DC H-Bridge Results

["Tesla list" <tesla-at-pupman-dot-com>]

Original poster: "S & J Young by way of Terry Fritz <twftesla-at-qwest-dot-net>" <youngs-at-konnections-dot-net>

List,

I tried an H-Bridge TC configuration, patterned after Larry Robertson's work
of a couple years ago.  I used the same 4 x 23 inch twin TC setup powered by
the same twin MOT filtered doubler which puts out 8 to 10 KV DC, depending
on the load.

It works fine, giving roughly the same performance as the other DC setups I
have tried - about 56 inch streamers at max power.  The H-Bridge
configuration has the advantage of roughly doubling the power supply voltage
to the tank cap, similar to a charging reactor.  The other advantages are no
ballasting required, and BPS ranging from as slow as you can make your RSG
go to as fast as you dare run it.  It gives you a lot of flexibility to try
various combinations of BPS and power supply voltage.

Basically, and H-bridge functions like a dpdt switch.  In position 1, it
supplies the power supply voltage to the series connected tank cap and
primary.  In position 2, it reverses the polarity of the DC to the primary
circuit.

I reworked my RSG to function as a dpdt switch.  It has 8 rotor electrodes,
with adjacent pairs connected together.  It has five stationary electrodes,
spaced the same as 5 adjacent rotor electrodes.  Electrodes 1 & 5 connect
together and go to the power supply.  #3 also goes to the power supply.  2 &
4 go to the series connected primary and cap.  Draw a picture to see how it
works.   It produces 8 bangs per revolution (4 voltage reversals).  The # 3
electrode gets much hotter than the others as it is doing double duty.  It
would be better to use 6 stationary electrodes, but I didn't have room for a
6th one.

The disadvantages are more losses in the spark gap, and a more complex
mechanical RSG.  Each bang fires four gaps.  Also, it requires the power
supply to have a couple of microfarad of low esr output filter cap.
Finally, I don't believe one could replace the 4 gaps with TSGs
successfully.  I have tried that with a spdt setup and ended up with
tremendous BANGS as the DC supply filter cap would discharge through both
TSGs.

The other thing I notice is that it doesn't seem to quite double the DC
voltage across the tank cap.  For example, when the DC supply is providing
1472 watts (8 KV at 184 ma) at 500 BPS, the 26 nF tank cap power calculates
to 1664 watts (0.5 * .026 mF * 16 KV * 16 KV * 500 BPS)  - obviously not
correct.  Using 15 KV in the calculation works, so the multiplier is about
15/8 = 1.88 instead of 2.

As an aside, MOT DC supplies get more efficient as the load increases.  For
example, at 200 BPS, the DC watts out at 9 KV was 630 and the AC KVA was
966.  So efficiency is 65%.  Why so low?  Because the MOTs at full AC input
draw a lot of current with no load due to the skimpy core design.  At 1472
watts out the KVA is 1816 which is 81% efficiency.  I believe as load goes
up, the core becomes less saturated and a higher percentage of primary power
makes it into the secondary.  No doubt PIG or PT powered supplies are more
efficient.

Bottom line - the H-Bridge does have some advantages, but overall, I prefer
a charging reactor configuration.  This only requires a spst RSG, or a TSG
with a hefty air blast to quench it.  And the majority of the parts can come
from salvaged microwave ovens.  My charging reactor is 3 MOTs in series, and
I have had zero voltage breakdown problems with them, and they only get
slightly warm with long runs.

--Steve Young