[Coco] Re: EPROM eraser wanted

[Gene Heskett]

On Sunday 16 April 2006 02:14, Stephen H. Fischer wrote:
[...]
>> That one should be a quartz tubed lamp, most of the germicidals
>> were. I have NDI how long it will take it to wipe a 2764 though.  It
>> takes more than the minimum to get all $FF's out of it, but it can
>> be overdone according to my reading on that quite a few years ago
>> now. The effect is that eventually it won't program regardless of
>> how many write pulses you bang it with.
>
>I am having a problem understanding your comments.
>
>Are you saying that the G8T5... will not erase EPROMS?

No, read that again, I said it was a quartz tube, meaning it will work, 
I just don't know how long it would take.

[...]

>You can convince many people that a simple task is one that requires
> huge $$$$.
>
>For the limited number of EPROM types that CoCo'ers would want to
> program the parts cost is trivial.
>
>Five dollars is too much I suspect even today if the very old parts
> have not risen in price. Using today's parts, it may be a simple one
> chip solution (PIC).
>
>While I have not constructed (As far as I remember) the simple
> programmer that I determined would be all that is necessary and I
> have not written the software I can supply the basis of that
> conclusion (A more parts intensive burner that I have and used
> several times for the SYM-1). Both the hardware and software needed
> is very simple once you understand the simple task. The CoCo is
> better for the task than a PC and if I remember correctly, I planned
> to write the software in BASIC.

Mine had to interface with a z80 board with an ss50 buss IIRC, and had 3 
latches to hold the address and data during the write pulse, and a 
timer to control the pulse duration, read-write operations being 
switched by turning an lm317 up and down to get the write voltage, then 
25 volts but soon lowered to 17 volts by newer chips coming online.

Write pulse duration was set by the timer, the timer being triggered by 
the write strobe to the data latch with a couple of gates to delay so 
the data was well setup by the time the lm317 went up.  IIRC that put a 
hold on the z-80 until the write pulse was over, at which point my 
software did a trial read to see if the data was good, if it was then 
it shifted the count of how many pulses it had taken to get that good 
read, and did half that many more write pulses blind as recommended by 
the chip people at the time.  The it went on to the next address and 
repeated the routine.

And from that you should be able to build a burner.  I remember one 
gotcha with the z-80, if all other access sigs were valid except the 
read, then it was going to be a write, since the z-80's write pulse was 
considerably delayed compared to its read pulse, so those conditions 
were then used to preset the burner for an upcoming write operation.  
The write pulse itself was then ignored.

>--
>Stephen H. Fischer

-- 
Cheers, Gene
People having trouble with vz bouncing email to me should add the word
'online' between the 'verizon', and the dot which bypasses vz's
stupid bounce rules.  I do use spamassassin too. :-)
Yahoo.com and AOL/TW attorneys please note, additions to the above
message by Gene Heskett are:
Copyright 2006 by Maurice Eugene Heskett, all rights reserved.