Water Sparkplug

[gazzzwp]

Apologies!

Smw1998a and Aaron

Apologies for my previous assertion - I really thought that frequency would be the limitation.

So please could I ask at what frequency was the video made? Would it wear out the spark plug? Is the current to high to be feasible? Does it still ignite water? What is the next step?

Gazza

[multiuser]

@smw1998a

Hello! Congratulate, well successfully (http://de.youtube.com/watch?v=vdxBbb-tR3s)! Unfortunately I do not have so much luck to copy a such circuit. Can you give us please an exact connection diagram, which you used in the video? thanks in advance!

YouTube - Videos von multiuser0815

[smw1998a]

Water Sparkplug II

Hello all,

At the moment I don’t have a spark plug with no internal resistance to test. The circuit does react when water mist is sprayed on the electrodes but I only get reliable firing with a smaller gap.

The frequency of the pulses from my circuit is 20Hz. The voltage is 12v and the current draw is 1 amp. That’s one pulse every 50 milli seconds. This frequency is well within the limits of a power relay but, unfortunately, this is the maximum output I can get with the charge pump I have in this circuit, which isn’t very efficient.

The charge pump is a 28VA 230v-12v transformer. The 12v secondary winding is connected to a 12v battery and 2N3055 transistor driven by a pulse generator. The generator consists of two 555 timers, one set as an astable running at a frequency of 1.4kH and the other is an edge triggered monostable with a pulse width of around 0.01 milli seconds. The astable triggers the monostable and the monostable drives the base of the transistor. The 230v primary winding is connected to the FWBR and the 4uf 400v capacitor.

The logic circuit controls the pulse generator, switching the charge pump on for 40 of the 50ms period, then switching the charge pump off, once this is done a pulse is sent to the opto coupler to discharge the capacitor which takes less than 1ms, there is a rest period and then the cycle starts again.

I found the circuit quite hard to tune. The balance between charge pump and discharge is very fine. To little separation and the SCR will not switch off correctly. To little duration on the charge pump and the cap wont hold enough voltage. To greater delay between charge and discharge and the cap will lose voltage.

I don’t have all the answers but solid state is more reliable than a relay. If the cap can be charged fast enough then pulse frequency could be brought down to 5ms and less, well beyond the switching capability of a power relay. Only then can a true assessment of heat and wear be made and alterations to the circuit, where necessary, can be carried out.

I will attach a component description of the discharge circuit. The logic circuit is very much a work in progress and is far from complete so this will take some time. I hope to get the circuit down to the 5ms range or less.

All The Best Lee…

[lighty]

@smw1998a

I was planning on making a somewhat more elaborate circuit however I don't have time at the moment to develop it but what I can do is to respectfully help you to refine yours:

1. Use triac (preferably alternistor sort) instead of thyristor. It would allow capacitor/primary circuit more free oscillations. It should also be
2. Your optocoupler/transistor combination is not impervious to self-triggering or latching up. You should rather consider using dedicated triac triggering optocoupler incorporating diacs. One without the zero crossing detector that I often employ is MOC3021 and I suggest that you use additional snubber network to prevent uncontrollable triggering.
3. What you should also do is to employ another alternistor between AC power supply and diode bridge. You could use another MOC30xx optocoupler but this one with zero crossing detector (so that you don't need additional serial resistors).
4. You should employ additional circuitry (bistable logic comes to mind) to have those two triacs (the charging one and discharging one) working in opposite cycles. It's necessary to prevent latching up of the discharge triac when the capacitor is being discharged and AC mains suddenly start pushing current into the system. The reason is that if primary impedance is sufficiently low the current will flow from bridge (effectively from AC source) and back through primary possibly keeping the discharge triac opened.
5. A slight pause have to be introduced between those two cycles (during which both triacs are not driven) in order to allow for the charging triac to close completely during the zero crossing point of the AC cycle.
6. The periods of charging cycle can be calculated by common formulas that takes into account AC voltage and frequency as well as the capacitor capacitance. If you calculate it properly you can manipulate charging time in order to have faster charging by introducing higher AC voltage but shortening the charging cycle. That would allow for higher charge/discharge frequency.

One must also have in mind fact that triacs and thyristors are not high frequency devices. Also, the bigger the current rating the slower it is to open thus lowering it's maximum frequency even more. As I can see you already know that.

If you have time and knowledge you can easily make suggested adjustments and significantly raise efficiency and max frequency of your circuit. If not, please be patient since I will most probably have some spare time next week to develop full circuitry myself. I would prefer to use microcontroller (as I'm sure you would too) to allow for simple, precise periods tweaking but I guess I will use one or two 556's in order to make things simple and cheap enough for amateurs to build.

That being said I respect your idea and I just want to help you refine your circuit (that is if you need my help). I will gladly discuss further development with you over PM or mail (in order to keep the thread clean from unnecessary details) but I'm off untill tuesday. Keep up the good work and please don't be offended by my suggestions.

