PDA

View Full Version : Problem reading/writing EPROM



aha2940
03-28-2015, 08:04 PM
Hi

I am trying to burn an EPROM (ST M27C160) but it's consistently failing. I noticed that when I read a blank M27C160 EPROM on my reader, it always reads like this:

FFBF FFBF FFBF...

all filled with this pattern. Of course, when I write the EPROM, it writes fine, but it fails the verification. Funny thing is: the bits that fail the verification are not always the same that show wrong when the EPROM is empty. Sometimes they are, but sometimes they check fine.Things I have tried:

- Using a different version of the software (TOPWin 6 and TOPWin7, my programmer is a TOP3000)
- Using a different PC to connect the programmer. Same problem.
- Using different EPROM models: I tried reading empty D27C1001D, M27C4002, 27C010A, and D27C1000A, all read perfectly (all FFFF)
- Using different USB cable to connect the programmer. Problem persists.

I am at a loss here guys, any suggestions are welcome.

Thanks!!

kuze
03-28-2015, 08:17 PM
You're using the ac adapter to power it right?

aha2940
03-28-2015, 09:25 PM
You're using the ac adapter to power it right?

Yes, however I have not tested that it's working. Any idea how many volts should it throw? I see no marks on either the programmer or the adapter itself.

Thanks!!

bustedstr8
03-28-2015, 09:40 PM
The programmer only needs the extra power for writes, reads should be fine with just the 5V on the USB.

aha2940
03-28-2015, 10:11 PM
Most EPROMs write at either 9v or 12v

I tested the adapter and is throwing 5V. It seems to be the right voltage, so that should be OK, however no clue how the adapter may get 12V from a 5V adapter.

kuze
03-28-2015, 10:17 PM
Yes, however I have not tested that it's working. Any idea how many volts should it throw? I see no marks on either the programmer or the adapter itself.

Thanks!!

Not sure, I know some chips require 12v. You could try topwin8 and updating your firmware to the latest version. I haven't had any issues with 27c160 chips on mine.

aha2940
03-28-2015, 10:30 PM
Not sure, I know some chips require 12v. You could try topwin8 and updating your firmware to the latest version. I haven't had any issues with 27c160 chips on mine.

Do these thingies have firmware? where do I get it? how do I update it? I tried to use TOPWin8, but the laptop I use with the programmer does not have ability to run it, therefore I'm stuck with 7 or 6.

Thanks!!

kuze
03-28-2015, 10:41 PM
Do these thingies have firmware? where do I get it? how do I update it? I tried to use TOPWin8, but the laptop I use with the programmer does not have ability to run it, therefore I'm stuck with 7 or 6.

Thanks!!

Maybe it doesn't update the fw, but there's an 'update' tool in topwin that will get you the latest version, the English manual covers using it. You can get the manual from their site. http://topwin6.com/Soft-1.html

Xian Xi
03-29-2015, 01:09 AM
The programmers use DC-DC voltage converters. I use mine solely on the USB power and I write 160 and 322s no problem.

When you say it's reading blank as FFBF, is the programmer verifying it as black or are you just reading the 160? Either the 160 is bad or you might need a better eraser, use one that actually has a bulb and not LEDs.

fremen
03-29-2015, 04:53 AM
Hi dudes,

Your problem can be the eprom itself. Sometimes the eprom is dead. I assume that you are referred to the same eprom. Have you tried it with another 160?

@James the problem is when try to verify. I suppose that the blank check is passed first.

Un saludo

aha2940
03-29-2015, 10:18 AM
Hi guys, I'll try to clear the questions here:

- I have tried 5 different EPROMs. 2 I have cleared them myself, the other 3 came clear from the buyer. All 5 exhibit the same FFBF pattern.
- The blank test fails. It fails in the second byte, because it reads BF instead of the expected FF. Same byte fails during the verification (after writing).
- The EPROMs write without throwing any error.
- My EPROM eraser is a small chinese plastic box, which has a mechanical chronometer and a fluorescent tube inside (looks like a fluorescent tube, I know it is not).

