Hydrogen engine ignition delay

I am looking for creative ways to delay the ignition on a single cylinder engine, by using an electronic circuit.

Here is what and why:
Having researched and worked with Hydrogen as a combustion enhancer (a long discussion I will leave out,) I am attempting to convert an existing Honda GX270 (9HP) motor to run on Hydrogen. The mechanics and fuel delivery are simple. What I am having difficulty with is proper ignition.

Hydrogen is such a fast burning fuel that it requires ignition timing to happen almost at top dead center and a Honda engine such as this uses a fixed magneto to make a spark (no points or anything.) Without some pretty advanced machinery, changing the timing is not possible (factory set at about 28 degrees before top dead center.)

I have already created a simple circuit that uses a hall effect sensor, a transistor driver, a transistor, a standard auto ignition coil, and a battery, to detect the existing magnet on the flywheel and create a spark. (It just turns the transistor on as the magnet goes by and off when it leaves. The width of the magnet is about 2 inches, which becomes the ignition "dwell" and the spark happens when the coil field colapses.)

What I would like to do, AS SIMPLY AS POSSIBLE, is to add something between the detection and the driver, to delay turning the transistor on by a preset and adjustable amount. Since this engine will be a constant RPM, I am not concerned with varying the amount with differing RPMs.

I have tried using a PIC microcontroller, but the programming got over my head. If I were to make something production, I would probably hire someone to do it that way, but I am now just doing this to experience the process first hand. In the mean time, I thought of a using a 555 timer, but am not quite sure how to do it. I'm really NOT much of an electronics person - I have tended to learn what is necessary to get something accomplished... So, inevitably, I blow a lot of parts up.

Anyone who has any clever ideas, or can guide me on setting up a timer to do this, I'll appreciate any guidance. It may be that this is just too hard to do without using a microcontroller or without mechanical fabrication. If that's the case, at least I will have tried myself.

Thanks,

Chris Helvey

Actually, it sounds like you want to be able to vary when the transistor turns OFF, as from what you've said that is when the spark occurs.

Correct?

This will be kind of dicey to do when the engine is first starting, because it'll need a relatively long delay, and until you get past idle it'll be tough to get the average RPM. As the engine gets up to speed and you can get an idea of what the RPM is, the delay won't have to vary much.

Is it an electric start engine? If so, does it have a ring gear on the flywheel? That would be a perfect target for a Hall-effect sensor to pick up what the actual speed of the engine was.

After that, it's just a matter of counting the teeth after the the other Hall-effect sensor is picked up, and letting go with the spark after enough teeth have passed. You could count the teeth with something simple, like a 4518 (BCD) or 4520 (BIN) dual up-counters. Feed the relevant outputs to a NAND gate (the 4068 is an 8-input NAND); when the terminal count is reached, the transistor is turned off and the plug fires.

Pull start. :-(

You kind of lost me on the counter.
Would I use two Hall effect sensors then?

I have to wonder if there is a cheap automotive aftermarket product that could do this.

Thanks,

Chris

There are two hobby sites for CDI ignitions that might help. "Bigboat" is involved in both and a key name to look up. The New Opensource CDI site has all the links you need.

I know you said that you didn't want to do the programming. On those sites, you will find people who may do it for you, or you will find links to commercial units that do the same thing (e.g., CH Ignitions, Wyoming, USA). The main reason you are likely to find helpful information there, as opposed to automotive sites, is that you are dealing with a single cylinder engine using a Hall sensor, which is standard fare for models.

As a far out suggestion, are you sure you need a spark at all? Will a glow plug work. You could start on methanol or ethanol, then switch to hydrogen.

John

HI,

An easier method maybe, mechanically.

In go-karting, (dealing with Honda engines), advancing the timing is almost a must for more power. But with Honda GX engines the timing gear is next to impossible for the average karter, to remove or advance on a few degrees. (As stated by yourself)
So another method is used, by filing a slight amount off the flywheel key, thus moving the flywheel a few degrees on the shaft, changing the timing. The degrees of advance/retard can be worked out by the amount you remove from the key.The flywheel is then lapped to the shaft, when done correct, and torqued down, the flywheel wont move.

This should work in opposite for you,i.e. retarding the timing.

There are sites that detail this. I think the keys can be even bought ready filed for precise advance/retard.

Might by easier.

Regards NM

He's at 28° BTDC and wants to go to 0°. That's more than can be done with the key. Maybe a machine shop could re-broach the timing gear/pulley. However, if he goes with a Hall sensor, adjusting the timing is no problem. John

True,
I overlooked the 28degrees to 0.
A key will give about 10 maximum.

Apologises for false hope.

NM

ChrisHelvey said:

What I would like to do, AS SIMPLY AS POSSIBLE, is to add something between the detection and the driver, to delay turning the transistor on by a preset and adjustable amount. Since this engine will be a constant RPM, I am not concerned with varying the amount with differing RPMs.

Click to expand...

What about a simple monostable mutlivibrator? Obviously, one with good stability would be required. Maybe something like an SN74121?

Why don't you just move your Hall-effect sensor over so that it will cause the spark to fire close to TDC? You had to add it to begin with - just move the thing over.

If your flywheel fits onto the crankshaft with a taper then moving the flywheel is the simplest method, you don't need the key at all it is only there to locate the flywheel in the correct position, as NM2008 said you may need to lap the flywheel to the shaft so it won't move. Many years ago English Villiers engines had no key, the flywheel was simply held by the taper. Paint some timing marks on the flywheel or crankcase at TDC and 28deg. before TDC and you'll know how much to rotate the flywheel.

How did this go ?
Did you find a suitable solution ?