[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: OLTC update - Coupling figured out :-)
Original poster: "Terry Fritz" <twftesla-at-qwest-dot-net>
Hi Ken,
At 02:13 PM 8/25/2002 -0700, you wrote:
>Terry (& all)-
>
>Figure -25 appears to show the secondary voltage reaching max. at the 2nd
>half-cycle and that is, indeed, extremely good. I should think you'd be
>able to see that directly on the scope by just laying your probe out as a
>little antenna, as I do.
I can see it in my plane wave antenna which is just the same thing as a
wire antenna but with a flat frequency response. That is how I first
noticed this. what we see is the normal ringup of the secondary but it
runs out of gas suddenly. It "was" going to ring up much further and
longer but it suddenly had no more power.
>
>If my pet theory, as to voltage rate-of-rise vs. spark-progression time,
>is true, however, that high rate may not do you much good in increasing
>spark length, at just 30-odd KHz. Getting up there in 1 cycle at 140 KHz
>would be a different thing entirely.
The fast ringup thing just shows there is not enough primary power or too
much loss. The actual energy transfer time is sort and not very efficient.
>
>As to why my system takes its 30 cycles to ring up--don't know yet. My
>secondary sits directly on the primary, with spacing of perhaps 1" from
>the top of the primary bundle to the first secondary turn and having the
>same nominal 12" diameter. Certainly, if I could drive it with a sine
>wave with all else being equal, I suppose I would. But the MOSFETs have
>to turn on hard, of course. Also, it strikes me that the harmonic
>content of my voltage square wave will not elicit a lot of primary
>current because the impedance of the primary is a lot higher for those
>components. So...let it be square, as far as I can see.
It may be just fine. I was just reminded of the wild voltage transients
Paul's programs predict in some situations. A square wave case would be
interesting to check.
>
>And as to harmonics "fighting" each other...say what? Does such a thing
>happen?? Not per anything I learned 50 years ago...but then one forgets,
>or never learned in the first place, a lot.
I don't know if that could happen or not. Be interesting to study it.
>
>I'd made a rough measurement many months ago that seemed to show that I
>got the best coupling to the secondary with a (larger diameter) primary
>located about 6" up (on a ~36" coil). Once I get my system up & running
>again (after the current irksome task of changing out 8 MOSFET-driver
>transistors for huskier ones), perhaps I'll make such a coil.
>
>And by the way, what is the mechanism that will cause your losses to
>decrease with increased input current?
The IGBTs tend to have linear losses (unlike FETs that have I^2 loss) with
increased drive current while the coil's power goes up as a square function
of current. So ten times the input voltage gives ten times the loss but
100 times the power. It appears that the coil becomes more efficient
overall with the more power it has available.
Much more to study in all this ;-))
Cheers,
Terry
>
>Ken
>
>On Sun, 25 Aug 2002 12:26:20 -0600 "Tesla list" <tesla-at-pupman-dot-com>
>writes:
>> Original poster: "Terry Fritz" <twftesla-at-qwest-dot-net>
>>
>> Hi Ken,
>>
>> I am using only about a 75 volt firing voltage right now (starting
>> out real
>> slow). But perhaps at these low levels the losses are higher and
>> loosing
>> power to the primary coil which just "looks" like very high
>> coupling. From
>> the scope trace at:
>>
>> http://hot-streamer-dot-com/temp/OLTC08-24-01.gif
>>
>> The secondary losses are fine, but not sure about the primary. I
>> will
>> check into this. I may just have to keep cranking it up and the
>> problem
>> will go away ;-)
>>
>> YEP! Your right! I ran MicroSim with more realistic "low power"
>> losses
>> and got:
>>
>> http://hot-streamer-dot-com/temp/OLTC08-25-01.gif
>>
>> That is indeed what I am seeing. The coupling is fine after all.
>> There is
>> NO problem. I just need to keep cranking the power up and the
>> losses will
>> reduce naturally.
>>
>> Thanks for the insight here!! It would have taken "me" a long time
>> to
>> figure this one out ;-))
>>
>> I am surprised your coil has such a long ring up. Unless the
>> coupling is
>> very low, which a I doubt, Perhaps the square waves don't couple as
>> well.
>> Only the Fo sine component may be doing the coupling while the
>> higher order
>> harmonics of the square wave are either not coupling or "fighting"
>> each
>> other. An interesting and unknown problem, exciting a two coil
>> system with
>> square waves instead of sine waves... Maybe Paul's program could
>> analyze
>> such a case since the harmonics in the secondary may easily come
>> into play
>> in such a case. Simple MicroSim models may not see the true action
>> there.
>>
>> I will try to run some models on this and see if I can figure
>> anything out.
>>
>> BTW - I think I know of a very easy way to measure the primary
>> current.
>> Just a loop of wire under the primary (or near it) to a scope probe.
>> The
>> voltage on the loop should be proportional to the current (or maybe
>> it
>> needs a load resistor?). A simple and very useful instrument whose
>> details
>> will have to wait for another day...
>>
>> Cheers,
>>
>> Terry
>>
>>
>> At 09:31 AM 8/25/2002 -0700, you wrote:
>> >Terry (& all)-
>> >
>> >I've been following the OLTC saga off & on. But it's just occurred
>> to me
>> >that you may find yourself up against the situation I've found with
>> my
>> >s.s. coil: As you may recall, I apply a ~1200 V pp square-wave
>> burst, of
>> >up to ~6 ms duration, to a 3-turn (untuned) primary circuit for
>> each
>> >spark. With a 140 KHz secondary Fr and a 6" x 24" smooth
>> (Landergren)
>> >toroid, it takes ~30 cycles of constant (not exponentially
>> declining!)
>> >excitation to bring the toroid potential up to the spark break-out
>> level.
>> >
>> >You are applying, I believe, ~680 x 2.8 = ~1900 V pp, initially,
>> to a
>> >3-turn primary circuit--incorporating much less resistance,
>> admittedly,
>> >than mine--but your excitation must (necessarily)
>> exponentially-decline
>> >quite rapidly. I'd think you might require more or less those 30
>> cycles
>> >to pump up the voltage & I fear that the decline of your primary
>> voltage
>> >may preclude that.
>> >
>> >Were you able to do any simulations on that?
>> >
>> >Ken Herrick
>> >
>> >On Sat, 24 Aug 2002 23:02:13 -0600 "Tesla list" <tesla-at-pupman-dot-com>
>> >writes:
>> >> Original poster: "Terry Fritz" <twftesla-at-qwest-dot-net>
>> >>
>> >> Hi All,
>> >>
>> >> Today I set up the coil for action:
>> >>
>> >> http://hot-streamer-dot-com/temp/OLTC08-24-02.jpg
>> >>
>> >> I powered up the coil...
>> >
>> >[snipped]
>> >
>> >
>> >
>>
>>
>>
>>
>