Whats going on?

[thundair]

I built this gasket for a Moto Guzzi as a test bed to see if an embedded igniter with four paired electrodes would work as a perimeter ignition.
I knew that I would have a timing problem but I am back firing through the exhaust at on/off throttle. The Lambda numbers are ~19. I took 3 of the plugs off line and I am just firing one of the embedded igniter's, and it still is backfiring at on/off throttle with a Lambda of ~16. I took the same plug wire and put it back to the original plug and all is well, with a Lambda of 14.

I will deal with the timing problem later, now I would just like to get one electrode working

I don't know anything but the people that do.

 

[Zapster]

Wouldn’t the perimeter be the worst place to try to ignite a mixture? It seems like you may have some fluid stagnation at this edge. Are you sure that you have a fresh charge at this edge when the engine runs; if not, there may be no spark problems. How can you be sure that it is a spark problem? Put a scope on the lead and see what is happening electrically compared to firing a typical sparkplug.

 

[btrueblood]

Explain a bit more please. Are you trying to drive an arc (spark) from the bent wires to the cylinder wall? If so, are you sure the wires are not shorting out when the gasket is compressed?

Finally, how are you driving the electrodes. I could imagine a coil/magneto designed for a single spark plug could get a bit overwhelmed when driving four spark gaps. Weak spark could explain misfire at WOT/max. cylinder pressure...

 

[thundair]

The gasket is ceramic with a .004" copper on each side it torqued down fine and we re torqued to make sure it was at a stop with no internal breakage. All four are firing between 8-13000 using a MAC tool clamp on
I have a 4 prong coil that is 3.75 ohms the stock one was 3.50 ohms.
I don't think there is a problem with it firing as it will go to red line running on one cylinder on one electrode.
Although it will not idle on that cylinder


I don't know anything but the people that do.

 

[thundair]

I just realized the pic I attached does not have the gaps cut yet. There is a .032 gap cut at each location. We test fired it and had a nice blue flame at all points.

I don't know anything but the people that do.

 

[btrueblood]

Ah, good.

 

[MikeHalloran]

Unless I missed a meeting, the gap between the root of each electrode and the adjacent metal surface, head or cylinder, has to be smaller than the .032 gap that you intended to fire across.



Mike Halloran
Pembroke Pines, FL, USA

 

[thundair]

Sorry about the quality but here is a pic of the first firing. The spring clamps are to hold the copper in place they are held with head compression when used.

With .040 clearance to the head we feel that any errant spark will soon carbon over and go back to the nice clean tantalum. The gasket thickness is .120" and the tantalum wire is .040" The old gasket was .047" and we took that difference from the head and cc'd the head to regain the volume and resultant compression ratio.
Although the squish has changed it has not shown to be a problem yet and will probably rear its ugly head when it is running well enough to get it back on the dyno

I don't know anything but the people that do.

 

[CCycle]

Probably at idle the residence volume left over from the exhaust stroke is a significant percentage of the charge. You are not getting enough stokeometric mixture out to the edge of the cylinder to the ignition point.

 

[btrueblood]

"I have a 4 prong coil that is 3.75 ohms the stock one was 3.50 ohms. "

Um. This still isn't clear to me, sorry for being obtuse...but, each spark gap should have its own coil (at least in my perfect whirled), is that your setup?

 

[thundair]

Right now I am using the stock coil with one electrode in the gasket, and the engine will rev to red line on the one cylinder.

It wont idle on that one and will back fire if I open and close the throttle quickly.

My Lambda readings say that I am creating a lean condition but why/how would that be by changing the spark location.



I don't know anything but the people that do.

 

[curtis74]

Changing the spark location changes where the flame front begins. In effect you have shortened the distance from where the flame starts to the piston crown. Mean Effective Pressure theoretically shouldn't change, but you may need to play with initial lead. My guess is that you will need more initial timing to counteract the fact that the flame kernel won't be starting at an area of ideal stoichiometry.

