Testing CF cards

[Angus Bliss]

Jon recommended I buy the Monoprice CF adaptor as I was not prepared to get the StarTek card with the missing pin 20. I bought from Amazon but what I got is not labelled, not even the box it came in. Does not look like the Monoprice I found on the net, but it works.

I have bought 4 Cisco second hand cards from HK which seem to work.

Anyway I downloaded the Stephen C Cousin programs SCMon_CF_Info_code8000.hex and SCMon_CF_Test_code8000.hex.

The CF cards seem to be from various manufactures.

The test program would seem to be a good stress test as it runs for hours. Took 12 hours to check one 256MB card.

I have the source of the Info program, but cannot find the Test program source. Does anyone know if and where I can get it? As it runs for many hours, I would like to modify it to log only the errors and continue unattended. Non of my cards is error free but some succeed on a second attempt from $04 errors, sometimes not. I call these hard and soft errors.

How does CP/M handle this? Do the CF cards handle bad sectors that the Test program is showing up?

I think the Test is the entire card, but of course CP/M is only using part of card from sector zero I assume, so errors on the higher sectors are OK?

Has anyone got any technical info on CF card errors.

Thanks in advance for any information.

Angus

[Steve Cousins]

Hi Angus

You can find the source code as part of the Small Computer Monitor / Small Computer Workshop download, here:
https://smallcomputercentral.wordpress.com/small-computer-monitor/

LiNC80's CP/M only uses the first 128Mbytes of the CF card so faults beyond that should not be an issue.

Not sure if CP/M understands the concept of bad sectors. I'd guess not but have never investigated this.

Steve

[Angus Bliss]

Thanks Steve,

I knew CP/M was not intelligent enough to figure out disk errors. I was assuming the controller chip in the card might do something much like disk drives do without consulting/informing the operating system unless failure is imminent.

I knew my Morrow Disk Jockey back in 1980 just retried 10 times and then giving up. Looking thru the current CP/M source I can't see retries, but assume they are there.

I have found a package called DUF05 on the net which can scan a disk and all sectors that report errors are rolled into a file in User 15 but I have not tried it in anger yet. I did use it to fix up a directory that one of my experiments had smashed a little. Can post it if anyone wants it. Written in good old TDL Assembler.