Re: Basic Stamp Controlled Spark Gap

["Tesla list" <tesla-at-pupman-dot-com>]

Original poster: "Jeremy Scott by way of Terry Fritz <teslalist-at-qwest-dot-net>" <supertux1-at-yahoo-dot-com>

 > This will probably need to be modulated so
 > interference is not an issue.
 > 1/4" may even be a bit large of a hole if you want
 > real fine adjustment of
 > your timing or phase.

I want the hole to look like an electrode.
If the hole is large, there will most likely
be a ramp up and ramp down of intensity --
like during an eclipse :) I should be
able measure this period accurately, and
it should be fairly close to what happens
during the 'real' electrode presentation.
I realize the real presentation 'rings down'
and isn't the same wave shape, but their
time periods should be the same.

 >  > hooked up to an input pin. This will create a
 > stream
 >  > of pulses that can be measured. RPM and electrode
 > face
 >  > time will be calculated in the stamp. The hole
 > will
 >  > be some fixed number of degrees away from an
 >  > electrode, the basic stamp will use this (along
 > with
 >  > RPM) to calculate exactly 'when' a rotating
 > electrode
 >  > passes
 >  > a fixed point.
 >  >
 >  > Speed control -  H-Bridge push/pull type circuit
 >  > using power MOSFETS and an output from the stamp.
 >  > This and the tachometer feedback will keep the
 >
 > I made a leg of a H bridge and scaling it times 4
 > and making a real PCB for
 > it was more of a hassle than just getting a ready
 > made bridge. It was not
 > shielded and has only "crashed" once so far. I'm
 > sure if I put it in a box
 > and kept it more than a few inches from my spark
 > gap, it would be fine
 > forever. Speed control has so far just been a linear
 > taper variable
 > resistor.

 > For a PWM drive, you need want a fixed chopping
 > rate, and a variable on/off
 > timing for each clock cycle. Mine was 2.4kHz because
 > it sounded cool when
 > connected to a speaker. The real controller I ended
 > up using is also 2.4KHz,
 > but some are up in the tens of kHz.

Did you have an AC or DC motor?

 > This will be intersting because the signals you want
 > to measure are really
 > noisy and complex, unlike say the voltage on a
 > battery as it slowly drains.

Yes I figured that :(  At miniumum i should be
able to detect zero voltage at 60Hz (reversal)

 > You may be able to turn this from a hardware into a
 > software problem, and
 > change from measuring capacitor voltage which is
 > probably going to be
 > complex and require mega-isolation, to measuring
 > current flowing into the
 > capacitor from your power transformer with a current
 > transformer on one HV
 > lead of your transformer. You can in software assume
 > that:

 > - charging the capacitor to full voltage takes
 > current. As long as current
 > flows from the transformer to the capacitor, it's
 > charging, or discharging
 > providing reactive current back into the
 > transformer, or you have some sort
 > of fault. We don't want this, we want to use this
 > energy in the tesla coil,
 > not send it back into the mains.
 >
 > - if current flow is zero, you have no power, the AC
 > input is crossing zero
 > and he capacitor is fully discharged.

 > - if the direction of the current from transformer
 > to capacitor is changing,
 > the peak AC voltage is dropping ( as the sine wave
 > ramps to zero before
 > reversing) and it's a good time to fire that
 > capacitor into your tank
 > circuit as it's fully charged and won't charge
 > anymore.

I used to assume that the capacitor's peak voltage
occurs at the same time as that of the power supply.
That's true when unlimited current is available, ie,
what MicroSim will tell you if you don't find away
to limit the power supply. But I do agree that one
should dump the capacitor before the half-wave ends.
It's also easier on the power supply since the gap
effectively shorts it when there is no voltage :)
(Minimal 'hum' in your streamers?)

 > I'm sure more logic is involved here, but software
 > projects are never as
 > easy as they first look. Again, you're probably
 > going to have to filter out
 > the RF noise and fun stuff like that as you'd need
 > to with measuring the
 > voltage on the cap itself, but the possible
 > advantage here is your are
 > measuring small currents with no complex isolation
 > requirements, and direct
 > voltage readings may be useless. The lead from even
 > a neon sign transformer
 > through some sort of magnetic core + a few turns of
 > wire = instant current
 > transformer.

Well, the stamp I've got has 40 (forty) input/output
pins.... so I'm going to try and slap a sensor on
everything I can. :)