Downright Criminal RAM

[Tristan Mumford <xtristan.xmumford@xgmail.xcom>]

As some of you may know I have been trying to resurrect a Clone ][ with an
unspecified problem. I was actually kindly given a replacement board to
fit.
However as I was pulling apart the clone a simple test occurred to me to
rule out teh two biggest suspects. The power supply and the RAM.

A little history:
Ages back the clone was put in a shed for many a year. I cannot even recall
why I put it there in the first place.

Late last year I got hold of it again and tried to power it up. Anyway it
didn't really work. On top of that more and more RAMs seemed to be going
heater mode on it.
Besides that the power supply was reading erratic voltages. Most notably
on -5 and -12v.

I then checked it with a lightly modified AT power supply. All voltages were
good but the screen looked different. With the clone supply lots of
exciting garbage would pulsate on the screen. With the AT supply all was
white.

An interesting result.

Now lets's just jump forward a bit.
It looked like it was either a dodgy powersupply causing RAM to fall like
dominoes or something else that I suspected but considered a longshot.
Then the idea of a dead easy test occurred to me. Just pull out all the RAM
and check the voltages.

So I did exactly that. All voltages were as they should be. Satisfied with
this I added some cards and tested again. Still all voltages were fine.

Just now I completed populating the board with RAM taken from another clone
board. It booted.

If you are wondering what happened so am I. However I have a theory which
aligns with behaviour I have observed in other circumstances.

The 4116 RAMs use just about every voltage rail available. Both positive an
negative. Apart from possibly -12v. To be honest I couldn't be bothered
checking.
Computer powersupplies generally have weak negative rails. I.e.  they supply
very little current.

What I believe happened is this:

One day during a poweron, or powercycle or something one of the 4116 RAMs
developed an internal short. They are quite old after all.
This internal short changed the internal bias of the chip and allowed some
current to flow between at least some of the different potential rails.
The shorting of the 4116 caused the internal potential of the other 4116
RAMs to change because of the clobbering / positive energising of the
negative rails (or it just shorted out into the address / databus). The
shift in voltages caused more to fail and so on.

The reason the AT supply was reading normal voltages was it had negative
rails strong enough to overpower the draw incurred by the damaged chips.

Anyway, that's my story.

Tristan.

-- 
-----> http://members.dodo.com.au/~izabellion1/tristan/index.html <-----
=====       It's not pretty, it's not great, but it is mine.       =====