Author Topic: W27C512 EEPROM Frustrations  (Read 7154 times)

0 Members and 1 Guest are viewing this topic.

Offline CRCLARKETopic starter

  • Contributor
  • Posts: 16
  • Country: ca
W27C512 EEPROM Frustrations
« on: January 05, 2020, 06:31:16 am »
I am relatively new to eeprom programming, but i am gaining plenty of experience.... I have been working with the w27c512 and I am having some real struggles with erasing the eeprom. I have built an Arduino breadboard programmer based on Ben Eaters, along with an erasing module, wired it properly and followed the voltage requirements. I tested it and it works well.

The problem is, out of 10 chips, only 3 work properly, 3 won't erase, and 4 went up in smoke during erasing. I am not sure whats wrong, are they defective chips?  am I not erasing properly? I do not use the verify algorithm, is it required? 3 erase just fine. I also don't use any timing to erase, just 2 push buttons. One for VPP and the other for CE...... I tried erasing at 12v and it works on the good chips, but had no effect otherwise.

I am once again 1 short on chips and the project is shelved.  |O :-[

I have just ordered a 20 pack of w27c512, luckily there pretty cheap but the wait sucks. I really hope I don't have this much trouble with the next batch

Has anyone had this trouble before?

Are there any tips or tricks to erasing these eeproms?

Thanks!
 

Offline Nerull

  • Frequent Contributor
  • **
  • Posts: 694
Re: W27C512 EEPROM Frustrations
« Reply #1 on: January 05, 2020, 08:05:31 am »
If you're burning chips than you likely do not have something wired correctly.
 

Offline Ian.M

  • Super Contributor
  • ***
  • Posts: 13218
Re: W27C512 EEPROM Frustrations
« Reply #2 on: January 05, 2020, 09:22:50 am »
The W27C512 datasheet says it requires +14.0V +/-0.25V Vpe to erase on /OE and A9,  will draw 30mA from Vpe  and the abs. max. voltage those pins can tolerate is only +14.5V.

That's not a lot of margin between a successful erase and blowing the chip, so your Vpe switching circuit must provide clean edges free from overshoot, ringing etc.  Also the spec says the /CE pulse to trigger erase is 100ms +/- 5ms, so there's no way you'l meet those specs with a simple pushbutton. 

At the minimum you'll need a debounce circuit for the button, an active Vpe switching cicuit that can handle >30mA with negligable voltage drop, and some monostables to sequence Vpe and /CE with correct timing.    Best bet would be to do the sequencing and timing with the Arduino.
« Last Edit: January 05, 2020, 09:42:11 am by Ian.M »
 

Offline CRCLARKETopic starter

  • Contributor
  • Posts: 16
  • Country: ca
Re: W27C512 EEPROM Frustrations
« Reply #3 on: January 05, 2020, 05:25:11 pm »
I am pretty sure it is wired correctly because I have erased several eeproms successfully, and the eeproms that do work I have erased several times successfully.

I am operating within the voltage tolerances, and have even successfully erased at 12v. I am aware of the 100us CE pulse but it does not specifically state a maximum time for VPP or A9. No issues reprogramming the chips, but only the ones that survived erasing made it to reprogram. and they have repeatedly worked as I was expecting. but it is total hit and miss, mostly miss. I feel it has something to do with that A9 voltage. I wish I knew how it worked, I added a current limit resistor to A9, didn't help, though it made the failures less catastrophic.

I can setup some timing for the erase, but I am skeptical about there being a difference....... and I don't think a denounce would help, because so what if it erases multiple times, but I can test it.

3 of the chips work reliably with my setup.... so far anyway, maybe there just tough, but I would expect them all to be identical

I will have to wait till new chips arrive to test because the chips that work, always work and the chip that don't, don't respond to anything so I have no way to differentiate the test

So I got some time to brainstorm till the new ones show up, so lay it on me!



 

Offline james_s

  • Super Contributor
  • ***
  • Posts: 21611
  • Country: us
Re: W27C512 EEPROM Frustrations
« Reply #4 on: January 05, 2020, 05:29:45 pm »
Buy a TL866 and use that, they're only about 40 bucks and will easily save you money in the long run if you are blowing chips.

Also where are you getting the EEPROMs? If they're from China they're likely pulls, many might not be EEPROMs at all. I once bought a couple of EEPROMs that turned out to be random 28 pin ICs that had been blacktopped and re-marked, I wiped them with acetone and the black paint came off the top.
 

