A way to erase OTP components?

From: Bob Shannon <bshannon_at_tiac.net>
Date: Wed Jan 22 17:50:01 2003

I currently work for a company that makes the X-Ray supplies for airport
scanners. We have the full X-Ray
emitter heads at test loads for the -180KV, 15ma supplies in-house.

These scanners place the high voltage under computer control, they don't
run at any one preset level. I gather
that the operating point shifts as the tubes age, and perhaps also with
the material being examined under software
control.

But crank one of these babies up to full power, and the X-rays will
generate very high charge levels in the die and
cook the part.

The question now becomes, does the date go away before the transistors do?

On a related note, I recently erased a (windowed) 27C512 using a UV led.
 I did'nt think it would work, but it did.

Eric Smith wrote:

>Bob Shannon wrote:
>
>>X-rays will erase the part at high energy levels, but this will also
>>degrade the part to the point where it
>>will fail soon, or die during erasure.
>>
>
>A friend had a large quantity of programmed 27C512 OTP EPROMs, with
>contents of no use to him. He tried to erase them using a commerical
>X-ray machine, of the type used for inspecting assembled circuit boards
>(to check solder joints on BGA packages) and was unable to produce any
>noticable change to the part.
>
>This subject has been discussed on sci.electronics multiple times in the
>past, and the consensus was that the frequency of Xrays is such that
>they have no direct effect on the stored data. But at very high
>exposures it is definitely possible to damage the part.
>
>Note that the exposure levels used for baggage inspection at airports
>is far too low to cause any measurable effect on EPROMs. They are
>generally set to substantially lower levels than medical X-rays
>equipment.
>
>The normal erasure procedure for EPROMs is to use short-wave UV light,
>nominally at a wavelength of 2537 angstroms (about 254 nm). Unlike the
>more common long-wave UV (black lights), this *is* dangerous to your eyes;
>never look into the light. The light bulbs and the window on the EPROM
>are made from quartz rather than glass, because glass blocks most of the
>short-wave UV. Suitable bulbs are commonly sold for germicidal use.
>
>In the mid-1980s, at least one Japanese semiconductor company offered
>plastic windowed EPROMs. The package was a normal epoxy DIP package,
>except that a square area in the center was molded from a different
>plastic. The window was translucent (not transparent) to visible light.
>These were less expensive than the usual ceramic windowed parts, but
>more expensive than OTP parts. However, they could only be erased a
>small number of times because the plastic window becomes opaque to the
>shortwave UV with prolonged exposure.
>
>I'd heard of these plastic windowed packages at the time, but never
>actually saw one until I purchased a coin-op video game board at an
>auction.
>
>Eric
>
>
>
>
>
Received on Wed Jan 22 2003 - 17:50:01 GMT