Re: How to care for an old computer?

[mjmahon@aol.com (Michael J. Mahon)]

Wayne Stewart wrote:

>Bill Garber wrote:
>
>> What would cause a ROM to fail other than
>> material failure. What is bit rot? Do you mean
>> like with an eprom? Small amounts of light getting
>> in over many years eventually switching a 0 to 1?
>> Never had that happen.
>
>The orinal ROMs used in II+, IIe and so forth I'm not really worried
>about. EPROMs are a different story. I've had a number of EPROMs fail,
>all covered and likely never exposed to UV. EPROMs do lose their
>porograming over time, even when kept in the dark. Fortunately EPROMs
>are easily duplicated. As long as you can get the same part. Some
>haven't been manufactured in years though.

EPROMs store data by trapping tiny charges in intrinsic (insulating)
silicon.

The reliability with which they maintain those charges is dependent
on 1) the degree to which they were programmed (the amount of
trapped charge), 2) the rate of leakage of the charge.

The rate of leakage is dependent primarily on 1) the purity of the
silicon, 2) the absolute temperature, and 3) the incident flux of
ionizing radiation (usually UV, but x-rays and gamma rays will
do nicely to erase an EPROM).

So there are several factors that affect the rate of an EPROM
"losing its mind".  The good news that a high-quality EPROM,
shielded from light and kept reasonably cool, can be expected
to maintain its programming for decades.

Of course, not all EPROMs meet those conditions.  ;-(

-michael

Check out amazing quality sound for 8-bit Apples on my
Home page:  http://members.aol.com/MJMahon/