The Cyber-Spy.Com Usenet Archive Feeds Directly
From The Open And Publicly Available Newsgroup
This Group And Thousands Of Others Are Available
On Most IS NNTP News Servers On Port 119.
Cyber-Spy.Com Is NOT Responsible For Any Topic,
Opinions Or Content Posted To This Or Any Other
Newsgroup. This Web Archive Of The Newsgroup And
Posts Are For Informational Purposes Only.
X-Mailer: Mozilla 4.79 [en] (Win98; U)
Subject: Re: Problems with MOSFET drivers
References: <firstname.lastname@example.org> <email@example.com> <firstname.lastname@example.org> <email@example.com>
NNTP-Posting-Date: Mon, 18 Nov 2002 17:15:05 GMT
Organization: AT&T Broadband
Date: Mon, 18 Nov 2002 17:15:05 GMT
Winfield Hill, Mike Poulton, Adam Seychell wrote:
>>> Then explain how your going to find the resonant frequency
>>> with knowing only the averaged magnituded of primary current.
>>> Especially for something as fluctuating as a tesla coil.
>> Step 0: start oscillator at frequency known to be low.
>> Step 1: read and store primary current
>> 3tep 2: increase frequency slightly
>> Step 3: read new primary current
>> Step 4: compare new current to old current
>> Step 5: if new>>old, change frequency more in same direction, by
>> an amount determined by the change in current (big change
>> in I gives small change in F).
>> Step 6: If new<> Step 7: If old<>new, but not by much, then leave where it is.
>> Step 8: goto step 3
This is a very crude control scheme typical of a bit head who is only
comfortable thinking in ones and zeros. The gain is non proportional,
slew rate limited, and is zero within the dead zone. It relies on the
physical inertia of disturbing objects moving about within the tesla
coil's field to provide the dominant pole to stabilize the system.
This may more or less work under most conditions - until a pesky arc
is drawn that breaks into fluctuations at the cycle time of the micro's
simplistic control loop. Then the control scheme might live up to its
common name of "bang-bang" control in a most exciting way.
However, with the right precautions, such as digital or analog low-
pass filtering of the current signal and some proportional feedback,
it might just get the job done. Whatever works, eh?
> Arghhh! Please let me suggest you measure the phase of the current
> out of your driver with respect to the voltage. You'll find that
> at resonance it's in phase, below resonance where the load appears
> inductive it lags, and above where the load looks capacitive it
> leads. Readymade for generating a servo signal, or at least a meter
> reading to tell you which way to tune your oscillator frequency.
Doesn't the current signal actually have both a resonance and an anti
resonance due to the uncoupled primary inductance feeding into the
stacked resonant tank? Has anyone ever run this experiment on the
test bench? If this is indeed true then current phase will only be
monotonic in the very immediate vicinity of the resonance and will
flop right over through the lower frequency anti resonance. Such
feedback signal behavior would be nearly as problematic to a control
circuit as that presented by attempting to use current magnitude.
In my opinion, Mike would do much better to use the integrated signal
from his sense coil that he wrote about in a prior post. That signal
should be completely monotonic within the control band and beyond and
be perfectly suited for automatically tracking tesla coil resonance.
How about it, Mike, have you yet measured the phase relationship of
sense coil signal to the tesla coil primary voltage? -- analog
Go Back To The Cyber-Spy.Com
Usenet Web Archive Index Of
The sci.electronics.design Newsgroup