How do you know EEPROM chips are good

Brevor

Joined Apr 9, 2011
297
SamEricson; Please accept my apologies, it seems you are not the person I was refering to. You seem to be in good hands with dl324. I didn't mean to sound harsh, I just finished up a 60 hour workweek and had a few drinks in me.
 

Thread Starter

SamEricson

Joined Apr 25, 2015
196
So you never had any issues with the check sum verification when using different universal programmers? or they all the same for the check sum verification

Some universal programmers are Closed Loop verification and others are Open loop verification, any way to tell?
 

dl324

Joined Mar 30, 2015
9,335
So you never had any issues with the check sum verification when using different universal programmers? or they all the same for the check sum verification
I have never used checksum for any device I've programmed. I always assumed the data I was programming was correct and insured that devices I programmed matched that data. If I needed to check for file transmission errors, I'd use something like MD5. The software I use for my hobbyist programmers calculates a checksum, but the algorithm isn't described. I have source for an early version of the program, but was never inclined to check the algorithm because it wasn't a feature I cared about.
Some universal programmers are Closed Loop verification and others are Open loop verification, any way to tell?
If it isn't in the documentation, you can put a scope on VCC while the programmer is operating to see if you can see any changes, or you could put a peak detector on the supply if it's changing too quickly, slowly, or you can't get the signal to sync.

When I was checking my Stag programmer, I was able to measure VCC and determine that the blank check and program voltages were incorrect.

For all EPROMs larger than 32Kb, program margin is checked explicitly. It's open loop for 32Kb and smaller. The primary motivation for the change was to decrease programming time. 32Kb required a 50mS pulse for each address (204.8S), if they used the same algorithm for 64Kb, programming time would be almost 7 minutes. With the improved algorithm, it was reduced to 1.25 minutes (for typical devices). It could be longer for devices with bits that were "difficult" to program.
 
Last edited:

takao21203

Joined Apr 28, 2012
3,696
I never said I did, that is why i'm trying to learn the correct way of erasing, reading, writing, checksum . etc. about EEPROM chips and ERPOM chips.

Most technicians I have worked with don't know either it seems.

dl324 is teaching and training me the right way because of issues i have been having.
most techs i saw dont entertain lamentation to this degree they fizzle the thing in a sec or toss it.
Reading the datasheet and worry about margins? show off the returned ics + equipment, proof you are no trol
 

Thread Starter

SamEricson

Joined Apr 25, 2015
196
At one of my last jobs we used these universal programmers Unisite XPI and a Unisite 288

What are some common problems you have seen in your experience with bad eeprom chips and eprom chips? bad chips in circuit and out of circuit

What did the EPROM chips and EEPROM chips failed from in your past experience? any common failures

a peak detector on the supply
What kind of peak detector did you use or did you make one?

Mostly I use the "Hold function" on my DVM meter for a peak detector not sure If this is correct what i am doing.
 

MrChips

Joined Oct 2, 2009
19,768
EPROMs and EEPROMs are either good, bad or flaky.

If the chip is good, your circuit may work if everything else is ok.
If the chip is bad, your circuit will fail to work.
If the chip is flaky, your circuit will work sometimes and fail sometimes.

Without knowing what is programmed on the chip there is no way your programmer will tell you if the chip is good or bad, checksum or no checksum.

A programmer cannot diagnose a bad chip, exceptions noted below. The only thing you can do is to copy the contents of a good chip on to another good chip.

If the programmer fails to verify a blank chip then the chip is bad.
If the programmer fails to verify the contents written to the chip, then the chip is bad.
 

takao21203

Joined Apr 28, 2012
3,696
EPROMs and EEPROMs are either good, bad or flaky.

If the chip is good, your circuit may work if everything else is ok.
If the chip is bad, your circuit will fail to work.
If the chip is flaky, your circuit will work sometimes and fail sometimes.

Without knowing what is programmed on the chip there is no way your programmer will tell you if the chip is good or bad, checksum or no checksum.

A programmer cannot diagnose a bad chip, exceptions noted below. The only thing you can do is to copy the contents of a good chip on to another good chip.

If the programmer fails to verify a blank chip then the chip is bad.
If the programmer fails to verify the contents written to the chip, then the chip is bad.
i saw a few flaky pics havw one in working circit which first driftts then hangs or self resets. all the time its powered up.

its 8pic core circuit 4x TXCO 18mhz

one of the chips freaks.

another has a broken io bit.

i had a 10f200 wgich randomly wold latch up and start heating up

just a few examples
 

dl324

Joined Mar 30, 2015
9,335
If the programmer fails to verify a blank chip then the chip is bad.
If the programmer fails to verify the contents written to the chip, then the chip is bad.
You can also read the chip at low, nominal, and high VCC and compare the data. Any charge loss issue should manifest itself; unless charge loss is too severe. If you can acquire the data from another source (e.g. backups or other devices), you can load the correct data into the programmer and do a program verify.
 

dl324

Joined Mar 30, 2015
9,335
What did the EPROM chips and EEPROM chips failed from in your past experience? any common failures
The primary failure in EPROMs/EEPROMs is oxide failure. Writing creates high local electric fields and causes oxide defects to develop. These can trap charge and affect the ability of a bit to be erased. Oxide can also breakdown which affects the ability to hold charge. Other logic is subject to the same failure mechanisms as other CMOS/NMOS devices (oxide failure, ESD, etc).

