Water Sparkplug

[gmeast]

inverter

500mA draw sure sounds good to me!! With that small of a parts list, and that small an amp draw, I could use two units If I needed to boost if it wouldn't keep up with the V-8! From what I see, the caps on the board are 47pF at 50V rate? The blocking diodes after T1 would be 600V rate at 500mA? That part looks like just a voltage doubler, so we just hook our existing ones to T1's primary, right? I will be very joyous if this will work for us. I guess the only question then is..............will it make good solid plasma? Later.............................Mike

Hi Mike,

I'm not sure right now, but the secondary is hooked to ground via a diode in Phil's inverter. That leaves us without isolation. What's your take on that?

Greg

[jstadwater]

Hey Greg,

I think, I have fried too many inverters before isolation was used, I'd be afraid to try it now without isolation. Essentially, the diode that goes to ground, and the 20uF cap that goes to ground.....are both conected via ground. We lift both components from ground and attach them together, they are still connected the same way, but now we have ground isolation.

He stated he has cyl 1,2, and 3 hooked up and working. He also stated that the coils for those cylinders fire HV+, and the others fire HV-. I don't know which spark circuit he is using, but since our MW caps shoot HV+ to the engine ground, might just get a fried circuit without isolation. With our circuit we would basically be hooking (if wired per schematic) the -Neg. end of the diode and the -Neg. end of the multiplier cap straight up to the +Pos of the MW discharge cap......I think that will let some smoke out of something!! Definitely pull those puppies from ground and hook them to each other instead.......and tap in right there to get your -Neg output DC lead..........That's my take on it........................Mike

[pmonk]

Hi Mike, Greg

Mike, the capacitors are .047uf. Greg, the transistors I used are 2N3055 and MJ2955. I doubled the wattage on the resistors just to be safe. I do not have to use a cooling fan on these transistors. Just use heat sinks and they barely get warm. I am using some 1N5408 diodes and 100uf 400v caps for the doubler. this give me 345 volts out of the doubler. From there I go through an inductor. It is just a fan coil I scrounged out of a microwave. Just disassembled and used the coil. It reads 90 ohms on my meter. On the other side of coil I have a 2uf microwave cap. Then I have a string of 55 1N5408 diodes. I loose about 20 volts through these diodes, so the end result is about 320 volts to the plug. I have the output tied to a three way splitter and use quad shield cable to go to the plugs. I am using NGK non-resistor plugs. I bought some spark plug checkers from Harbor Freight and took them apart so I could use the boot and the plug to assemble a cable for each spark plug. I have inserted a spark gap in the cable just as extra protection for my coil. I made these using a ball point pen amd some all thread rod screwed into the 2" section of pen. I have included an attachement of the cable assembly.

Later,
Phil

[Kinetix]

Hi guys.
I am glad that new ideas come around.
I'll post another oscilator schematic that works. I do not know the eficiency, maybe some values could be modified (C1, R4, R5 or C2) but I remember I made it and it worked. You also may try other types of NPN power transistors (e.g BU526A, 2N3055, etc.). Some of them will work better, some worse...
Oscilator will charge the capacitor C2 only when points are closed. So you don't need a power resistor or bulb to limit the current while capacitor discharges. Anyway, if you like the bulb setup more, the schematic can be modified accordingly.
All the best.

[dllabarre]

charge pump

While we're on it.
I found this schematic of a charge pump for plasma a couple months ago.
I don't remember where though.

DonL

[pmonk]

Hi guys,

I am not using the doubler circuit shown on the schematic I posted. I am using the circuit shown on the attachment. The diodes are 1N5408 and the caps are 100uf 400v.

[jstadwater]

That's the doubler arrangement I was talking about changing the schematic to, and NOT hooking to ground. Later..............................Mike

[gmeast]

output isolation

That's the doubler arrangement I was talking about changing the schematic to, and NOT hooking to ground. Later..............................Mike

Hi everyone,

This is all good to see.

Phil:
That's what I figured for the MJ2955. I ordered some. I'm going to go ahead with my present inverter/oscillator 'cause I'm messing with super-hydrating the inlet air with steam from recouped exhaust heat and don't need the charging capacity for anything else. But I'm going to build the unit you posted now for kicks and eventually implement it ... and thanks again for posting

My goal right now is to conquer an ugly oil leak in the Bug and get back to testing.

If anything 'good' has come from these inverter exercises, it's the honing of our collective abilities to understand, build and improve the tools we need and use. So ... it's ALL good !

Peace,

Greg

[magnetO]

Lee, smw1998a still here? Duration of spark?

Hello All,
I have attached two scope shots regarding the above observations.

All The Best Lee...