[gotoluc]

Hello All
@gazzzwp
Looks like some of you are discovering what I suspected a few days ago from my own trials - that the effect is not reproducible at faster frequencies.

I disagree. Switching speed can be increased however, so will current draw. How far the frequency of the ignition coil can be pushed in this circuit is unclear. In a car using contact breakers, the time of induction is limiting factor in the operating frequency of the coil. In our circuit the time it takes to charge the capacitor is the limiting factor using solid state switching. The relay slows the circuit considerably.

Here is a link to my solid state experiment.

YouTube - Water Sparkplug II

Regards Lee...

@smw1998a, excellent circuit development and video presentations.

Thanks

Luc

[Aaron]

scr discharged cap

What is the next step?Gazza

Follow the concept of what SWM is doing with the SCR discharged cap. As long as the power source keeps the cap charged up, you cannot get a frequency too high with an engine that the circuit cannot keep up with.

This is basically already what an off the shelf CDI does except it is charged from a 12v source.

[Aaron]

great spark vid!

Great vid and thanks for bringing awareness to Energetic Forum!

Hello All
@gazzzwp
Looks like some of you are discovering what I suspected a few days ago from my own trials - that the effect is not reproducible at faster frequencies.

I disagree. Switching speed can be increased however, so will current draw. How far the frequency of the ignition coil can be pushed in this circuit is unclear. In a car using contact breakers, the time of induction is limiting factor in the operating frequency of the coil. In our circuit the time it takes to charge the capacitor is the limiting factor using solid state switching. The relay slows the circuit considerably.

Here is a link to my solid state experiment.

YouTube - Water Sparkplug II

Regards Lee...

[rickoff]

Follow the concept of what SWM is doing with the SCR discharged cap. As long as the power source keeps the cap charged up, you cannot get a frequency too high with an engine that the circuit cannot keep up with.

This is basically already what an off the shelf CDI does except it is charged from a 12v source.

Aaron is quite right. CDI is perfectly capable of charging its main cap at red line maximum engine rpms. CDI basically takes the 12 volt battery input, uses a step-up transformer to boost the voltage to between 400 and 600V, and a charging circuit using a rectifier (to prevent premature discharge) charges the main cap. At this voltage range, the cap is charged nearly instantaneously. A triggering circuit then breaks the charging circuit, allowing the cap to discharge through the ignition coil primary.

Lee, have you measured the actual voltage applied to the cap through your ss charging circuit? Nice diagram, and I enjoyed your video.

Best regards, Rick

[Jules Tresor]

>I can't find the photo of S1R's spark plug. Can you post a copy here?

I also don't have any pictures of his plugs. I was just quoting his message, he never show picture I think ...
tq
MDG

[Peter Lindemann]

Great Demo, Lee

Lee,

Great demo of the SCR capacitor dump system. It shows that higher speed operation is possible and that solid-state systems are the way to go to get up to "engine firing speeds". It also shows, with out a doubt, that Luc has given us a "miracle" breakthrough with the HV diode trick!

Your control circuit seems to work very well, but it might be unnecessarily complicated. I remember when I was working with John Bedini on the first SCR cap-dump battery chargers run from the SSG, that the SCR was able to reset itself as long as the capacitor was being charged with intermittent pulses. As long as there was a pause in-between the charging pulses, the SCR would reset automatically, and we didn't have to turn the charge system off to dump the cap. Even some of the schematics Aaron has shown me of commercial CDI's don't look like they shut off the cap charging circuit when they dump the cap with the SCR.

Just thinking out loud here, but there might be a simpler way..... (Lighty, I consider your suggestions the "professional" way to go for a circuit that will work perfectly!!! I'm just trying to get more "backyard tinkerers" to try the experiment, too.)

Actually, that you are getting all of that PUNCH from a 4uf cap is very impressive!

Peter

[Aaron]

metal spheres

SWM,

Do you have any tiny metal spheres you could put that spark to? I might get some copper bb's to solder at the end of some copper rods. I want to see really what the spark difference is with the same power with and without spheres.

[smw1998a]

Water Sparkplug II

Hello All,
Thank you all for your kind comments and input. My motivation for the circuit was to see if I could replicate and then study the effect. I had no relays to hand that could handle more than 30v-50vDC, so I tried the solid state route as the frequency limitation of the relays is obvious.

As Lighty observed my component selection in the discharge circuit can suffer from latch up and self triggering, although the self triggering appears to have been eliminated by changing from CMOS 7555s to bipolar 555s. I have not suffered uncontrollable firing as yet. Latch up can be a problem but with a smaller spark gap, is much less so.

