I have an old Toshiba 2085-S603 laptop.
It runs fine on the internal 128MB ram,
but won't recognize the additional RAM in the socket.
I've tried several known-good 64 and 128MB modules
in PC100 and PC133. Nothing. It came with
a 128M Micron. Sticker says it has 256M. The 128 works in
other laptops. I bought it dead, so don't know
if it ever recognized the 128M.

I'm not the only one with the problem, but couldn't
find anything by google more recent than 2005.
"Buy Kingston RAM at $100 a pop," is not an option
on a $3 laptop.

I'd like to investigate the possibility that I have
a hardware problem. But the physical construction
makes it impossible to probe.

Would be nice if I had a memory tester that said
something like, "bit 2 is stuck". That would dramatically
reduce the amount of wire soldering I'd have to do to
bring out a probeable lead.

Memtest86 will let me enter a top memory address higher
than 128MB, but doesn't seem to change the test range.

Is there a memory test utility that will let me
read/write a physical location above what's recognized by the
bios?

Thanks, mike
--
Return address is VALID!

mike wrote:

Hi Mike,

If the BIOS does not recognize the RAM chips, then it can not be tested
by any software.

With laptops, RAM can be VERY specific. I had a couple of 1GB SoDIMMs I
wanted to move from a zd7010us to a zd8215us and it wouldn't work. I
knew the SoDIMMs worked since I pulled them from a workig laptop.

When I went to the Crucial & Kingston pages both, they did show 2
separate SoDIMM part numbers.

There use to be a cross-reference chart for RAM modules on the web. I
don't remember the URL, but you could Google it and see if the RAM you
are trying to use is in fact compatible with your laptop.

--

Sincerely,
C.Joseph Drayton, Ph.D. AS&T

CSD Computer Services
Web site: http://csdcs.tlerma.com/
E-mail: csdcs@tlerma.com

C.Joseph S. Drayton wrote:
OK, let's do a sanity check.
The BIOS determines how much ram by testing for it.
It plugs a number into a RAM location telling the OS how
much ram exists.
If the bios can plug that location, so can my software.
My code should be able to do anything that
the processor is capable of without any reference to or use
of the bios...once the code starts excuting.
I'm just too lazy to research it and write the code if it
already exists.

I did a bunch of googling. RAM vendors are not very helpful describing
what they have. But the most important number...the one after the $...
is a deal breaker for all of them.

Would be interesting to learn the actual differences in the modules.
Might just be a change who's only purpose is to prevent you from
buying ram elsewhere. The RAM modules in the HP4P printer are standard
simms with the jumpers changed so you have to pay HP $$$$ for them.
Fortunately, somebody figgered that out and published it. Thought
maybe someone had figgered out the magic for Toshiba ram.

I expect it's an exercise in futility. Probably in the contents of the
write-protected half of the serial eeprom.

But I had to at least ask.
mike

--
Return address is VALID!

mike wrote:

Hi Mike,

It is the BIOS firmware that is checking the bus. If the BIOS does not
see anything on the bus, how can any software (that runs under a
commercial OS) do so.

There are diagnostic disk that technicians can buy that do not have to
get information from the BIOS and can talk to the bus directly. All the
commecial OSes that I am familiar with must at some point receive
information from the BIOS as to what is on the bus.

--

Sincerely,
C.Joseph Drayton, Ph.D. AS&T

CSD Computer Services
Web site: http://csdcs.tlerma.com/
E-mail: csdcs@tlerma.com

mike wrote:
It is possible that your notebook was produced at the transition to
lead-free solder. Lead-free solder used during the transition
(2001-2002) has less resistance to thermal cycling and has a tendency to
crack. RAM sockets are prone to this cracking since they are typically
on the backside of the mainboard near the CPU or GPU, and are not
cooled.

One cracked joint in the heretofore unused RAM socket will make the
inserted RAM appear to fail. Unfortunately, there is no economical
solution to the problem unless you can find another $3 mainboard. Note
that there have been advances in the application and composition of
lead-free solder that now prevent degradation of solder joints.

Q