That might help your idle popping.

 

[bosscat]

Very interesting idea. What were you hoping to achieve? Power, economy, emissions? I'm not familiar with this engine design and don't know what shape the chamber is, so apologies if I make no sense, but this is my take on it:

Your lambda reading is pretty easy to explain - it's reading misfires or partial fires. As the burn is incomplete, air and fuel passes through the engine and the sensor sees the O2 giving a false lean reading. This is consistent with the popping and spitting I think.

The popping spitting/reluctance to run at low speed all read like very retarded ignition (or very lean mix). My suspicion (and this has already been mentioned above) is that the charge density at the edges of the chamber where the spark is, is not sufficient to light/propagate the flamefront easily or is much less turbulent than the rest of the chamber, which all leads towards very slow combustion (when any occurs at all) giving the illusion of retarded timing.

It occurs to me that when the designers first laid out the combustion chamber they will have planned for squish to push whatever is in the chamber towards the spark plug and the spark plug is probably fairly central. This unfortunately means that the engine has actually been designed to shift the charge away from where you are trying to light it?

You could try advancing the ignition (alot) at idle and partial throttle (why not use something like Megajolt or megasquirt n spark to give you a mappable ignition you can play with?). However I'm not sure this will help that much. What I wonder is whether it is worth running all 5 ignition points simultaneously. At idle the original plug does most of the work and then when the engine is working hard and charge densities and turbulence are much higher, the other ignition points can contribute and presumably speed up combustion and make it more complete. This might allow later timing without torque loss and possibly allow higher CR without pinking? Might also be advantageous to vary the timing of the original plug and spark ring independently, but that is getting way over the head of a bush mechanic like me......

I shall watch this with interest

Regards

Nick

 

[BrianPetersen]

I'm fairly sure that this engine is a 2-valve hemi-head, with the spark plug in the usual location between the valves but offset to one side due to lack of room dead-centre.

The perimeter of a hemi-head is usually a squish-band. If my understanding of what has been attempted is correct, it will be attempting ignition in the squish-bands.

The above explanation of what your wide-band is doing, is correct. It's a false lean indication caused by mis-fires.

 

[BrianPetersen]

One other thing ... as to why it's sorta working at higher revs but not at idle. The ignition system advances the timing as the revs go up. If it's firing at (say) 40 degrees BTDC then the piston is still a fair distance down the hole, so it's starting the fire and then squishing the fire into the main chamber after it's already lit. If it's firing at (say) 5 degrees BTDC at idle, the piston is practically all the way to the top, and it's trying to light the fire within the squish band.

Interesting idea, by the way ... I suspect the real problem is that the shape of the combustion chamber isn't matched to where you are lighting the fire.

 

[thundair]

I have a feeling I my have caused the problem by machining almost .100" of the head puts a big flat cold squish area in the quadra plug area.

I never considered running the spark plug as I am putting a 230 bar pressure sensor there.

I think you guys might have hit on something here as I now have the Quadra plug in the cold zone I may grind that out to make sure I induce some swirl in next to those electrodes.

I am working from a patent US7299785 and yes a nine year old orangutan can get a patent LOL


I don't know anything but the people that do.

 

[BrianPetersen]

Have you thought about the erosion that will happen when the spark jumps to the aluminum head or piston, if you use this system long-term?

 

[dgallup]

There are millions of patents that are unique novel original bad ideas.

 

[thundair]

dgallup ...amen brother thats why if I cant prove it out I am done with it, as it is not my first failure or hopefully my last...

Brian ....Although I am not doing this for the test the original idea is to have a circuit that is switching power supply. This is a totally independent circuit and will not see the engine components as ground. There are a lot of things to sort out yet, going from wishful dialog to reality.
I will repost when we get more done.

I don't know anything but the people that do.

 

[whitehendrix]

very slick.

i too will be watching this one..