ideas on rev limiter algorithm?

[gsz]

I need a methodology to control max engine speed while testing small two-cycle engines on dyno. what i'm thinking is to kill the spark intermittently at a specified rpm (like a stutter box).

questions are:
1) with a small two-cycle, carbureted engine is there a danger with turning the ignition on and off so as to maintain a max speed (back fires etc)

2) any recommendations on the control strategy? ie, if we turn ign off when rpm > x, and turn back on if rpm < y, what should the difference be between x and y (100 rpm delta if max rpm is 11,0000?)

thanks for any input.

 

[Lou Scannon]

If you have all the capability to do what you have outlined, there should be nothing stopping you from writing a rev limiting program with calibratable parameters, and experimenting (cautiously, at first), to find an effective calibration.
I'm surprised that your dyno control does not have rev limiting capability. If it doesn't, it might be something you could easily add.
Regarding backfires, they're always possible when spark is cut (especially intermittently) while fueling continues. If you're concerned about the hazard, I'd recommend installing flame arrestors in the intake and exhaust at the locations where you do not want flame to propogate past. Blow out panels in the intake & exhaust plumbing on the extinguished side of the flame arrestors might be a good idea also, to let pressure escape safely.

 

[reidh]

What type of dyno are you running? It may be easier to increase your dyno load so that the engine doesn't over-rev, as long as that doesn't skew the data you are trying to collect.

Reidh

 

[reidh]

hemi,

You beat me by 1 minute. I have to learn to type faster.

Reidh

 

[gsz]

thanks for the feedback. the system is labview based (national instruments USB hardware connected to a laptop) and is completely custom so implementation of virtually ANYthing is possible.

although i started out with closed loop speed control (rpm feedback was used to vary the water supply to the brake), i've fallen back to manual control like in the old days due to my inability to maintain good automatic speed control. i'm running a small 70cc 2-cycle racing engine which when it "gets on the pipe" sends the engine to redline in an instant, and the system just can't shove water into the dyno quick enough to maintain rpm.

the technique i'm using now is to rev the motor up to near redline without load, then slowly crank up the water while simultaneously increasing throttle as needed to keep the revs fairly constant. when the throttle is at wot, i adjust the water load slightly to vary the rpm.

having the rev limiter not only provides a safaty from free-revving if something happens (like coupling breaking), but conceivably i could simply hold the throttle wide open since the rev limiter would hold it "stable", then apply load until the revs drop below the rev limit. i would then begin my sweep downward.

what ya think?

 

[MikeHalloran]

I think you might foul or glaze the plug if you kept it pushed against the rev- limiter.



Mike Halloran
Pembroke Pines, FL, USA

 

[bradrs]

I have made several 2 stroke rev limiters in the past. Some were software based internal to the ignition, and others were hardware based, attached to the OE ignition.

Attaching one to the OE ignition can definitely result in a rough running, because of how they are attached. I have heard of backfires and broken reeds from them. But mostly, I have heard of lots of success with doing this, even with something that works by basically grounding the engine kill switch above a set RPM.

How are you planning on implementing this limiter? Do you actually have control of the ignition, or are you simply planning on externally attaching it to an existing ignition?

 

[dgallup]

Most vehicle rev limiters use a "softer" setup which might be overkill for your needs. Multi cylinder engines will cut individual cylinders first. Single cylinder usually cut every 3rd or 4th spark, then every other spark. This is to make the transition less abrupt. May be more trouble than it's worth for you.

 

[gsz]

i do NOT have "direct" control of the ignition, which is a PVL brand CDI. i plan on simply replicating the kill circuit by grounding the CDI box via a solid-state relay. when the daq system senses rpm above the max threshold, the relay closes and effectively kills the spark. when rpm falls below the min threshold the relay is opened and spark is returned. if i hold the throttle open then the engine should cycle between these limits at a frequency and amplitude dictated by the threshold rpm delta and the engine's characteristics (inertia and power output).

one of my concerns is actually HOW the ignition is returned... the cdi is grounded until some asynchronous point at which the grounded kill is removed. ...will the cdi provide a spark output at that instant (most certainly the wrong timing), or is the cdi smart enough to wait for the crank trigger before releasing the spark?

i also appreciate the thought on fouling. will keep it in mind. certainly the more elegant solution would be kill fuel supply, but without efi it aint gonna happen. :>[

 

[bradrs]

I didn't deal with PVL much, but I do believe they have analog and digital versions, with the digital being offered in a programmable version. Is that an option for you?

If not, your idea of a relay to trigger the kill switch should work. One 2 stroke rev limiter on the market, the Dyna DRL300CDI, does basically that. I've heard people say that particular one doesn't work with PVL ignitions, but the idea is sound.

The main issue then, is whether or not you are using a pVL that has a separate lead to power the CDI firing, and a seperate wire for the trigger. Some of the smaller 2 stroke engine ignitions use a single wire for both function. These could have a real problem with the firing at the wrong time like you were worried about. If it has a separate wire for trigger and charging, it SHOULD be able to avoid firing at the wrong time.

If your existing setup is a single wire for charging and triggering, I would convert to a different stator, if performance is more important than low cost.

 

[thundair]

Is there a way to put a capacitor that is switchable in-line with the hall effect from the crankshaft trigger?
You could then add a resistor to the out put to effectively retard the spark.
I have used this to be able to start a two stroke that was set on full advance.

Cheers





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