Offline wraper

  • Supporter
  • ****
  • Posts: 17952
  • Country: lv
Re: W27C512 EEPROM Frustrations
« Reply #5 on: January 05, 2020, 05:44:16 pm »
If you bought them from China probably you got mix of different ICs, most of them being OTP EPROM. It's quite often that old marking is removed and new marking is applied. I once ordered exactly the same UV EEPROMs from the same seller twice with one week difference. First batch was what it should be, second batch had much larger die. Turned out to be way older IC from different vendor and with  21V programming voltage instead of 12.5V. The thing was that despite differences in packaging, both had exactly the same marking with the same date code. Also happened with ATTINY24A which had the same marking on top but actually were mix of ATTINY24 and 24A. Also although they had the same lot/date code on top, on the bottom side it was different. While normally it should match. Turned out to be blacktopped and remarked. So even if marking does match, check if there are no differences in packages, if any marking on bottom side does match if they have the same date code on top. Also try rubbing them with acetone or some other organic solvent and see if marking comes off.
 

Offline wraper

  • Supporter
  • ****
  • Posts: 17952
  • Country: lv
Re: W27C512 EEPROM Frustrations
« Reply #6 on: January 05, 2020, 05:53:48 pm »
They are poorly visible on photo but seems some have dimple on corner and some doesn't. Also here for example is an ebay listing with obviously remarked IC on the picture. https://www.ebay.com/itm/10PCS-EEPROM-IC-WINBOND-DIP-28-W27C512-45Z-W27C512-45/123882775460?hash=item1cd7fd23a4:g:~9QAAOSwLpRdXpAi
Others may look much more sophisticated buy still remarked.
 

Offline CRCLARKETopic starter

  • Contributor
  • Posts: 16
  • Country: ca
Re: W27C512 EEPROM Frustrations
« Reply #7 on: January 05, 2020, 05:59:19 pm »
That's interesting because that is basically what happened
If you bought them from China probably you got mix of different ICs, most of them being OTP EPROM. It's quite often that old marking is removed and new marking is applied. I once ordered exactly the same UV EEPROMs from the same seller twice with one week difference. First batch was what it should be, second batch had much larger die. Turned out to be way older IC from different vendor and with  21V programming voltage instead of 12.5V. The thing was that despite differences in packaging, both had exactly the same marking with the same date code. Also happened with ATTINY24A which had the same marking on top but actually were mix of ATTINY24 and 24A. Also although they had the same lot/date code on top, on the bottom side it was different. While normally it should match. Turned out to be blacktopped and remarked. So even if marking does match, check if there are no differences in packages, if any marking on bottom side does match if they have the same date code on top. Also try rubbing them with acetone or some other organic solvent and see if marking comes off.

That's interesting because that is basically what happened, I first ordered 5 from china (everything comes from china) and had 2 bad? chips, so i then ordered 5 more about 2 weeks later from the same seller, and I think only 1 actually worked, which brings us to now, but the numbers on the chips are etched in, so I don't think they are repainted or rebranded. So the new batch I ordered from another seller, but still from china.

Does anyone know where to get eeproms NOT from china?
 

Offline wraper

  • Supporter
  • ****
  • Posts: 17952
  • Country: lv
Re: W27C512 EEPROM Frustrations
« Reply #8 on: January 05, 2020, 06:04:45 pm »
but the numbers on the chips are etched in, so I don't think they are repainted or rebranded.
They are marked with laser, which I can do at home. Old marking it either scrubbed with very strong solvent if it was originally paint, or sanded and blacktopped if it was marking made by laser. Then new laser marking applied.
 

Offline wraper

  • Supporter
  • ****
  • Posts: 17952
  • Country: lv
Re: W27C512 EEPROM Frustrations
« Reply #9 on: January 05, 2020, 06:12:27 pm »
Clueless naivety:





 

Offline wraper

  • Supporter
  • ****
  • Posts: 17952
  • Country: lv
Re: W27C512 EEPROM Frustrations
« Reply #10 on: January 05, 2020, 06:26:55 pm »
 

Offline Bud

  • Super Contributor
  • ***
  • Posts: 7276
  • Country: ca
Re: W27C512 EEPROM Frustrations
« Reply #11 on: January 05, 2020, 06:30:27 pm »
There is your problem. Buy from a reputable source.
Facebook-free life and Rigol-free shack.
 

Offline Ian.M

  • Super Contributor
  • ***
  • Posts: 13218
Re: W27C512 EEPROM Frustrations
« Reply #12 on: January 05, 2020, 06:42:56 pm »
Implement ID readout with A9 pulsed to 12V (see Table of Operating Modes' on page 7 and left side of 'Erase Waveform' diagram on page 13).  That should be non-destructive on any JEDEC compatible EPROM or EEPROM >64K.   

