Some CompactFlash doesn't work in True-IDE mode

Do you have a question? Post it now! No Registration Necessary

Translate This Thread From English to

Threaded View

Hi folks,

Today I found something interesting you might like to know.  Some
CompactFlash, even with the CF logo, doesn't appear to support
True-IDE mode.  The specific flash that we had was a Memorex branded
flash that, when plugged into an IDE port, comes up in BIOS with a
"32M CHH" name.

The funny part is that it forces itself to be the slave device.  Here
I am with a Geode board with an integrated CompactFlash socket on the
secondary master IDE and nowhere to plug anything else in, and the
flash comes up as the the slave drive.  If you put it into the "normal
IDE" to compactflash adapter board and try to run it as master, the
thing isn't even recognized.

This was fun to find, as it was simultaneous with a new flash creation
tool that uses the Win32 SYSLINUX to create bootable Linux
CompactFlash with a minimum of work from a Windows box.  When it
didn't work, the initial finger was pointed at SYSLINUX.  When I
pulled all the flash and checked them all, I soon realized that the
only flash having problems were those that came up with a "32M CHH"
tag, and they all had a 021 mfg code on the side.

These CompactFlash do say "Copyright 2000" on them so I suspect that
they were manufactured more than a few years ago.  They do, however,
have the CF Association logo on them.  None of the other
Memorex-branded flash I've got has this problem, so I hope that its
just some leftover crap from the early days.

Anyway, cheers.  I hope someone else someday finds this useful to
know.

--
Alex Pavloff - remove BLAH to email
Software Engineer, ESA Technology

Re: Some CompactFlash doesn't work in True-IDE mode
Quoted text here. Click to load it

Though devices sporting the CF logo must support True-IDE mode,they
need only implement the subset of commands in the CF 2.0 spec - are
you sure your driver isn't using others? Also,they may implement some
or all of the commands as no-ops,
Sprow.

Re: Some CompactFlash doesn't work in True-IDE mode

Quoted text here. Click to load it

It seems to me that at the bare minimum, the CF shouldn't force itself
to be the slave device in TrueIDE mode.  I don't have the CF spec in
front of me, but this seems like it would be a required function.
--
Alex Pavloff - remove BLAH to email
Software Engineer, ESA Technology

Re: Some CompactFlash doesn't work in True-IDE mode
IIRC, master/slave mode is determined by one of the lines during either
power up or reset. Along the same lines, CF cards cannot be hot-swapped
if you intend to use them in true-IDE mode. The input line that
determines true-IDE mode must be low at power up. Since the power pins
are longer than this pin, this pin is latched as a high.

Alex Pavloff wrote:
Quoted text here. Click to load it


Re: Some CompactFlash doesn't work in True-IDE mode
Quoted text here. Click to load it

[snip]

Alex

I had a similar issue wiht Memorex CF boards. After connecting -DASP
and -PDIAG to VCC things started to work.

Maybe this helps you too.

Markus

Re: Some CompactFlash doesn't work in True-IDE mode

Quoted text here. Click to load it

I didn't design the board or the socket these things are going into.
They're going into a stock Acrosser AR-B1551 board, a 3.5" Geode
"biscuit" board (of which there are numerous variations by other
companies).

Some more testing results in a bunch of Kingston branded flash of many
many different sizes and actual chip mfgs all working just fine, and
a stack of 5 021-coded "32MB CHH" flash not.
--
Alex Pavloff - remove BLAH to email
Software Engineer, ESA Technology

Site Timeline