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

Re: Basic Stamp Controlled Spark Gap



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

 >
 > If the motor has a shaft that sticks out the other
 > end, put your optical
 > encoder there.. keeping it away from the high
 > voltage/high current
 > circuitry of the primary.  You might also be able to
 > pick up pulses from
 > the internal cooling fan of the motor with a
 > suitable sensor.  Most motors
 > have some sort of fins cast into one or the other
 > end of the rotor.

Unfortunately the motor I have in mind is large and
only has one shaft :(

The optical encoder needs to look like a 'virtual
electrode' That is, the hole it senses has to be the
same diameter and rotate at the same speed (inches
per second around the circumference as well as RPM) as
the real electrodes. So say we have a disc with four
electrodes at 90,180,270,360 the virtual '5th
electrode' is going to be a hole of the same diameter
(1/4") somewhere between 90 and 180...(say 135)
sensors can be mounted on top of the motor
well away from the electrodes on the sides. If I use a
super bright IR led (or a bunch of them) and a black
tube/lens around the sensor, it should be easy to keep
things well spaced. The program will know that pulses
from this sensor are 45 degrees out of phase with
the real electrodes. The light through the hole
fairy accurately mimics the conductivity/time ratio
of the spark gap. (arc channel forms slightly before
electrodes are perfectly aligned and exists slightly
afterwards, so does the light beam -- except it's
intensity/time -- easily measured with a A/D
converter)

 > >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
 > >motor running at a consistent speed by varying
 > >the frequency of AC. This should compensate for
 > >any desyncronizing drag of the disc. How fast
 > >it rotates is yet to be determined. Initially
 > >planning on 1800RPM with eight electrodes for
 > >up to 240BPS.
 >
 > Since the motor is unidirectional, all you need is
 > one switching device..
 > you don't need 4 quadrant operation (i.e. reverse or
 > braking)...

You hook a synchronous AC motor up to a special H
bridge and you can feed it a pseudo-sinewave of
whatever frequency you want (up to a point)
It's RPM is a function of the frequency you
give it. The down side is that the motor has the most
torque at 60Hz (it's designed for that...)
and you have have a 120V DC supply :) (not hard..)

It may also be possible to vary the phase this way
by stretching just one of the sinewave 'pulses' (H
bridge makes a sinewave out of a series of hi-low
pulses)

I've since changed my design thoughts on this one :)
I found a 'digital potentiometer', which makes
controlling a triac and all that jazz (light
dimmer/speed control/phase shift) available.

  > >Phase control - R/C type servos will rotate the
 > >'fixed'
 > >pair of electrodes (connections to the power
 > supply)
 > >anywhere from 0 to 90 degrees in fractional degree
 > >increments.

 > Why not shift the rotor motor speed and phase...
 > rather than moving the
 > electrodes. no extra servos needed. Imagine you've
 > got the motor synced up
 > at 1800 rpm and at phase=0... if the motor is slowed
 > to 1799 rpm, it will
 > roll through 360 of phase shift relative to power
 > line in 1 minute.. 6
 > degrees/second.. so, if you wanted to run at 18
 > degrees behind, you just
 > run at 1799 for 3 seconds.. in reality, this is just
 > a phase locked loop,
 > and you don't need the step functions...

Yep, it all depends on what the basic stamp can do
and how quickly it can process the timing signals.
I've seen that the faster ones can send pulses that
are mere microseconds long, so it shouldn't be too
difficult.

 > >Timing signal -- this is the hard part, it involves
 > >two things: a) figuring out where in the
 > capacitor's
 > >charge cycle the spark gap is firing,
 >
 > An optical detector will do quite nicely for telling
 > when the gap fires.
 > very bright light, fast rise time.

That would work -- a coil of wire around the
the leads to the spark gap would sense it too.
I'm sure with 1000A of pulse discharge at RF,
SOMETHING would be induced into it that could
be read. (Observing proper distances of course.)

A sensor across the safety gap isn't a bad idea either
and can detect when things are going wrong and shut
the coil down before bad things happen. (ie long
before *I* notice something is wrong)


 > >and b)
 > >quantifiying the capacitor voltage when this
 > happens.
 >
 > Yes, this is a bit of a challenge..

In a nutshell, I want the stamp to know a) WHEN things
happen, b) QUANTIFY what is happening (ie time
duration, how much voltage/current ) then use that
information to automatically adjust what can be
adjusted -- rotating primary tap, spark gap RPM/Phase

Regarding the primary:

Would also be fairly easy to create a small 'crystal
radio' tuned to the secondary+toriod's resonant
frequency and move the primary tap such that signals
received achieve maxium gain. The problem there is
that
the frequency of the reciever must be re-tuned every
time the coil is turned on and the point of this
excersize is a 'plug'n'play' coil :)

Ultimately I want to be able to shut the thing off,
grab another toroid, throw it on the top, grab another
power supply, connect it up. Then, turn it on, and
watch the primary tap and spark gap MOVE to an optimal
point and stop, all while recording readings to the
stamp for futher analysis later...or...send it via
a serial stream to my laptop via RF carrier.










__________________________________
Do you Yahoo!?
The New Yahoo! Search - Faster. Easier. Bingo.
http://search.yahoo-dot-com