Before attempting to program a 'new' chip, check the device ID.  If you don't get the correct ID bytes, DA 08, its a remarked fake, not a genuine Winbond W27C512.  If it lets the holy smoke out, then its probably not even a non-volatile memory chip!  If it already has data in it, its recycled not new.
 

Offline CRCLARKETopic starter

  • Contributor
  • Posts: 16
  • Country: ca
Re: W27C512 EEPROM Frustrations
« Reply #13 on: January 05, 2020, 06:52:31 pm »

Upon closer inspection of the chips, maybe? probably? the have been rebranded. There are some definite physical variations in the chips, but even the working ones have some variances.

but knowing this information doesn't help a lot....

so i guess the lesson is:

There is your problem. Buy from a reputable source.


but none of shady sellers admit to being "non-reputable"

Where do you get Your eeproms from?

Implement ID readout with A9 pulsed to 12V (see Table of Operating Modes' on page 7 and left side of 'Erase Waveform' diagram on page 13).  That should be non-destructive on any JEDEC compatible EPROM or EEPROM >64K.   

Before attempting to program a 'new' chip, check the device ID.  If you don't get the correct ID bytes, DA 08, its a remarked fake, not a genuine Winbond W27C512.  If it lets the holy smoke out, then its probably not even a non-volatile memory chip!  If it already has data in it, its recycled not new.

Hey, that's a good idea!!!

also, less than half of the chips were erased when i received them
« Last Edit: January 05, 2020, 07:12:36 pm by CRCLARKE »
 

Offline wraper

  • Supporter
  • ****
  • Posts: 17952
  • Country: lv
Re: W27C512 EEPROM Frustrations
« Reply #14 on: January 05, 2020, 06:57:24 pm »
Clearly counterfeit. Packages are different but marking is exactly the same. Genuine components from different batches cannot have exactly the same lot, date code and other information (bottom row). Some ICs are obviously salvaged since there are soldering signs on their leads. Certainly try rubbing them with acetone, some look like they could be sanded and blacktopped.
« Last Edit: January 05, 2020, 06:59:46 pm by wraper »
 

Offline wraper

  • Supporter
  • ****
  • Posts: 17952
  • Country: lv
Re: W27C512 EEPROM Frustrations
« Reply #15 on: January 05, 2020, 07:12:39 pm »
but none of shady sellers admit to being "non-reputable"
On ebay there are basically none, especially in China. If you get genuine part one time, does not mean you get a good part next. Also most of counterfeits "work"*, so most hobbyist buyers don't even have any clue they were duped.

*"work" does not mean they meet specifications or work properly in all applications they should. For example audio opamps like AD797 sell like hot cakes on ebay but barely any of them are genuine. But most buyers leave positive feedback. And even if you have genuine parts, it's really hard to compete against 50 sellers with counterfeit crap. Barely anyone wants to buy at genuine component price. And the worst part is that even if you buy at high price does not mean you'll get genuine part. Though chances are certainly much higher.
« Last Edit: January 05, 2020, 07:16:46 pm by wraper »
 

Offline james_s

  • Super Contributor
  • ***
  • Posts: 21611
  • Country: us
Re: W27C512 EEPROM Frustrations
« Reply #16 on: January 05, 2020, 07:21:44 pm »
If you get parts that are clearly counterfeit, submit an item not as described claim with ebay and get your money back. The counterfeit sellers proliferate because hardly anyone bothers to do this. Just be 100% sure so you don't falsely accuse an honest seller.

With old parts like DIP EEPROMs it can be really hard to find a legit source for genuine parts. I have no problem paying for used pulls if they're sold as such. What bothers me are all the sellers passing them off as brand new.
 

Offline CRCLARKETopic starter

  • Contributor
  • Posts: 16
  • Country: ca
Re: W27C512 EEPROM Frustrations
« Reply #17 on: January 05, 2020, 08:32:40 pm »
I watched the video (posted above) about counterfeit components........ and it raised 2 questions to me.....

1. Why relabel something other than what it is? After all, it already has a label and proper function, it would seem easier to sell an orange as an orange rather than trying to paint it to look like an apple.

2. Unless the they have a fully automated system, it seems impossible to profit from manual chip salvage in bulk. If it took only a minute to refurbish each component, then even at a $1 each per component, the labor required is BARELY worth the return.