Thanks for your help so far guys, very appreciated.

kuze
03-29-2015, 10:29 AM
Blanking in topwin shouldn't be necessary as long as the EPROM has already been fully erased. I'd try erasing one for a half hour, then reposition it at a 90 degree angle and give it another half hour. Then read the chip in topwin to confirm it is fully erased before trying to write again.

fremen
03-29-2015, 10:34 AM
Hi,

Thank you for the clarifications. A couple questions, how long for erase a 160?? I use 40-45 minutes.
Do all of them have the same serial numbers?? Some of them have an X and some programmers have problems with them

Un saludo

pulstar
03-29-2015, 10:46 AM
Too much UV exposure to an eprom can damage it. Not sure how long a 27c160 should be exposed to blank fully, but an hour in total sounds a bit too much?

pulstar
03-29-2015, 11:41 AM
I'd write 00s to the EPROM before erasing it though, just to be safe.

With my two programmers (one a Wellon, the other a Top some number I can't remember) you can't write to it if it doesn't pass the blank check. And even when it is being written in this case, it's coming back as FFBF all the way through.

How would writing 00s protect it?

pulstar
03-29-2015, 12:42 PM
It has to do with how UV light charges the cells in an EPROM to set them to 1, and how UV light can "overcharge" a cell when overexposed, making it unclearable (stuck at 1).
By writing 0s, you're actually clearing the charges from it, so you run less of a risk of overcharging an already charged cell.

Also, are these CMOS EPROMs?
Those are a little finicky with supply and programming voltages (needs 6v and 12.5v).

Ah right, cool info, didn't know that...but if the eprom isn't writing properly it makes it a little difficult to write 00's (and also, once an eprom is written you can't change the info until you UV erase it anyway)...27c160's should be CMOS (I think) .

aha2940
03-29-2015, 12:54 PM
Hi

Some more clarifications:

- The "blank" on my programmer, only tests if the EPROM is blank, it does not blank it. It has to be blanked with an UV eraser.
- The EPROMs are all CMOS (they are 27C160, C is for CMOS AFAIK).
- I leave EPROMS about 15-20 mins on the eraser to get them blank. I read 5 minutes should be enough, and that too much time can damage the EPROM.
- My writer will write even if you don't do the blank test. I guess it will blindly write what it's on the buffer, even if it writes right or not.
- All the 160 EPROMS I've tested are ST M27C160-100F1. I have not tried other brands of 160s.
- I have tried reading the EPROMs as ST and also as generic 160 EPROMs. Same results with both.

Thanks!!

bustedstr8
03-29-2015, 01:18 PM
Try filling the buffer with AAAA and see if that burns and verifies.

aha2940
03-29-2015, 02:25 PM
Try filling the buffer with AAAA and see if that burns and verifies.

Hi

Just did the test, it writes and verifies fine.

Thanks!!

fremen
03-29-2015, 05:54 PM
Hi,

Your problem is the erase time, try 45 minutes.

Un saludo

aha2940
03-29-2015, 05:59 PM
Hi

Just did the test, it writes and verifies fine.

Thanks!!

I hate to quote myself, but I also tested writing to the same EPROM a buffer filled with 55 (without erasing it first). It wrote fine, and the EPROM now reads completely filled with 00 (that was expected, right?).


Hi,

Your problem is the erase time, try 45 minutes.

Un saludo

I'll try to erase the previous EPROM for 45 minutes and try again, however I've read online that such a long erase time can damage the EPROM...

Thanks!!

Edit: on a second thought...if the problem was the erase time, wouldn't the incorrectly erased bits be random? I find it strange that all the erased EPROMs get the same FFBF pattern, and that also the EPROMs I just bought, that I have not erased (and they say are new, but who knows...) show the same pattern. I'm starting to think something went wrong with my programmer...:(

bustedstr8
03-29-2015, 06:28 PM
I hate to quote myself, but I also tested writing to the same EPROM a buffer filled with 55 (without erasing it first). It wrote fine, and the EPROM now reads completely filled with 00 (that was expected, right?).

Yes



I'll try to erase the previous EPROM for 45 minutes and try again, however I've read online that such a long erase time can damage the EPROM...

Thanks!!

Edit: on a second thought...if the problem was the erase time, wouldn't the incorrectly erased bits be random? I find it strange that all the erased EPROMs get the same FFBF pattern, and that also the EPROMs I just bought, that I have not erased (and they say are new, but who knows...) show the same pattern. I'm starting to think something went wrong with my programmer...:(

That's what I'm thinking. I had you write AAAA because it should bypass the bad bit.

