[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Re: Re: Re: [TCML] SSTC full bridge control system question
Michael Twieg wrote:
Great, the program helped my figure out what I was doing wrong (I had
written a short mathematica script to do about the same thing). The
Drsstcd program outputs gave me great looking results; output peaked
just over 1MV with the first current peak being just over 1KA.
Strangely, the values used weren't incredibly far off (dropped K by
about 15%, dropped La by about 50%), so apparently the system is more
sensitive than I had thought. There must be some sort of bifurcation
in the system's behavior that allows it to only notch with zero
crossing under specific circumstances. Hopefully our rig will give us
enough tuning capability to hit the mark.
If the coupling coefficient is small the tuning is more critical.
Also, another thing I noticed during the simulation was that though I
deactivated the driving source at the first current notch, the second
current peak was greater than the first. Must be due to energy in the
secondary coupling back to the primary. Do you think it's safe to
assume that additional peak currents aren't a large concern, since at
that point the IGBTs will be conducting through their clamping diodes
and not their collector-emitter junctions?
This happens if you short-circuit the driver. If the driver transistors
have diodes across them, the
diodes return energy to the driver with less current than in the first beat.
I wrote this some time ago, and one of the cases studied experimentally
shows this phenomenon:
http://www.coe.ufrj.br/~acmq/tesla/drsstc1.pdf
With a driver with diodes, what happens is something as this:
Driver voltage and input current:
http://www.coe.ufrj.br/~acmq/tesla/finalcurrent.jpg
Driver voltage and output voltage:
http://www.coe.ufrj.br/~acmq/tesla/finalvoltage.jpg
The images were taken during tests with a low power drsstc (similar to
the default design of
the drsstcd program, with 2 cycles added). A half-bridge driver with
bipolar transistors powered by
180 V and free-wheeling diodes was operated for 10 cycles and left open,
with the diodes returning
energy. The maximum input current was of about 20 A, and the maximum
output voltage of about
60 kV. This system worked, but had quite low output voltage and
difficulties in producing more
perfect driver waveforms.
A problem that was observed was that after the interruption of the
driver the square wave produced
by the returning energy was causing the transistors to conduct by
excessive dv/dt, with large power
dissipation. Circuis had to be added to keep the transistors off.
This system didn't use feedback.
Antonio Carlos M. de Queiroz
_______________________________________________
Tesla mailing list
Tesla@xxxxxxxxxx
http://www.pupman.com/mailman/listinfo/tesla