RE: En:RE: Parallel Port Ram Tester from Brazil!

> -----Original Message-----
> From: owner-techtoolslist@flippers.com [mailto:owner-
> techtoolslist@flippers.com] On Behalf Of John Robertson
> Sent: Monday, December 29, 2003 9:49 PM
> To: techtoolslist
> Subject: Re: En:RE: Parallel Port Ram Tester from Brazil!
>
> What a great project!
>
> The RAM pattern test used by Fluke might be of interest here. It will
> identify stuck address bits - but I have no idea of exactly the
process
> used to generate the Fluke RAM test. Perhaps someone else here might
have
> some algorithms that they can recommend?
>
> I was just testing an Asteroids today and the 9010 showed a DCD error
(was
> a failing A10 on the Vector RAM) - which proved to be a problem with a
> replacement socket. RAM tests must be able to check that address lines
are
> all working as well as the data can be read out...so offset the
pattern by
> one bit for each address line to make sure you pick up these errors!

DCD error identification is very useful. I find that a lot of the RAMs
that fail (besides an all-out failure) fail on a DCD error. Your
algorithm will have to do something like start with an address range
that is all 00 (or all FF... ) and then write something like $AE to a
given address. Then it'll need to test each address line (XOR 1 into
each bit of the address?) and then read from that address. If it reads
your test pattern in any of the other addresses, then you have a
decoding error. Finally, you'd want to write back to your test address
$00 to clear it and move onto the next test address. DCD testing can
probably be done just by testing addresses:

0000 0000
0000 0001
0000 0010
0000 0100
0000 1000
0001 0000
0010 0000
0100 0000
1000 0000

for a real quick test. A through test would be to test it on each
individual address. The Fluke 9100A manual may give you some ideas of
tests.

Other tests to consider are "burn in" repetitive tests that you could
leave running for a while. Ideally, I'd like to hook up a lot of chips
and then run the test. Possible?

--James Bright
www.QuarterArcade.com
Restored Arcade Games for your Home

If you ever want to (un)subscribe yourself with TTL, you can send mail to:
<Majordomo@flippers.com>
with the following command in the body of your email message:
(un)subscribe techtoolslist
or from another email account, besides xxx@yyy.com:
(un)subscribe techtoolslist xxx@yyy.com
Received on Mon Dec 29 22:16:37 2003