[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Re: [TCML] Solid state efficiency, was: mini Tesla coil specs]



The magnetizing current is one reason i ditched the un-tuned primary
approach.  When using IGBTs, it seemed to be particularly stressful, though
i suppose i was pushing some rather huge power levels with only a pair of
SOT-227 packaged IGBTs.  Anyway, the magnetizing current meant i was
hard-switching something like 20A out of 60A peak primary current at
120khz.  The IGBTs just dont put up with that very well and soon let you
know about it.

It seems that the topic of this thread is no longer applicable to the
subject at hand.  I'll leave that up to you, Ken.

Steve

On Mon, Nov 16, 2009 at 9:38 PM, Ken or Doris Herrick <kchdlh@xxxxxxxxx>wrote:

> Steve (& all)-
>
> I've been trying to stick with an untuned primary + secondary-current
> feedback so that I could easily switch secondaries without any adjustment.
>  I have two, one with Fr of 100 KHz and the other, about 120.  The idea is
> to just be able to plop one into place, then the other with no fussing.
>
> My (simulated) switching occurs uniformly at ~0.44 us after
> secondary-current z.c. for a 4.3 us half-period  I believe I was remiss in
> reporting a phase-change:  The phase appeared to change because the shape of
> the primary current was changing due to the magnetizing current.  I've now
> re-checked the phase between the secondary feedback signal's z.c. and the
> primary's z.c.'s for both the tuned & untuned condition. Tuned, they start
> out 90 deg. out but within 45 us they are in-phase & remain that way.
>  Untuned, the same 90-deg. out initially with a triangle-wave of current.
>  Then the current shape gradually approaches sine but the z.c.'s never
> coincide.  Always 45 out or so.
>
> Perhaps I might go with tuned.   Gotta study it.  Stay tuned.
>
> KCH
>
>
> Steve Ward wrote:
>
>> Ken,
>>
>> This lack of perfect zero current switching is exactly why i use primary
>> current feedback to determine when to switch the bridge of transistors.
>>
>> But... im confused.  Your system is an un-tuned primary, just an inductor
>> which can be modeled as being part of the secondary coil.  How, then, can
>> your primary current be out of phase with the secondary current?  While
>> there is some "magnetizing" current in the primary (being 90* out of phase
>> essentially), it should NOT be so much to cause that much hard switching
>> (though it is a pain in CW systems).
>>
>> I suspect there is something not correct about your feedback loop, and you
>> are getting a 90* phase shift.  Have you verified that your transistors
>> are
>> switching at the secondary current zero crossing (or within .5uS of it)?
>>
>> Switching at the primary current *peaks* doesnt seem elegant or like the
>> correct solution.
>>
>> Steve
>>
>>
>> On Sun, Nov 15, 2009 at 3:43 PM, Ken or Doris Herrick <kchdlh@xxxxxxxxx
>> >wrote:
>>
>>
>>
>>> So almost immediately...a follow-up:  Belatedly I thought to resonate the
>>> primary & try it again.  This time, a) the current's a sine-wave all the
>>> way; but b) switching /starts out /near current/ /z.c/./ and after about
>>> 4
>>> cycles it's shifted to current-peaks, where it stays.  The current is, of
>>> course, a lot greater due to the series-resonance, causing the
>>> secondary's
>>> output to rise faster.
>>>
>>> But a) where's the advantage, if switching drifts toward current-peak.
>>>  And
>>> b) how to handle the greater current?  (Bigger hardware, that's how!)
>>>
>>> And finally:  Whichever way one goes, it seems as if one will have to put
>>> up with switching near current-peaks--unless one can produce the spark
>>> within the 1st 4 cycles or so in a DR design.  I haven't simulated that
>>> yet.
>>>
>>> Comments?
>>>
>>> KCH
>>>
>>> -------- Original Message --------
>>> Subject:        Re: [TCML] Solid state efficiency, was: mini Tesla coil
>>> specs
>>> Date:   Sun, 15 Nov 2009 13:19:09 -0800
>>> From:   Ken or Doris Herrick <kchdlh@xxxxxxxxx>
>>> To:     Tesla Coil Mailing List <tesla@xxxxxxxxxx>
>>> References: <cc2218e80911111413md7c0b57k355ad83baa7e78a4@xxxxxxxxxxxxxx>
>>> <
>>> 4AFC3E64.90209@xxxxxxxxx> <
>>> cc2218e80911130718k750de578t197b1d7d7677aaa9@xxxxxxxxxxxxxx> <
>>> 4AFDCA69.4090001@xxxxxxxxx>
>>>
>>>
>>>
>>>  From KCH-
>>>    While I have the attention of a few s.s.'ers here, perhaps someone can
>>> shed
>>> light on this:  I'm simulating what amounts to a half-bridge driving an
>>> untuned-primary t.c.  I record waveforms of a) MOSFET current and b)
>>> primary
>>> current.  I notice the following: At the start, the MOSFETs switch at
>>> current peaks and the current waveform is triangular.  As the primary
>>> current--and secondary voltage, of course--builds up, the current
>>> waveform
>>> gradually changes toward a sine shape and the switching events shift
>>> toward
>>> current-zero-crossing.  I have a feeling that that's the case in the real
>>> world.  Is that to be expected?
>>>
>>> KCH
>>>
>>>
>>> _______________________________________________
>>> Tesla mailing list
>>> Tesla@xxxxxxxxxx
>>> http://www.pupman.com/mailman/listinfo/tesla
>>>
>>>
>>>
>>
>>
>>
>
> _______________________________________________
> Tesla mailing list
> Tesla@xxxxxxxxxx
> http://www.pupman.com/mailman/listinfo/tesla
>
_______________________________________________
Tesla mailing list
Tesla@xxxxxxxxxx
http://www.pupman.com/mailman/listinfo/tesla