[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [TCML] a conceptual SSSR
Ken,
This has indeed been done before. There was a SSTC book (in electronic
form) written by Gary Johnson:
http://www.eece.ksu.edu/~gjohnson/
I believe he did something very much like you propose, which is to base-feed
the bottom of the resonator with relatively high voltage. He also has a
number of excellent papers about tesla coils in general on that site.
The main concerns are:
1) the magnetizing current of the transformer will add more out of phase
current to be switched by your FETs/IGBTs. By the way, what is the
magnetizing inductance of TX1? Surely its not 32uH (that sounds more like a
leakage inductance value). Unless... this is an air cored transformer and
not ferrite as i have assumed? I have not studied the possible
advantages/disadvantages of using air vs ferrite core for such an
application. Using ferrite would have far less magnetizing current, which
sounds good to me, and would also reduce leakage inductance which serves to
lower the driver impedance and avoids too much extra resonant voltage rise
on TX1's secondary.
2) the design of TX1 is not trivial for high power and high voltage
operation. You already propose something that should output 9kVpk (or
significantly more if the leakage inductance is significant), at 140khz even
9kV can deliver considerable power into corona, which will quickly destroy
transformer insulation. So oil submersion/vacuuming might be required for
high reliability, hard to say.
Other than that, i dont doubt that it can work. Performance will probably
be limited by how much current you can supply, and how much voltage TX1 can
sustain. If you need more power, just increase the ratio of TX1.
Steve
On Tue, May 10, 2011 at 4:32 PM, Ken Herrick <kchdlh@xxxxxxxxx> wrote:
> Greetings, all-
>
> I've been away from the List for quite a while (old age, lagging
> motivation, death in the family) but have still been off-&-on pursuing the
> single-resonant scheme (OK, OK!, I know... but it keeps me out of the bars).
> Some may remember the photo I posted years ago, before I "improved" the
> design so much that it quit forever:
>
> http://dl.dropbox.com/u/28799314/KCH_tcg1.jpg
>
> That was a time-exposure, of 5-8 sec as i remember, plus a 2nd (double)
> exposure with flash.
>
> (First time I've used dropbox, so I'm hoping all can view these OK--when
> I've gone beddy-by & am off-line.)
>
> I'm now back at the task, having totally redesigned & rebuilt the primary
> apparatus. But I have thought of a scheme that might ease the considerable
> burden on my power MOSFETs; here's the simulation schematic and I would
> invite comments on it:
>
> http://dl.dropbox.com/u/28799314/KCH%2010%20May%2011%20ckt.jpg
>
> The secondary apparatus is L1 + (R2 = L1's dc resistance) + (C2 = the
> ambient capacitance) + the top electrode. I presently have that + all the
> rest in hardware, built & somewhat operating in my current redesign, except
> for TX1. It's the use of TX1 that I'd like to get comments on. Very likely
> others have done this kind of thing years ago.
>
> The whole thing is, obviously, a feedback loop, to sustain oscillation at
> the "secondary's" Fr. I put that in quotes because, in this concept,
> there's no primary coil as such. Instead, I take the 1800 V p-p that I
> already have (from E1 in the simulation) and feed it thru a 1:10 transformer
> (TX1 + R3), to directly drive the "secondary" (which is to be my existing
> 12"-dia. coil whose in-place Fr is ~140 KHz). TX2 is just a small toroidal
> common-mode filter I've had on hand; it works well in the hdw.
>
> Here are simulation waveforms, each showing toroid output, 1800 V p-p drive
> from E1, and E1's output current:
>
> http://dl.dropbox.com/u/28799314/KCH%2010%20May%2011%20graph%201.jpg
>
> http://dl.dropbox.com/u/28799314/KCH%2010%20May%2011%20graph%202.jpg
>
> Graph 201 shows the beginning of the ring-up and 202, the end (stopping,
> merely for convenience, at 200 us). Notice that well before 180 us,
> switching has shifted so as to occur near current-z.c. Also, the p-p
> E1-output current is markedly below that obtained in my more-complex
> simulation of the current setup, which is less TX1 but including a 4 uH
> primary coil of 4T on 12" diameter. The peak current shown in these graphs
> can quite comfortably be provided by my present driver setup so I'm tempted
> to try adding in a real TX1 before I go much further.
>
> If anyone wants to check this out in SiMetrix, contact me off-List & I'll
> email the file.
>
> Ken Herrick
> Oakland, CA
> _______________________________________________
> Tesla mailing list
> Tesla@xxxxxxxxxx
> http://www.pupman.com/mailman/listinfo/tesla
>
_______________________________________________
Tesla mailing list
Tesla@xxxxxxxxxx
http://www.pupman.com/mailman/listinfo/tesla