I can buy almost any new component or even IC's for a $1 or less.

I am pretty sure a person could sell salvage components as is and with less effort , and make better money. why would anyone refurbish a capacitor? for money no less?   :horse: :palm:
 

Offline Bud

  • Super Contributor
  • ***
  • Posts: 7276
  • Country: ca
Re: W27C512 EEPROM Frustrations
« Reply #18 on: January 05, 2020, 08:42:29 pm »
There is your problem. Buy from a reputable source.


but none of shady sellers admit to being "non-reputable"

Where do you get Your eeproms from?

I mean industrial parts suppliers, Digikey, Mouser, Arrow, Element14, and such. If you get your parts from random sellers on .bay and .express, you will always be in situations like the one you are in right now, that is - you will not know if the problem relates to bad parts or to your designs. You can eliminate the parts variable in the equation and focus on designs.
Facebook-free life and Rigol-free shack.
 

Offline CRCLARKETopic starter

  • Contributor
  • Posts: 16
  • Country: ca
Re: W27C512 EEPROM Frustrations
« Reply #19 on: January 05, 2020, 08:51:21 pm »
DigiKey is my favorite site! They are not always the cheapest but they are reliable and best of all they have quick shipping. Especially for the average Canadian Eskimo

but they do not offer a lot of options for eeproms.
« Last Edit: January 05, 2020, 08:55:32 pm by CRCLARKE »
 

Offline wraper

  • Supporter
  • ****
  • Posts: 17952
  • Country: lv
Re: W27C512 EEPROM Frustrations
« Reply #20 on: January 05, 2020, 09:03:05 pm »
1. Why relabel something other than what it is? After all, it already has a label and proper function, it would seem easier to sell an orange as an orange rather than trying to paint it to look like an apple.
Because you can sell counterfeit (functionally similar) or salvaged component as new and genuine. That's besides cases when internals do not match in function. Even in those cases you can "create" what's currently in demand.
Quote
then even at a $1 each per component
Done by cheap and fast labor. If you create $100 of components per hour with $10 labor, it's a quite decent profit. In cases when components are counterfeited from scratch, not salvaged from e-waste, process is mostly automated like in normal IC production. Don't forget that you can buy stuff from China for less than $1 including shipping. Compared to that process of counterfeiting is a gold mine.
Quote
I can buy almost any new component or even IC's for a $1 or less.
There are tons of counterfeit components sold for <5 cents a piece.
 

Offline wraper

  • Supporter
  • ****
  • Posts: 17952
  • Country: lv
Re: W27C512 EEPROM Frustrations
« Reply #21 on: January 05, 2020, 09:05:12 pm »
but they do not offer a lot of options for eeproms.
Because you want to buy obsolete stuff no longer in production. Your best bet is compatible component from smaller manufacturer which produces alternatives for obsolete components.
why would anyone refurbish a capacitor? for money no less?   :horse: :palm:
Because nobody will buy used with few exceptions like producers of counterfeit power bricks. But don't expect they will pay the same price.
« Last Edit: January 05, 2020, 09:45:26 pm by wraper »
 

Offline wraper

  • Supporter
  • ****
  • Posts: 17952
  • Country: lv
Re: W27C512 EEPROM Frustrations
« Reply #22 on: January 05, 2020, 09:18:52 pm »
Actually recently I found counterfeit (e-waste) 330uF 450V capacitors in power supplies used in elemental analyzers which cost around $30k. Made by German manufacturer (fuck you Eltra https://www.eltra.com/). Although those idiots outsourced PSU to some dodgy Chinese manufacturer.
 

Offline CRCLARKETopic starter

  • Contributor
  • Posts: 16
  • Country: ca
Re: W27C512 EEPROM Frustrations
« Reply #23 on: January 05, 2020, 10:11:11 pm »
i can see the value in some components, but for the most part it seems like setting up to counterfeit money, but only producing dollar bills.....

with a bit more effort you could counterfeit something worth the profit, or just make an authentic product......

and i have no issues with used components, i use them often in repairs, but i know they are used and they work just fine... usually, but that is the risk of used. 
 

Offline amyk

  • Super Contributor
  • ***
  • Posts: 8526
Re: W27C512 EEPROM Frustrations
« Reply #24 on: January 05, 2020, 10:45:08 pm »
Besides the discussion about fake parts, I would also recommend investing in an oscilloscope. Even genuine parts are unlikely to tolerate timing and voltage outside of spec. "Externally timed" EEPROMs like these can be easily damaged if the erase/program pulses are too long.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf