Re: Final Capacitor Test Results


>I wonder whether there isn't some kind of resonant charging going on
>and adding the resistance completely killed the Q of the input choke
>(welder)? What do others think? How could adding that resistance kill 
>the system so completely? Please, if anyone has ideas let us know. I 
>for one hate unexplained mysteries. 

I can confirm observing this with my 9mH choke and 5 or 10 ohms of
parallel resistance. I've been planing to measure the 220 primary and
HV side of my system to find it's explanation but, my garage is almost
empty now for a new garage door installation tomorrow. I'll try to get
in some LV and HV measurement in the next day or two.

>It was interesting what Richard 
>Quick said about synergism in the system. At 60Hz, the cap reactance
>is about 106kOhms, and the reactance of the primary coil (assume 
>around 100uH) is a fraction of an Ohm. So the capacitave reactance
>transformed back through the pole pig to its primary is the dominating
>reactance seen by the mains. I wonder if this was in resonance with
>the welder? Any ideas on this one are most welcome. I'm only sorry I
>can't be there to measure it.

--------- begin, includes of some of my previous musings -------

Ah Ha! 65^2 * 5 ohms = 21K ohms * 6nF => RC = 127uS about 8kHz. maybe
that was why my resistive only ballast worked so poorly?

And when the gap is on, the choke is charging. When the gap opens, the
choke ramps up the voltage on the pig MUCH faster than just a plain
resistive ballast.

---------------------- a couple of days later------

220V in. 10 ohms resistive -> pole pig primary: pig secondary into
rotary (2 times 0.03"/0.7mm) spark gap. No Cap, no TC primary. Lets
just see if we can strike and quench this thing? Very poor gap
conduction. A few firings per second, at up to 2670 presentations per
second! That's 10K RPM. 20A in the 220V primary circuit. All reactive

Took out the resistive ballast. Put in 9mH of inductor. (my foot in
mouth inductor of last week) Less than 10A 220V primary current, but a
very smooth firing, with LONG 3" power arcs from stationary electrode
A to stationary electrode B. 

A and B are one and a half gaps apart. C and D are 180 degrees from A
and B.

Put in 10 ohms of resistive ballast, parallel > 20A; series < 20A but
not much difference in gap quenching.

Tried no ballasting. Figured that the rotary would quench;) ;) ;) ;)
One shot mode;) Pegged the 100A meter on my current shunt - tripped
the 50A circuit breaker. I didn't have my eye on the gap - I missed
it's performance.

Then I tried the above with my 6nF cap across the pig secondary. What
a difference in the firing rate and in the quenching. All combinations
were smooth, short (1/2" max arc, as opposed to 3+"), and so much
sharper in sound that I had to put on hearing protectors!

------------------ end include ------------------------

My current (npi) theory is that when the shunt gap conducts, it's
impedance ~10? is transformed by the pig to ~0 and charges the ballast
inductor. When the gap quenches, the magnetic field around the ballast
inductor collapses and boosts the primary voltage to the pig much more
rapidly that a resistive ballast will (see the first paragraph of my
included text). Putting in to low of a  resistance in parallel with
this choke will over dampen it, eliminating it's boost effect.

Also, as you can see from the last paragraph of my included text the
TC cap does have an effect in this circuit.