[conspire] Another motherboard was _not_ burned out today

[Rick Moen]

Redirecting back.  (I'm inferring an intent to post rather than send
private mail.)

----- Forwarded message from Ross Bernheim <rossbernheim at speakeasy.net> -----

Date: Mon, 16 Feb 2015 17:06:21 -0800
From: Ross Bernheim <rossbernheim at speakeasy.net>
To: Rick Moen <rick at linuxmafia.com>
Subject: Re: [conspire] Another motherboard was _not_ burned out today
X-Mailer: Apple Mail (2.2070.6)

Ridk,

I did Tempest and EMI testing for years at a location in Palo ALto and the supposedly
good power was fine for most things, but we had to test the power line on equipment 
to assure among other things the didn’t put too much noise onto the power line. We 
had to measure the noise on the power line first. 

The bottom line was that the noise on the power line was well above the levels
that we had to test to. Even a 100dB filter on all three lines and a 200ft ground 
were not enough to eliminate the noise down to a usable level!

Several years later I was repairing RF generators for another company. They
worked fine in the morning. When the inspector came by in the early afternoon
they failed. Traced it to low voltage on the AC line. Seems we were at the end
of the power distribution line. Called PG&E and they sent a senior and very 
knowledgeable tech. He adjusted the transformer on their line to give us the
highest voltage he could. It didn’t help enough. We had to buy a very heavy 
duty adjustable transformer and put it in the line and with the inspector 
watching each time, I adjusted the voltage to the specified minimum 
voltage for the generators. They always passed.

As to MOV’s. A MOV is fairly fast. It won’t stop extremely short duration 
spikes. However a single MOV doesn’t work. You need at least two MOV’s
on each line a lower level MOV and a higher voltage MOV. Also these 
devices have a limitation on power absorbed in an event before they burn
out. So they must be replaced after a major event.

Another class of device to consider is a saturated transformer. At normal 
voltages the magnetic circuit is saturated so any higher voltage or spike 
cannot transfer through the already saturated magnetic circuit.

Ross


> On Feb 16, 2015, at 1:59 PM, Rick Moen <rick at linuxmafia.com> wrote:
 
[snip a quoted copy of my entire prior post]

----- End forwarded message -----