What kind of peak detector did you use or did you make one?;
Just a modified voltage follower.

Mostly I use the "Hold function" on my DVM meter for a peak detector not sure If this is correct what i am doing.
The hold function would only capture peak if the button was pushed when the signal was at peak. Otherwise, it just holds the reading that was being displayed when the button was pressed.[/quote]
 

MrChips

Joined Oct 2, 2009
19,768
You can also read the chip at low, nominal, and high VCC and compare the data. Any charge loss issue should manifest itself; unless charge loss is too severe. If you can acquire the data from another source (e.g. backups or other devices), you can load the correct data into the programmer and do a program verify.
It is known practice to read back and verify ROMs at higher voltages to detect marginal programming. I think this is beyond the scope of what the TS is looking for.
 

dl324

Joined Mar 30, 2015
9,335
It is known practice to read back and verify ROMs at higher voltages to detect marginal programming.
Not to nit pick, but some ROMs are fuses or are "masked" (manufactured with a certain data pattern).
I think this is beyond the scope of what the TS is looking for.
Seemed to me that the OP was interested in increasing his knowledge. Just my opinion, I could be wrong...
 

Thread Starter

SamEricson

Joined Apr 25, 2015
196
Seemed to me that the OP was interested in increasing his knowledge. Just my opinion, I could be wrong...
Yes I am interested in increasing my knowledge and hope others will read this and help other members also

When there is RAM or ROM issues on a circuit board, sometimes techs will put in "Blank" EEPROM or EPROM chips and use a test program or a fluke 9010A to flood the Blank EEPROM chips with 1111's or 0000's. They are checking to see if the EEPROM chips are shifting data.

I have no idea why they would use blank EEPROM chips or how they can flood the EEPROM chips with 1111's and 0000's to check the shifting of data.

It's a way of testing EEPROM and EPROM chips in circuit.
 

dl324

Joined Mar 30, 2015
9,335
It's a way of testing EEPROM and EPROM chips in circuit.
In circuit test/program capability needs to be designed in. Otherwise, you need to remove the chips to test. If you try to power an EPROM to test it in circuit, you'll power up the rest of the board. Even if you disconnect the EPROM power pin from the board, you'll find that any inputs being driven by the EPROM will power up any CMOS logic on the board through the input protection diodes. If you power the board, you'll be fighting other logic for control of the EPROM.
 

Thread Starter

SamEricson

Joined Apr 25, 2015
196
There was some circuit boards I have found in my past that have "Inverted the EEPROM code" and used Logic IC chips with inverted inputs or outputs that is why they inverted the code in the EEPROM chips.

The schematics didn't use the inverted logic chips , but the repair person swapped them out and used inverted logic chips which the repair person inverted the EEPROM chip.

How do you invert the EEPROM file code? so the starting address is the ending address
Invert the ROM contents
Invert the ROM entries
 
Last edited:

MrChips

Joined Oct 2, 2009
19,768
You are going around in circles.

There is no way to check an EEPROM if you do not have information about its contents.
 

dl324

Joined Mar 30, 2015
9,335
The schematics didn't use the inverted logic chips , but the repair person swapped them out and used inverted logic chips which the repair person inverted the EEPROM chip.
I can't fathom why someone would do that; well, I could, but it would be an end of the world scenario...

How do you invert the EEPROM file code? so the starting address is the ending address
Invert the ROM contents
Invert the ROM entries
If I had to invert data, I'd write a program to flip the bits. I'd do the same to reverse addresses, but I really can't fathom why you would ever need or want to do that.
 

takao21203

Joined Apr 28, 2012
3,696
weird thread mixing up EPROM and EEPROM and talking to test EPROM in circuit

EPROMS are out of fashion for quite a while.

every whacko EEPROM programmer will give error message.

There was some circuit boards I have found in my past that have "Inverted the EEPROM code" and used Logic IC chips with inverted inputs or outputs that is why they inverted the code in the EEPROM chips.

The schematics didn't use the inverted logic chips , but the repair person swapped them out and used inverted logic chips which the repair person inverted the EEPROM chip.

How do you invert the EEPROM file code? so the starting address is the ending address
Invert the ROM contents
Invert the ROM entries
eeprom + logic ICs? also weird.

you wouldnt mind to post schematic?
 

#12

Joined Nov 30, 2010
18,167
weird thread mixing up EPROM and EEPROM and talking to test EPROM in circuit

you wouldnt mind to post schematic?
This TS will never post a schematic or a brand name or a model number. He wants Universal Truths that apply to All Circuits Everywhere, so there is no benefit to saying which circuit he is talking about.
 

dl324

Joined Mar 30, 2015
9,335
EPROMS are out of fashion for quite a while.
There's still a lot of equipment using them. I have quite a few pieces of equipment from the 70's that are still quite useful. If the EPROMs lost their data, that equipment would be useless.

Many auto and motorcycle owners would agree that they're still useful and required. My "old" BMW wouldn't run without a 27C256...
 

takao21203

Joined Apr 28, 2012
3,696
This TS will never post a schematic or a brand name or a model number. He wants Universal Truths that apply to All Circuits Everywhere, so there is no benefit to saying which circuit he is talking about.
because there is nothing? i was thinking, it might become apparent some day.

thats why i like weird circuits someone just copies it but doesnt know all details about it.
 
Top