User-agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.9; rv:24.0) Gecko/20100101 Thunderbird/24.6.0
On 9/19/14, 6:06 AM, Dave Boyle wrote:
We're all on the same page. I have a Cortex M0 based microcontroller and
I'm setting up a system for assembly language programming using mcu on
eclipse. I have an optocoupler that I will protect with a series
capacitor and connect to the ac line to get a pulse stream of 120 per
second. I'll get a hall effect sensor that will hopefully detect my 6
copper clad tungsten electrodes spinning on my disk. If not then I can
embed some bits of metal in between the electrodes which is where I need
them anyways. Now I have a second stream of pulses that I can speed up
and slow down by outputting to a power control device like the vfd. My
algorithm will be based on a division by six but there will be lots of
other things to do as well. I'll switch my nst's by a solid state relay
so if the mcu ever fails they will disconnect immediately. This should
become the standard method of phase syncing any motor into synchronous
action. If I had a lab full of scopes and test equipment I might try and
do the same thing with a cmos phase locked loop that sells for 29 cents.
But I like the control that I get with an mcu.
magnetic sensors might not be the best approach (as others have commented)
Optical would work well, and you can use the cheap 1mm plastic fiber
(aka TOSlink) for short distance runs, so that the circuitry isn't near
the spark gap or TC primary.
However, it occurs to me that another approach might be to just
optically detect the light from the spark in the gap. You can easily
get the line voltage sync into the MCU, and what you can do is make the
sync pulse from the spark detection line up relative to the peak of the
line voltage (since that's presumably what you want)
BTW, the "death to NST" occurs when you don't get a spark (or it's at
the wrong time), and the voltage comes up too high. So what you would
do is set up your logic so that if you don't detect the spark, you shut
down the NST. You've got 8 or 10 milliseconds to decide and that's
loads of time with modern MCUs.
A potentially more reliable way to do the switching, by the way, is to
generate pulses for the SSR (at the correct phase relative to line
voltage), and just inhibit the pulses. Rather than having a static
on/off that you hope to turn off quickly enough. Although both will
work. A side benefit of the "explicit pulse to fire the SSR" (which is
likely optoisolated SCRs back to back) is that you can do phase control
power adjustment (by delaying the trigger pulse from zero crossing).
Now, of course, being a triggered spark gap guy, myself, I'd just say
why bother with whirling pins of death (aside from the obvious visual
appeal) and use your MCU to generate trigger pulses to an ignition coil
that fires a triggered blown gap.