Try writing 5555 on a blank eprom if it's a bad bit I think it will read 5515

aha2940
03-29-2015, 07:25 PM
That's what I'm thinking. I had you write AAAA because it should bypass the bad bit.

Try writing 5555 on a blank eprom if it's a bad bit I think it will read 5515

You were absolutely right. After writing a buffer of only 55 to an EPROM, I got the EPROM filled with 5515. Now the problem is I don't know if the programmer is not writing the files right, or is simply reading them wrong...any way to know?

bustedstr8
03-29-2015, 07:51 PM
It's definitely not reading that bit correctly, you wouldn't fail to verify every blank you had.

Only way I can think of to verify that you are still writing the proper data is to use another programmer or find someone to read a 5555 filled chip.

If you can't find anyone close I have no problem doing it.

aha2940
03-29-2015, 08:48 PM
It's definitely not reading that bit correctly, you wouldn't fail to verify every blank you had.

Only way I can think of to verify that you are still writing the proper data is to use another programmer or find someone to read a 5555 filled chip.

If you can't find anyone close I have no problem doing it.

Thanks for your help, bustedstr8. I'll consider your offer and maybe send you a couple chips to test, however that's not an option for every EPROM I write :)

From what I've read, this M27C160 chip works with having a bit more than 1 million of 16-bit sections, which are addressable individually. So, you enter a number, from 0 to 2^20-1, in binary form, on the A lines (pins 1-10 and 33-42), being 0 a voltage of 0 (or very near it) and 1 a voltage of about 5 (or close to it) and you get the bits that section (sector?) contains, on the D (Q) pins, which seem to be pins 14 to 30 on this EPROM model. Therefore, it seems my programmer is reading wrong always the 6th bit (starting from 0) on all the 16-bit sections it reads (it always read 0, no matter what). Am I getting this right? if so, maybe this thing is easily fixable...what do you guys think?

Thanks!!

bustedstr8
03-29-2015, 10:16 PM
Yes, it looks like D6 is always stuck at 0.
I have no idea how easy it would be to fix. If 27c160 verifies on another machine, you will probably be fine burning them, just not reading them.

aha2940
03-30-2015, 02:58 PM
Yes, it looks like D6 is always stuck at 0.
I have no idea how easy it would be to fix. If 27c160 verifies on another machine, you will probably be fine burning them, just not reading them.

Damn, this is a bummer. I had not in my plans that this programmer would break so soon. Funny that it works with smaller EPROMs, however it does not work with these 160s, and I just tested with an M27C322, same FFBF pattern. Chinese POS LOL. Just curious: can anybody please post a pic of their TOP3000 boards? I seem ho have found something funny on the bottom board of mine...

Thanks!!

P.S: Can anybody please recommend a programmer that will not break my bank but is trusty and good quality? just in case I have to replace this one completely.

purple
03-30-2015, 03:11 PM
On some eproms reading and writing voltages can be different, sometimes I get error when verifying but usually write is successful. Some 160 and 322 types need just a bit more than 12v, try different psu's and see if it does make difference also try not blank checking before you write. (maybe this makes them corrupt?)

mainman
03-30-2015, 04:08 PM
Damn, this is a bummer. I had not in my plans that this programmer would break so soon. Funny that it works with smaller EPROMs, however it does not work with these 160s, and I just tested with an M27C322, same FFBF pattern. Chinese POS LOL. Just curious: can anybody please post a pic of their TOP3000 boards? I seem ho have found something funny on the bottom board of mine...

Thanks!!

P.S: Can anybody please recommend a programmer that will not break my bank but is trusty and good quality? just in case I have to replace this one completely.

The only bullet proof programmer I can recommend is the chipmax2 but it will break your bank at $545 but it is well worth it if you use your programmer a lot, excellent quality.

Both the top and vp programmers are hit and miss, you go cheap and get what you pay for.

Fox1
03-30-2015, 04:14 PM
The only bullet proof programmer I can recommend is the chipmax2 but it will break your bank at $545 but it is well worth it if you use your programmer a lot, excellent quality.

Both the top and vp programmers are hit and miss, you go cheap and get what you pay for.

You can get a TopMax (Parallel port) for under $200 and it will do more than the Chipmax2. Of course, you need a parallel port hah

aha2940
03-30-2015, 04:20 PM
You can get a TopMax (Parallel port) for under $200 and it will do more than the Chipmax2. Of course, you need a parallel port hah

The parallel port is not an issue, I have an old laptop with parallel port (heck, it even has serial port!) dedicated for usage with the programmer. I'll research a bit about this.

Thanks!!

mainman
03-30-2015, 04:34 PM
You can get a TopMax (Parallel port) for under $200 and it will do more than the Chipmax2. Of course, you need a parallel port hah

Yeah but it's not a good idea to have a programmer that is locked to one machine, the fact that my programmer is locked to xp is the only reason I am updating it despite the fact it has served me faithfully for years and despite I still have multiple machines that run xp for legacy support.

That parallel programmer might seem like a good ideal now but it will show it limitations and age quickly to you.

channelmaniac
04-06-2015, 01:14 AM
You can get a TopMax (Parallel port) for under $200 and it will do more than the Chipmax2. Of course, you need a parallel port hah

AND you will need Windows XP.

The eeTools parallel port programmers are not supported with PCI slot parallel cards or with USB to Parallel adapters. You have to have an on-motherboard parallel port.

I have 2 of the TopMax programmers and just sold my 3rd for $245 locally. That's the discount I would get from eeTools on trading it in towards a TopMax 2, which I'll be getting very soon. :)

mainman
04-06-2015, 02:22 AM
AND you will need Windows XP.

The eeTools parallel port programmers are not supported with PCI slot parallel cards or with USB to Parallel adapters. You have to have an on-motherboard parallel port.

I have 2 of the TopMax programmers and just sold my 3rd for $245 locally. That's the discount I would get from eeTools on trading it in towards a TopMax 2, which I'll be getting very soon. :)

Holy, do you really need a $800 programmer channel. What is wrong with the chipmax2

Btw there is a remap work around to use legacy devices hardwired for old motherboards parallel port with pci cards but you would still be locked to xp.

Xavier
04-09-2015, 08:03 PM
Well then it sounds like your not selecting the right eprom or it's not selectable, your system/software isn't compatible with it or it's not getting enough power (in that order).

I know on mine (different then yours) there's some small one like the 2716 that the most common brand ones just won't work on my system.

Those Chinese erasers are weaksauce, just crank that timer all the way...it still won't erase half of them. Do it again half way to make sure most are erased or real quick check if they're erased then take out those ones and do the rest for half way. I think these things run on both 220/120v and at 120 it just give them enough juice to really do anything. I had big one for awhile something like a spectraline and it erased most everything in 5-7 minutes / held dozens of chips as well. The bulb burned out on it though and all the ones I could find cost more than I paid for the unit itself.

I always check erase, program and verify to make sure I have a trustworthy programming.

aha2940
04-09-2015, 08:34 PM
Well then it sounds like your not selecting the right eprom or it's not selectable, your system/software isn't compatible with it or it's not getting enough power (in that order).

I'm afraid I am selecting the right EPROM in the software (it is selectable), the software is made by the same people who made the programmer and it is getting power from the USB port + the external adapter. The eraser I have is chinese, but even EPROMs that were sold to me as new, show that same pattern. 1 or 2 chips, I doubt. I have tested so far 25 blank EPROMs, all show the same FFBF pattern. This thing is broken, it's not something I have made wrong. I'll get a new one when I can.

Thanks!!

Xian Xi
04-09-2015, 08:49 PM
Check the soldering to the socket on it. It could be a broken joint or something else.

The Webmiester
04-10-2015, 09:07 AM
I have a TOP3100 and a GQ-4X and I'm constantly switching between them for reading and writing different chips. For example, GQ-4X will not read old 2764, 27128 etc chips that I have. But then the TOP3100 won't write 27C040s well.

Have you tried manually setting the address and reading the bits? Put the chip on a breadboard, power it (no program power), set the address to 0, set the mode to 16-bit, and read off the data bits to see if your D6 bit is being written.

There's also always a chance that the D6 pin in the programmer isn't great, maybe try something like only inserting the chip half way.

Lastly, I know the GQ-4X has a mode where you can test each pin on the programmer, like turn them on and off and measure with a meter. Maybe the TOP programmer software has that too, I can't remember off the top of my head.