Searching \ for 'EPROM erasing question' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/index.htm?key=eprom+erasing+question
Search entire site for: 'EPROM erasing question'.

Truncated match.
PICList Thread
'EPROM erasing question'
1998\11\06@121754 by keller

flavicon
face
Is it possible for an EPROM (16C63) to check out as blank, program, and
verify correctly but not be fully erased?

I am working with the 16C63JW.  I have 6 that I rotate through my eraser as
I am programming,  occasionally it seems like things just go screwy.  Am I
not leaving them in the eraser long enough?  Please let me know your
thoughts on this.

Thanks
Steven Keller

1998\11\06@125050 by Bob Blick

face
flavicon
face
> Is it possible for an EPROM (16C63) to check out as blank, program, and
> verify correctly but not be fully erased?

Yes. Erase at least twice as long as it takes to pass a blank test.
Otherwise weird stuff will happen(oscillator not starting, lockups, etc).

I have the timer in my Datarase ][ set at 20 minutes.

-Bob

1998\11\06@125054 by Engineering Department

flavicon
face
>Is it possible for an EPROM (16C63) to check out as blank, program, and
>verify correctly but not be fully erased?
>
>I am working with the 16C63JW.  I have 6 that I rotate through my eraser as
>I am programming,  occasionally it seems like things just go screwy.  Am I
>not leaving them in the eraser long enough?  Please let me know your
>thoughts on this.


I think so Steve.  I've had the same problem with the little '54s.  For some
reason 15 minutes under a 4w lamp will generally give me a chip that tests
as "blank" but screws up when programming it.  Using 30 minutes I don't
recall running into the same problem until the PIC gets quite old (lotsa
programming/erase cycles -- I don't know the exact number but it's over 75).

Cheers,

Win Wiencke
Image Logic Corporation
spam_OUTImageLogicTakeThisOuTspamibm.net

1998\11\07@142745 by Peter L. Peres

picon face
On Fri, 6 Nov 1998, Engineering Department wrote:

> recall running into the same problem until the PIC gets quite old (lotsa
> programming/erase cycles -- I don't know the exact number but it's over 75).

It should last well over 1000x. Clean the 'quite old' 54JW window
mercilessly with alcohool and put it in to erase for an hour, then try
again. The thin grime that accumulates on top of the window is extremely
efficient at stopping UV.

Peter

1998\11\07@142802 by Peter L. Peres

picon face
On Fri, 6 Nov 1998, Steven Keller wrote:

> Is it possible for an EPROM (16C63) to check out as blank, program, and
> verify correctly but not be fully erased?

Yes. I let them sun themselves one hour every 5 to 10 erase/program cycles
to avoid it. When I did not do that I did see strange things after a few
cycles (no pattern). I used 54JW, 711JW and 508JW. By the time I moved to
711 and 508 I already used my one hour sunbath method and never saw
strange effects again.

I suppose that an incmoplete erasure leves some charges that can migrate
back and forth after a while. I have also designed & built my own 54
programmer and wrote the soft for it so I did a lot of checking on erase
patterns, what is erased first etc. There is almost no pattern, you can't
guarantee that the fuses are erased last (on a 54JW), and in general, just
leave them in there long enough ;(

Peter

1998\11\09@065218 by Caisson

flavicon
face
> Van: Steven Keller <.....kellerKILLspamspam@spam@theramp.net>
> Aan: PICLISTspamKILLspamMITVMA.MIT.EDU
> Onderwerp: EPROM erasing question
> Datum: vrijdag 6 november 1998 18:15

Hello Steven,

> Is it possible for an EPROM (16C63) to check out as blank, program, and
> verify correctly but not be fully erased?

Yes.  It depends on the time between Erasing and Programming the part.  If
you take a part directly outof a (working :-) Eraser and than Blank-check
the part it could show up as erased, but some minutes later it could fail.
I'ts sort of "bouncing back". That's why (combined with other reasons) most
people "over-erase" EPROM memory.

If you do experience "strange" behaviour, try verifying the part.  If it
fails you could be experiencing "bounce back".  By the way ... Do you
verify with 6 & 4 volts or with 5 volts only ?  (PicStart + only verifies
with 5 volts.)  "Professional" programmers check with 6 & 4 volts to have a
bigger error-margin.

Greetz,
 Rudy Wieser

1998\11\16@172942 by John Payson

flavicon
face
> Is it possible for an EPROM (16C63) to check out as blank, program, and
> verify correctly but not be fully erased?

|If you do experience "strange" behaviour, try verifying the part.  If it
|fails you could be experiencing "bounce back".  By the way ... Do you
|verify with 6 & 4 volts or with 5 volts only ?  (PicStart + only verifies
|with 5 volts.)  "Professional" programmers check with 6 & 4 volts to have a
|bigger error-margin.

You should blank-check any PIC device with a voltage at least
as low as any voltage from which the device will be operated,
if not lower.  The ROM cells seem to be more sensitive to their
"programming" state when operated on low voltages than higher
voltages, so an address which was programmed "0" and then mostly
erased may show up as erased when read at +5V and as "programmed"
when read at 2.5V.

Ironically, while Microchip emphasizes the importance of checking
for write/erase margins for production jobs, it's actually the J/W
parts used in development where erase margin is a real issue.  OTP
devices start out thoroughly blank (they'd better!) so the test for
erase margin is largely superfluous.  Window parts, however, may of-
ten be less-than-thoroughly erased and can cause bizarre behavior if
this isn't detected.

More... (looser matching)
- Last day of these posts
- In 1998 , 1999 only
- Today
- New search...