RE: SS Variac (Low Pass Filter)

["Tesla list" <tesla@xxxxxxxxxx>]

Original poster: Sparktron01@xxxxxxxxxxx 

Hey Steve

I'm running tonight ... got tired of playing games so bumped power up some...
Mounted bridge and IGBT on a heatsink.  Input transformer is 120X240 Pri.
16x32 sec (wired 120Vin => 32Vout) .  LP filter is a 1.5" OD ferrite with
~60T #16 magnet wire (measured ~30uH); Capacitor is 8uF, 660VAC mylar
out of a HID ballast.  Output transformer is 120VAC primary, 240x480V
sec. (wired 120 =>480V out).  Load is a 300W incandescent bulb.  The
transfomer does have some inductive reactance (as well as LP filter).

Vo is adjustable from <1Vrms to ~90Vrms (~ 150-200W power throughput).
Ran for ~30 minutes to see if something would detonate.  The Bridge + IGBT
heatsink was running about 30 deg F above ambient.  Transformer(s) and
Vo cap was "cool as a cucumber"...  :^)    The LP inductor was taking it on the
chin and was hottest thing in circuit (maybe 50 deg F above ambient, might
be saturating, HF wise).  OBTW operation was very acceptable with NO
RC snubber across bridge (which I will install) or line filters + varistors on
line side and primary side of HV transformer being powered by this
contraption.  Output waveform had some HF in it, but considering (1) we
are processing real power, and (2) LPF has not been otpimized for PWM
carrier or power throughput, either I'm REAL lucky, or circuit as some real
promise (at least below 1.0kVA).  In real circuit, I will use a TL494 for
independent voltage and current feedback adjustment and soft start
capability.  Neat to hear the transformer (and 300W bulb) "sing" at about
5kHz...   :^D

I can provide pics of lashup, waveforms, etc.

Best Regards

Dave Sharpe, TCBOR/HEAS
Chesterfield, VA. USA


> Original poster: "Steve Conner" <steve.conner@xxxxxxxxxxx>
>
>  >The basic topology is a full wave DC bridge in
>  >series with hot leg, with a 600V, 40A IGBT across +/- of bridge as a
>  >commutating
>  >switch.
>
> Before you do any more work on this, please give some thought to the $64,000
> question- When you're driving an inductive load (like a NST etc) through
> this device, the current will try to keep flowing when the IGBT turns off.
> Where will it flow?
>
> I tried to figure it out but couldn't think of any path for it. I concluded
> that the diode bridge or IGBT would be destroyed by inductive kickback when
> running with an inductive load, so I gave up on this topology.
>
> If anyone can prove me wrong I'd be happy as it's a very simple and
> attractive circuit. I know a guy who successfully used it to do PWM dimming
> of filament lamps but of course they are pure resistive loads.
>
> Steve Conner
>
>