Homopolar Generators (N-Machine) by Bruce de Palma

[mikestocks2006]

It doesn't mention anything about OU or not, but the amount of current drawn sure is impressive:

1.6 million Amps

Nature 204, 1079 (12 December 1964); doi:10.1038/2041079a0
Tests with Solid Brushes on the Canberra Homopolar Generator
R. A. MARSHALL*

Research School of Physical Sciences, Institute of Advanced Studies, Australian National University, Canberra.
*Present address: c/o Large Generator and Motor Department, General Electric Co., Schenectady, New York.

EARLY in 1963 a programme was begun with the view of establishing whether or not solid brushes could be used instead of liquid metal jets to collect the required current (1.6 million amp) from the rotors of the Canberra homopolar generator. Within the first few weeks of investigation it became clear that there were no insuperable problems. Many pilot tests were conducted and details of the design will be published later. Since then, brush gear has been designed, manufactured and installed, and large-scale tests have been made with this gear in the homopolar generator. The behaviour of these brushes has been very satisfactory.

Blamey, J. W. , Carden, P. O. , Hibbard, L. U. , Inall, E. K. , Marshall, R. A. , and Oliphant, M. , Nature, 195, 113 (1962). | ISI |
http://www.nature.com/nature/journal/v204/n4963/abs/2041079a0.html

[PulsedPower]

It doesn't mention anything about OU or not, but the amount of current drawn sure is impressive:

That is their main use, high current DC, the one in Canberra had a short and bent all the buss system due to the magnetic forces between the conductors, estimated fault current 6MA,  Some newer machines make much more current but I do not know of any with a higher energy capacity, the Canberra machine is huge, it has a 2000Kw Dc generator to drive the magnet and bring it up to speed..

[AbbaRue]

It would seem the best approach to building one of these would be to build a bunch of smaller units and connect them in series.
It would be very difficult finding semiconductors that could switch such high currents,
which would be needed to step up the voltage to a more useful level.
Also would be easier on the brushes to have a smaller current.

[sm0ky2]

Brushless contacts, and the Return-curcuit at a Tangent, to minimize Back-EMF resultant from the drawn current.

[sm0ky2]

multiple disks on a common shaft, to maximize output.