Notes of Meeting with Howie Pfeffer on Active Shunt Test - September 4, 2002
============================================================================

	Peter Garbincius and Howie Pfeffer

	Howie showed me the hardware and the test setup in the booster gallery
east side lab. A circuit sketch is included in my logbook PHG22-142. The FMI 
dipoles are simulated by two transformers giving 7 mHenry and 50 milliohms.  
The FMI dipoles will have an ohmic loss of about 11.5 volts at full Tevatron 
current. In parallel with the FMI dipole (or simulated) load is a 50 mOhm
resistor (20 * 250 mWatt 1 ohm resistors in parallel) all in parallel with
a bank of FET switches (7 for now, may add a few later) and associated load
resistor and bias supply.  

	The idea is to have approximately 230 Amps bypassed around the FMI 
dipoles at full Tevatron current.  Of this 230 Amps, approximately 155 Amps
would pass through the resistor bank and 75 Amps through the FET bank.  A 
transductor, comparator, and gain amplifier produce the FET gating.  This
system has been operated with 37 amps through the simulated load PLUS 37 amps
through the resistor bank, adding to 75 amps through the FET's.

	Although these are switching FETs (running at approximately 1/2 of
their rms power rating) they do operate stably (no current oscillations) 
without excessive temperature excursions.

	Howie intends to add a few more FETs to bring capability up to
100 Amps through the FETs.  Water cooling will be added to the FETs and
resistor bank (< 2 KW through resistors).

	Howie discussed the controls and specifications (accuracy and limit
of drift) a week or so ago with Bruce Hanna and Jerry Annala.  They desired
1 amp at 4300 Amps, which corresponds to 1 Amp at 300 Amps (0.33%) through 
the shunt, an easily attainable specification.  Howie will use a 468 
reference card with a standard low-Beta quad regulator using their own 
D-to-A and OP amps, and which will NOT require temperature stabilization.

	There are a few things still required, such as optical fiber coupling
of the reference to the FETs and an isolation power tranformer for the bias
power suppoly (10 v @ 150 A bias supply, either linear or switching?)  

	Redundancy and Spares is an issue, whether to have components or
entire systems.  Howie is leaning to having spare sub-systems such as: FET 
bank, controls, bias power supply, tranducer, optical fiber (isolation link),
etc. 

	The system will be packaged in (now empty) relay racks in B4 Service
Buidling.

	Howie discussed quench detection and stability.  Since each QPM cell
covers each of the 24 houses, the B42-B49 QPM cell at B4 will have the 
FMI/shunt system hanging off its end, so it will minimize any spurious quench
signal.  Of course, the shunt system and cables have to be protected from
the full Tevatron current.  HOW?

	Continued testing will check temperatures and stability at up to
300 Amp total shunt/100 Amp FETs.  There will also be an attempt to more
realistically test simulate the Tevatron and FMI dipoles in terms of L, C, 
and R.  It is noted that the operation of this shunt system has been simulated
in SPICE.

	Finally I asked Howie if it will be ready by October 1 => Yes,
what then about backup => will, will yes, will start ordering materials for
backup.  I ended by reminding Howie that it is imperative to make sure this
works fully and reliably before committing to removing the existing abort
Lambertson magnets.

Follow-up Tuesday, September 10:
===============================

week of 9/9 - Howie Pfeffer's father is ill, he will be gone

at suggestion of Dan Wolff, spoke to Julius Lentz:

	last week added extra FET and rearranged resistors 
		for better heat dissapation
	is still waiting on delivery of water cooled heat sinks
		can work with air, but water cooling is better
	"not yet on paper" => requisition and design?
		isolation transformer
		opto-isolation for reference (update obsolete components)
		current transducer (replace borrow)
		spares and backups

Jim Ranson said that Howie has not put in requisition for pulling cables
	from B49 to B4 Service Bldg - about 500 feet each, 4 cables, 2 per pole

I relayed all of this to Dan Wolff and stressed testing, backups, spares 
	since this element will keep the whole Tevatron program off the air
	Dan agreed that this system could not be ready for
	1 October, would not be ready for 15 October, 
	likely ready for 1 November, would be ready for Christmas

see also notes of 10 Sept meeting


Completed Action Items (4sept):
===============================

	SHUNT:  better dates for readiness, reliability, full backup
	11/4:  Howie says installation at B4 by 20 December

	how to protect shunt/cables from 4500 Amps and quench
	11/4:  Howie says dont worry about it