Hi Lee,

still looking sometimes in?

Could you explain, or could anyone other who is familiar with scopes explain, please, how long is the spark duration on the scope screen shot?

Time scale (horizontal) is 50 mikroSeconds per dev (=? div, division, dividor?)
So, how long is the spark duration?
a) Is it ~2.2dev * 50us = spark duration ~110us ??
a) At 10dev = 1 scope screen length ??
or
b) Is it ~12dev * 50us = spark duration ~ 600us ??
b) At 50dev = 10 grids = 1 scope screen length?


Why needing?
Have no scope, but need to know, how long normal spark (~110 or ~600 us), and as you, Lee, found plasma is about ten times faster, how long is plasma spark (~11us or ~60us) ??

Why needing?
Having destroyed lots of diodes. Most of them, I guess, already "shocked" on soldering: Did make those long chains of 1000V diodes, shorted the wire because of total length, brought too much heat in them on soldering. Half lethal. Driving, getting increased exhaust temp (diode chain not far from exhaust). Full lethal.

Diode current, diode surge current IS NOT the problem with this 1000V diodes (N5408), I guess.

Max. diode surge current WILL BE perhaps a problem with the more expensiver HV diodes.
Only found adequate NTE517. What diode else?


For current limiting during plasma with pre resistor, and not destroying too much HV diodes, the question:

How long is spark duration time on foto: ~110us or ~600us??

Thanks a lot
magnetO

[Ansis Freimanis]

110us

F(frequency)=1/T(period)

You have 110 microseconds, but You must calculate frequency.
frequency = x (we dont know how it is), we calculate it:

x(F)=1/110x(1x10-6)=1'000'000/110=9090Hz=9.09KHz

If You have 1 ms= 0.001 Sec
F=1/1x(1x10-3)=1000/1=1000Hz=1KHz

Be, happy!

[insane4evr]

Use the squares

Volts/div or Time/div:

On the scope screen (graticule) markings, use the squares, not the tiny division lines on the cross-hair. These tiny division lines are fractions of your volts/div or time/div.

It is like your tape measure. It has numbers every centimeter which have larger lines but it also has smaller lines every millimeter but no numbers. Not sure if this clarifies or confuses.

[magnetO]

110us and use squares. okay

Volts/div or Time/div:

Hi Ansis Freimanis,
hi insane4ever,

so you are saying both the same: answer is a).

(Not the charging time for cap or frequency of re-charging is in the moment important to me. But spark duration, and Lee said plasma spark is 10 times or even more faster, but how fast exactly.)
http://www.energeticforum.com/renewa...html#post26418

Thank you
magnetO

[gmeast]

EXAMPLE: VexUs Charge / Discharge Trace

Volts/div or Time/div:

On the scope screen (graticule) markings, use the squares, not the tiny division lines on the cross-hair. These tiny division lines are fractions of your volts/div or time/div.

It is like your tape measure. It has numbers every centimeter which have larger lines but it also has smaller lines every millimeter but no numbers. Not sure if this clarifies or confuses.

Hi all,

Here is an example of one of the charge / discharge curves from my 4-cylinder VW engine with the VexUs plasma spark. You can glean a HUGE amount of information about "what's goin' on" from a simple trace like the one below:

The horizontal and vertical scales are:
50 VDC for each major division / square on the vertical scale
and
.0025 seconds (2.5 ms = 2.5 thousandths of a second) per division / square on the horizontal scale.

It shows a rather classic charge of the 'dump' capacitor (curvy part) and then the plasma discharge when the spark plug fires (nearly vertical). The plasma lasts until the voltage across the gap has diminished to about 10 - 20 VDC ... that's amazing to me. It only lasts a few nanoseconds (very steep slope)-a 200 VDC drop in a few nanoseconds !

It shows a pulse period of (3 x 0.0025) = 0.0075 and that is a frequency of:
1 / 0.0075 = 133 Hz.

There are 4-cylinders so:
120 / (4 x 0.0075) = 4000 RPM
(the "120" is all of the constants and conversions reduced to a single term)

... just some real-world application sort of stuff. Peace,

Greg

[Ansis Freimanis]

to MagnetO

I will say- intensity of discharge. In moment when you start hear loud sound - than your spark energy is more than speed of sound- 343 m/s.
P.S. I am not engineer of sound barrier, but it is true- You hear it with your ears.
Be happy!

[Ansis Freimanis]

to Greg

Yes the discharge is very fast- it is like shoot with fire arms- sound barrel is broken, and we can hear "bang".
P.S. In oscilloscope we can’t see that, but we see absolutely precise- like a cut with super sharp knife- discharge- absolute "zero" in nanoseconds, from 174 Volts...
Cheers!