Most people use the slight gap, and let the electric field pull the water up to the electrode method I believe. I use the submerged electrode, with a glass tube over each wire so that only about .1" of tip is showing. Then pump the water through the chamber at a rate of about 1 gallon per hour for every 30 mA of current. I also cool the water down to just above freezing before it enters the chamber.
Marshall [email protected] wrote: > Thanks for the response Marshall. > > I've seen a few circuits that use a 555 timer with a power transistor > or an SCR on the output to drive an autotransformer that puts out > about 20 KV. I was hoping that I could use stuff I had lying around > the house for the HV source. I guess I need to break down and buy a > neon sign transformer. Do you submerge both of your electrodes or > leave a small air gap between one of them and the DW? Any advice is > greatly appreciated. > > Thank you for your time. > Andy Scott > > Date: Mon, 02 Sep 2002 11:51:23 -0400 From: Marshall Dudley > <[email protected]> To: [email protected] Subject: Re: CS>HVAC > > > An auto spark coil would be problematic. First it would be difficult > to get the current you need, since they step the voltage up by over > 1,000:1 I believe. It would require an amp to generate a milliamp of > current on the secondary. Second, the coil is made to operate at high > frequencies (that is a rapid rise and fall time). To allow the silver > time to aggregate into particles, or move sufficiently away from the > electrode before reversal (and it will reverse, even if the waveform > is asymetrical unless you put a high voltage diode in the secondary) > would require very high voltage. > > Lets take an example. To make a gallon an hour at 60 htz requires abou > 10K volts and about 25 mA of current. With a spark coil, which > typically has a 10 microsecond pulse before the leakage inductance > shorts it out, it would require 50,000 pulses a second to maintain the > same duty cycle. But this would quickly burn the coil out, since it is > made for duty cycles of maybe 1/100 of that max. So if we run it at > 500 pulses per second, we are running at about 1% duty cycle. The > amount of current necessary would need to be about 100 times large for > the same production rate. Thus we need about 1.4 Amps of secondary > current. With the step up these have, that would require about 3,000 > amps on the input. And since the frequency is about 8 times as high, > the current needs to be about 8 times as high as well, or about 30,000 > amps. Of course if you want to make it at a slower rate, then you > could use a lower amperage, of maybe 1 amp, and make a gallon in about > 30,000 hours. > > Now I could be off by a factor of 2, or even an order of magnitude on > some of these estimates, but the result would still be the same. I > believe it would be impratical. > > Use a 15 KV neon sign transformer like I do, and it will work fine. > > Marshall
