the original is the only one that works in the circuit boards (I've tried a couple different boards too)

If the Dataman is crook, why does the data read off all the chips ok and the checksum ok ... yet not one of the 14 I've tried have worked in it's location.

Please Help !!!

Thanks, Steve E

[Dataman (Neil Parker)]

Hello Steve E,

What speed are your latest batch of devices rated at? I ask because it sounds like the S4 is working properly. These 27C256B devices can range in access speeds anywhere from 45ns to 250ns. Perhaps your application requires a slightly faster access speed from the devices in order for it to work?

If you have checked this and do not believe it to be the cause of the problem please let me know and we can look into it further.

Regards


Neil Parker
Dataman Programmers

[Guest]

Hi Neil,

Thanks for your reply. The eproms I am using are all 120 or 150ns. They were actually working chips that came out of the boards and have just had a later version of software installed.

Since posting my question I have found someone to burn them on their programmer and they have all worked ok in their respective target boards.

So, I am very certain the issue is with my S4, and have just spent the last few hours reloading the library and experimenting.

What is strange is that the master's checksum is different to what I had recorded for it some time ago, but the master works in the target board every time ... so I sort of ignored that. Big mistake I think.

When I load the master into ram, the checksum is 00340604, and when I burn the eproms, their checksum is the same also ... yet none of the copies worked in the target boards. At the end of each 'burn' it comes up as 'same'.

I was also diligent in making sure I chose the correct algorithm for each chip.

Just once an hour ago after reloading the library, the checksum of my master came up as the original, so I tried to burn a copy but it failed part way thru'. Then the checksum was back to 00340604.

Hoping you can point me in the right direction.

Thanks, Steve E

[Guest]

Something I forgot to mention, was that most times when I reload the library it comes up with ...

DATAMAN S4 RY?
LIB 3.00
128K BYTES RAM

What is the 'RY?' all about?

Steve E

[Dataman (Neil Parker)]

Hi Steve E,

Ok there are few things you can try to locate the problem. Please run the RAM test on the S4 (FUNC INFO > ESC > ESC > ENTER) If you can run the test 3 times and let me know the result.

How are you loading the new libraries into the S4? Is it from library PROMS or via the serial cable directly into the S4? The "RY?" on the S4 display after loading a library is simply the end of the text that was there previously. Turn the S4 off and back on to refresh the display.

It is also worth locking a device in the ZIF socket and using your thumbs slide it from left to right a few times to clean the contacts, this often helps with intermittent checksum and programming problems.

If you repeatedly take a checksum of a device in the socket, is the checksum consistent or does it change?

Regards,

Neil Parker
Dataman Programmers

[Guest]

Hi Neil,

Thanks for your detailed reply and sorry for my slow response. Scored a virus on the pc!!!

Before I got your reply, I had decided that the S4 was faulty so took the cover off to check it out. Gave it a good clean and also reseated the IC in it's socket.

When I did a checksum on the IC I was trying to copy, it was still the same as above, so tried it a few more times. For some reason I thought to flex the board a little, and then the checksum was correct.

It was always either one of two ... 00340604 or 0036AB38 (correct number)

I couldn't find any dry joints and I haven't dropped it or been rough with it so not sure where my problem is. Both batteries were replaced about a year ago.

I've only tried it a couple times since and it now seems ok.

Ok there are few things you can try to locate the problem. Please run the RAM test on the S4 (FUNC INFO > ESC > ESC > ENTER) If you can run the test 3 times and let me know the result.

Ran the RAM test and it passed each time.

How are you loading the new libraries into the S4? Is it from library PROMS or via the serial cable directly into the S4? The "RY?" on the S4 display after loading a library is simply the end of the text that was there previously. Turn the S4 off and back on to refresh the display.

Loading from my PC via a serial cable. I had saved it (I thought) on a 27C256, but couldn't get it to load back in from that.

It is also worth locking a device in the ZIF socket and using your thumbs slide it from left to right a few times to clean the contacts, this often helps with intermittent checksum and programming problems.

OK, I've done that now too. Thanks.

If you repeatedly take a checksum of a device in the socket, is the checksum consistent or does it change?

As above, during this problem it has only ever been one of two numbers .. either the correct one 0036AB38, or 00340604.

Thanks for your help. I am about to do another batch, so will see how I go, but am still interested to know what you think.

Regards, Steve E

[Dataman (Neil Parker)]

Hello Steve E,

To be honest it does sound like a dry joint or similar. I would suspect the ZIF socket initially. If you want to return the S4 to us for repair please follow the direction on our website.

If you have any questions, please let me know.

Best regards,

Neil Parker
Dataman Programmers

[Guest]

Hi Neil,

Thanks again. I got through most of them ok, but the problem reappeared, so I resoldered the ZIF and the ram sockets.

It worked ok after that, but I only had a couple to do, so time will tell.

Thanks for all your help.

Regards, Steve E