The frequency of the charge pump can be a problem with the SCR as Peter noted. The main problem is a very short duration pulse into a low resistance winding, transformed across a high resistance winding at 1.4KHz in a mains transformer designed for 50Hz. At this time I’m more concerned about the charge pump than I am about the discharge side of the circuit. While I don’t doubt the discharge circuit can and should be improved, I’m at the edge of my abilities component wise. I do hope to get a few more tens of hertz out of this arrangement though.

The need to stop the charge pump before discharge was two fold in this circuit. One; to try and prevent the SCR from not switching off and two; To reduce the current draw of the circuit. In early experiments I left the charge pump on all the time and dumped the cap manually by shorting the positive line or the negative line. This worked well but when I added the SCR there was a momentary very heavy current draw of three to four amps just after the moment of discharge, if the circuit didn’t latch up.

The circuit, as I have it at the moment, with the charge pump switching off before the cap dump, draws 1 amp of current from the battery. In my video, I put a analogue meter between the battery and circuit and the needle doesn’t move when I change from a standard HV spark to the diode enhanced HV spark although this is difficult to see. The two still photographs show both types of arcs. The thinner arc had it’s brightness brought right up within Photoshop, whereas the diode enhanced spark created its own light. Both shots were taken at 1000th of a second with indoor light.

The effect is real and astonishing. Connecting the HV diode drew no more current from the battery. As the circuit contains an SCR, multiple oscillations between the cap and coil are unlikely, I think.

As Rick and Aaron pointed out, what I have built is a very basic CDI system. It may be possible with a true CDI system (not a solid state switched induction coil) the inclusion of the HV diode may be all that is required to produce the effect in that system. Although, in my experience, the coils from a true CDI system are much smaller than the coil I used (which was inductive). Whether you could run an engine from this remains to be seen. Is the water actually igniting or is it rapid expansion through localised super heating and vaporisation around the arc?

Aaron, I will look for some spheres but tests I have already carried out showed that the standard arc will happily jump a gap of 25mm or more at this frequency and slightly higher (any further and I get arcing between the negative input of the coil and the HV output). Connecting the HV diode kills the arc and no arc is evident although everything is working as it should with gaps greater than 12mm. The smaller the gap the more reliable the enhanced spark is. In fact, with a much smaller gap I could probably increase the frequency as is. As Peter said, what I’m getting from a 4uf cap is already impressive. Photo caps work well but take to long to charge so frequency drops right down.

Rick, I’m back at work this week but I will try to scope across the cap and check voltage and see what I can see. I already lost a DMM to this circuit.

The best to you all Lee…

[ren]

Hi guys,

I have my hands full with other replications at the moment (Im not superman like JETIJS is ) but I stumbled across this the other day while looking for parts.

I dont know if it is of any help or has already been mentioned, I just saw Cap Discharge and thought of this thread, along with other Bedini devices...

Jaycar Electronics

[adamhy]

Hi, smw1998a,

forgive my stupid, which H11D1 do you used in your circuit?

Hello all,

At the moment I don’t have a spark plug with no internal resistance to test. The circuit does react when water mist is sprayed on the electrodes but I only get reliable firing with a smaller gap.

The frequency of the pulses from my circuit is 20Hz. The voltage is 12v and the current draw is 1 amp. That’s one pulse every 50 milli seconds. This frequency is well within the limits of a power relay but, unfortunately, this is the maximum output I can get with the charge pump I have in this circuit, which isn’t very efficient.

The charge pump is a 28VA 230v-12v transformer. The 12v secondary winding is connected to a 12v battery and 2N3055 transistor driven by a pulse generator. The generator consists of two 555 timers, one set as an astable running at a frequency of 1.4kH and the other is an edge triggered monostable with a pulse width of around 0.01 milli seconds. The astable triggers the monostable and the monostable drives the base of the transistor. The 230v primary winding is connected to the FWBR and the 4uf 400v capacitor.

The logic circuit controls the pulse generator, switching the charge pump on for 40 of the 50ms period, then switching the charge pump off, once this is done a pulse is sent to the opto coupler to discharge the capacitor which takes less than 1ms, there is a rest period and then the cycle starts again.

I found the circuit quite hard to tune. The balance between charge pump and discharge is very fine. To little separation and the SCR will not switch off correctly. To little duration on the charge pump and the cap wont hold enough voltage. To greater delay between charge and discharge and the cap will lose voltage.

I don’t have all the answers but solid state is more reliable than a relay. If the cap can be charged fast enough then pulse frequency could be brought down to 5ms and less, well beyond the switching capability of a power relay. Only then can a true assessment of heat and wear be made and alterations to the circuit, where necessary, can be carried out.

I will attach a component description of the discharge circuit. The logic circuit is very much a work in progress and is far from complete so this will take some time. I hope to get the circuit down to the 5ms range or less.

All The Best Lee…

[smw1998a]

H11d1

Hello adamhy,
Here is a link to the data sheet

http://komponenten.es.aau.dk/fileadm...Opto/H11D1.pdf

Regards Lee...