On Sat, 20 Mar 2010 18:09:54 -0700, John Larkin
wrote:

On Sat, 20 Mar 2010 15:02:24 -0700, Archimedes' Lever
wrote:

On Sat, 20 Mar 2010 13:32:09 -0700, John Larkin
wrote:

On Sat, 20 Mar 2010 09:27:19 -0700, Archimedes' Lever
wrote:

On Sat, 20 Mar 2010 06:05:08 -0500, John Fields
wrote:

On Fri, 19 Mar 2010 18:00:08 -0700, Archimedes' Lever
wrote:

On Fri, 19 Mar 2010 12:38:45 -0500, John Fields
wrote:

Excellent proper use of a ferrite bead on one leg of each device.

---
Sorry, but those are ring-tongue terminals soldered around the emitters
of the transistors.

JF

Yeah, I saw that. It looked so similar to an amplifier I made some
years back, that with just a glance I thought I saw the same thing.
We had transzorbs on each device as well. Oh well... so much for quick
glances.

Does that mean I'm *not* blind as a bat?

John

No, but if you think that "rackmount" "solution" you have where the
transformer weighs a bus load is a good design, you're nuts. It either
MUST be used as the bottom element in the rack. It should actually be
two rackmount modules, considering the fact that it is obvious that lead
length doesn't bother you. We had an amplifier module that had two 2kVA
units in the bottom of it. It was on wheels, and would never have been
considered as a rack mount capable installation, and those transformers
together weighed less than the monster you sport.

It's for MRI, so my customers always buy and rack three at a time.

The leads are of course trimmed to length when the transformer is
wired up.

The checks do clear, so it's a good design.

John

Ours was 1000.00 Volts @ 1500 Watts and fed a CAT scanner transducer
set. It was a bank of about 18 FETs driving the output multiplier
transformer. I hand wound those on a two inch diameter pot core for the
prototypes and the first few production items, until we got the contract
Mfgr taught how to construct them properly, so they would not fail. We
got the noise figure down to well below the customer spec. That was
Philips. It was a real nice power supply. For all I know, they have
since sold hundreds of them.

I think we ended up at 17kHz (35kHz ripple peaks). Ripple was the
important factor, as was precise voltage regulation. Ripple voltage of
1mV, and 10mV regulation, and temperature coefficient of 2ppm/ºC over the
operating temperature range.