Water Sparkplug

[gmeast]

3055

Hi, everyone.
Guys, as I recall, the 2N3055 can drive much more power that it does in this circuit that you experiment with right now.
It seems that they do not open totally. So, maybe you should lower the value of the 220R/0.5W resistors. Put 100R/5W instead and compare the results. That will probably modify the oscilating frequency. Of course it will draw a little more power when unloaded, but will give more power at full load.
The bigger the current injected in base (gate), the bigger the current through the transformer, so the bigger the power transferred through it.
2N3055 can drive up to 15Amps (C-E junction) if fed with 7Amps in base.
Best wishes.

Hi everyone,

Looking at the trace I see it is a little 'roundy' under load and that makes me think also we need more base current (ain't turning full 'on' - in a linear mode). The 3055's are warm, but the numbers are better than with the other oscillator. I get 22 - 25 watts with a 1K Ohm R load and around 150 VDC on the scope. Also my Amp meter reads around 1.8 - 2.5 Amps and 12.5 VDC at the battery so the numbers agree. I agree, lets open it up to 100 Ohms on the base and square it off. I'm running a 1 uF electrolytic along with the 12kR and am at 57 Hz ... That RC agrees w/theoretical too.

Thanks for everyone's input,

Greg

[magnetO]

HV Polarity? Depends on Ignition System.

hi,

but both of them lit up.

graham

Hi Graham,

it would be good to know what kind of ignition your engine is driven from.

A modern transistor ignition does multi spark per engine stroke. The 1., 3.,.. goes up(or down); 2.,4.,.. goes down (or up).

The standard CDI (capacitve discharge ignition), CAN make multi spark (down and up) per engine stroke.

Kettering Ignition
Kettering Ignition
does normally make only one direction.

So, what igniton is in your engine?


@insane4evr,
@kinetix,

I did grasp deep at the trash can, and get the thrown away diodes again. Made a string out of some half-lethal diodes, but all at about 20MOhm (no real bad ones, no good ones), and (again) only 25 diodes. The string blocks. It works!

Havent drive much km, but it works.

Not sure what kind of diodes (NTE517 ??) next order will be.

magnetO

[sa14785528]

hi Jetijs

thanks for the speedy reply! thats a great way of testing, should have thought of it myself, damn!

cool, so im almost there! wana see that big plasma ball!

thanks

graham

[Jetijs]

I am using 0,05uF caps and the oscillation frequency is 2,23KHz.
This is a waveform of the input at 75W load on the output. The base resistance is 220 Ohms. Input power 12.6V 3.4A. The bulb fillament just starts to glow dimmly.



And this is when I lowered the base resistance to 50Ohms. Now the 75w bulb turned on at moderate brightness. The amp draw increased to 5.3A.

[sa14785528]

hi magneto

thanks for the help, i used jetijs' method and got the system running! that plasma spark is a beautiful sight and sound!

cheers

graham

[gmeast]

Oscillator

Hi everyone,

This oscillator Phil (aka pmonk) posted is great. The power consumption, load values, scope data all match. I don't understand the heat on the 3055's but it went down and the output power went up when I went from 220 Ohms to 110 Ohms on the Base of the 3055's (thanks for suggestion Kinetix). Even if I need to cool the xistors that's fine because it consumes far less power than the other oscillator.

Greg

[Jetijs]

I agree with that.
Greg, did I understand correct and your oscillator is running at 57Hz?
also, what transformer are you using?

[Chris31]

Hi everyone,

This oscillator Phil (aka pmonk) posted is great. The power consumption, load values, scope data all match. I don't understand the heat on the 3055's but it went down and the output power went up when I went from 220 Ohms to 110 Ohms on the Base of the 3055's (thanks for suggestion Kinetix). Even if I need to cool the xistors that's fine because it consumes far less power than the other oscillator.

Greg

Many post ago I have suggested the transistor should be driven hard else you will never use its full switching potential, as a result you are going to get drop in efficiency.

For some reason no one understood what I meant.

Well its sorted now, thats all that matters

[Chris31]

Chris,

These are the specs for the plug:

Champion RDZ19H
10mm
3/8" reach
5/8 socket

I called Champion to find a non-resistor plug. When I told him I wasn't too concerned with the reach (determines if it's a hot/cold plug), he asked me what I was burning. I told him I was trying to ignite water vapor. 'CLICK'. He hung up on me.

Dave

LOL what did you expect?

Anyway as for the plugs, unfortunately this plugs seems a rare one. I cant find NGK or bosch making an equivalent, champion does not list a non resistor version either.

[mxt]

Chris,

These are the specs for the plug:

Champion RDZ19H
10mm
3/8" reach
5/8 socket

I called Champion to find a non-resistor plug. When I told him I wasn't too concerned with the reach (determines if it's a hot/cold plug), he asked me what I was burning. I told him I was trying to ignite water vapor. 'CLICK'. He hung up on me.

Dave

Hello Dave,

Here is what I found. The difference is that your plug is with tapered seat but Non-resistor plugs are with gasket seats. Check this link: Spark plugs

Miks

[Ursine]

I may end up drilling the center out and epoxying in another electrode.

Dave

[gmeast]

transformer and Hz

I agree with that.
Greg, did I understand correct and your oscillator is running at 57Hz?
also, what transformer are you using?

Hi Jetijs,

Yes, I am using the 12k Ohms and a 1.0 uF cap on the PNP transistor's base circuit ... it is the RC. The scope actually reads 100 Hz but then jumps around. The total square wave period is 17.5ms though and the wave is not sine and it is jagged so the scope screws up and you have to compute it.

The Transformer is a Tamura PF24-10. It is 24VA.

http://home.pacbell.net/gmeast/VW/PF24series.pdf

Greg

[Jetijs]

Thanks Greg
Can you point me to a formula how to calculate proper RC?

[gmeast]

RC

Thanks Greg
Can you point me to a formula how to calculate proper RC?

Hi Jetijs,

I'm sure you already know this but it's the product of the resistance to the flow of current charging a capacitor and the capacity of that capacitor and regardless of the voltage potential ... RXC.

Bleed current from 12 VDC with a resistor into a capacitor that has a common ground and it will take '1 time constant' to charge that Cap to 63% of 12 VDC. Take the same resistor and capacitor and charge with 100 VDC it will still take exactly '1 time constant' to charge that cap to 63% of 100 VDC. So voltage doesn't matter except for component rating.

1000 Ohms and a 1 uF capacitor:

1 time constant is R x C = 1000 x 1/1000000 F:
1000/1000000 = 0.001 seconds.:

Charging from 100 VDC you will get to 63 VDC in 1 ms.

Because of the non-linear charging curve it will take '4 time constants' to fill the cap and '5 time constants is considered 'steady state'.

In the inverter circuit there is interplay between the 220 Ohm base resistor, leakages from both transistor pairs, the transformer resistance and a host of other stuff so it's not a simple calculation to determine the actual frequency from a design end.

Hope this helps,

Greg

[maxc]

When I use a 3055. I take a nte128 or a nte129 too dirrect drive it. You get more gain. Then you can drive it with a 1000ohm resistor.