On 1/6/2013 5:00 PM, klem kedidelhopper wrote:


Further, after clearing the password a few weeks ago, the computer
seemed unstable. So we copied a few needed files to a memory stick
and my son dumped the hard drive and installed Windows 7 from a disc
he had, that among others had an OS version on it for "Toshiba".

Not clear what that means. I've seen universal install disks for XP.
They address the OEM activation issues, but probably don't have all the
drivers you need for every version of every Toshiba computer ever made.

This afternoon I looked at the power supply on the scope. This supply
is rated 19VDC 3.95A. With no load the output read 19.2V. There was a
small 20mv.low frequency triangle waveform present. Approximately
every 6 usec. there was a 160mv. spike. I took a better look at the
spikes. They actually looked like ringing pulses with a width of about
80 usec. I then loaded the supply with a 4 amp resistive load. The
voltage dropped to 18.9V and the spikes increased in amplitude to
about .900V. every 6usec.

Assuming you have a quality scope and probes and know what you're doing,
0.9V sounds like a lot, but probably not killer.

Now given a SMPS that's charging LI-ion batteries I don't know whether
or not the above constitutes a "noisy" supply or if these kind of
aberrations are to be expected. So I figured I'd substitute my bench
supply with linear regulation as a test.

I set up 19 V with a current limit of 4A. I then connected this to the
computer and after hitting power on I immediately noticed the current
meter surging to 100ma. every second or so. The computer booted up
and worked for about 2 minutes and then just shut down. Afterwards I
was not able to get it to come back on with the bench supply. Thinking
that trying to use a linear supply to power a device with LI-ion
batteries probably wasn't the smartest thing to do I then connected
the suspect Toshiba supply and the computer booted up and never froze
during a 1 hour period of using it.
I'm chasing a ghost.

So now what I do know is that the power supply does spike up to .900V
when its fully loaded. I don't know what the actual load of the
computer is, and since I couldn't simulate this I figured I'd go for
worst case. However I think that I can assume that it must also spike
somewhat when connected to an already powered on computer as well.
Whether or not it's supposed to do this or if this is my problem is
anyone's guess. It's pretty clear that the computer does not like to
run on pure DC though. Lenny

I've seen some wacky designs, but never seen a laptop that wouldn't
run off a DC bench supply.