Archives

P&G is pulse of energy, glide, pulse & glide. The glide can be longer than the pulse, sometimes 3 or four times longer.
The insight engine is three cylinder. What if the thing would run at reduced power on one cylinder. Fire a cylinder, cut the fuel injector for the next two cylinders, fire again. A simple chip could cut them off or turn them on. Want a little more power fire two, off one. Will the thing shake to pieces, or could it work???:rolleyes: ;) Willie what do you think?

About a year ago, I conducted a "Cylinder Deactivation System" by which I ran my Integra on 3-cylinders using a toggle switch in the cabin to cut power to one injector on the fly. After much research, discussion, and testing, it turns out that the timing of no combustion in one cylinder was causing moderate-to-heavy vibration and rocking of the engine unless running over 4K RPM. So, I ran it 50 miles on 3 cylinders, keeping it in 3rd gear or else the shimmy was horrendous. Furthermore, the cylinder that was pumping air only was fooling the oxygen sensor into thinking that a lack of fuel was being supplied and it ran rich. I got a whopping 21 mpg on that experiment and deemed it too violent for engine longevity.

The odd number of cylinders might be an advantage for engine balancing, but the O2 sensor issue should be addressed. It's fairly simple to test -- simply disconnect an injector from the harness clip to see if it runs (at all or smoothly). You will get a CEL of "cylinder #x misfire". Great idea though. I thought if HEMIs could do it, well I could too! Turns out theres more to it...

RH77

I have a relay that cuts all injectors off, for the FAS. It would be a piece of cake to have a cylinder cut off that was more selective. The Honda system has dynamic vibration control using the IMA motor as a damper. It would be interesting to see how well it would work with a dead cylinder or 2. May be a good thing to play with at Up Your Volts 3 in 2 weeks?
There may be a way to cut the signal from the O2 sensor during the cylinder cut so that the mixture stays where it should be?
Wish I had more timeto play.

*stares into crystal ball*

I see many vibration problems in your future...

The inline three-cylinder is one of the worst engines as far as vibration and now you want to change the number of active cylinders on the fly? Good luck :p

I did say crazy idea;) But MIMA was a crazy Idea and it worked, so sometimes one has to try just to see what happens.

The Honda system has dynamic vibration control using the IMA motor as a damper.

That's interesting, and really neat -- it could be the saving grace in this test.

There may be a way to cut the signal from the O2 sensor during the cylinder cut so that the mixture stays where it should be?
As you know, it's tricky to fool a Honda emissions system. I'm guessing that disabling it may throw it into "open loop", which we don't want (if hybrids do that --honestly, I'm not very familiar with the innner-workings, but I'm learning :rolleyes: ). The solutions would be to hijack the sensors, and create your own voltage to emulate a normally operating system; so, if you have 2 sensors, the first sensor before the cat would need to show less oxygen (so perhaps something that lets the original signal be squelched (a resistor?), so the peaks and valleys remain, but at a lower point on the chart, perhaps) The waves of voltage to show "normal operation" in the first sensor, and the second should have a rather flat line to show that the cat is working (easier, just splice-in a voltage that it likes). I've heard that a slight variance between these 2 sensors, in Hondas, will freak out the emissions system and go into intermittent open loop (I had a problem on a '99 TL quite similarly).

It's definitely trial and error, if you have the time. It would probably be one of the first of its kind in an Insight. :)

RH77

Watching things with an OBD scanner would tell us pretty quickly if the solution was working. One could even use a cap to store the good values, and disconnect the sensor input when the cylinders were off, like a sample and hold circuit?

Any ECU would have to be clued in that this is happening, for
starters, or the charge of plain old air will totally screw the
closed-loop aspect.
.
But assuming that could be dealt with, how about injector-cutting
TWO out of FOUR cylinders instead? That will still make the torque
pulses applied to the crank more sporadic, but may run more evenly
overall with less external shaking in the mounts.
.
The Honda VTEC hack to leave valves completely closed seems like
it would better support this sort of thing than a more normal
valvetrain, but for best adaptation would probably have to be
switchable per-cylinder. Electrically-actuated valves will
doubtless make this so much easier...
.
_H*

Another solution a bit more elaborate, but definately in the relm of feasability.
We connect a Microcontroller to the fuel injector pulse as an interrupt line.
Two analog inputs of the micro are watching the two O2 sensors voltage, and will store a data table of the O2 changes with normal firing. Basically does a frame grab. When the dead cylinders try to fire, the processor plays back a real time replica of both O2 sensor pulses to the ECM, via a high speed PWM output that with a low pass filter, duplicates that good signal.

That way the computer in the car would not know that things were different?

It could work. The micro only cost $10 that could do a nice job with this.
Being an odd cylinder count ,(I am talking about my 3 cylinder Insight, not the 4 cyl civic or prius) I wonder if one of the cylinders woulden't find a reasonable balance compared to the two off cylinders. Only way to know is to try it.:cool: ;)

Hello everyone,

This is my first post here, so here it goes,

Mike,

If you are trying to deactivate a cylinder you still have the pumping losses when the cylinder is compressing the air, this is wasted energy, the other cylinder would have to work harder to compensate for this, a better way would be to FAS, the engine is off, and there are no mechanical losses thru the engine.

The biggest loss of energy in an engine is heat, thru the cooling system, and the exaust system, FAS eliminates the exaust completely, this is the biggest gain you can have. IMO

the way I see it....

reducing the number of active cylanders..... While tricky... has been done in other vehicles... during times when you still need power but less of it..... kind of like an alternative way to what lean burn does ... less fuel for less power but when you don't need the extra power it is ok and more fuel Efficient .... yeagh you still have pumping losses.... but 2 cylander instead of 3 should be a little under 30% less fuel at the same fuel ratio... so you could force a lean burn like mode without the trickier stuff needed to balance Lean burn ratios and temps.... you could also combine it with lean burn .... you would get even less power per engine cycle but... sometimes you need less power and just like Lean burn it could lead to better fuel economy.... The other benefit is that if it works... you could even start out in 2 cylander mode when the Engine is Cold and you would never get normal Lean burn operation.

I think it will be tricky to say the least... but there are benefits to be had... One question might be how much of a benefit ... since the variation of rich / normal / Lean burning modes already in an Insight gives you some of these same kind of benefits as deactivating cylanders.... I don't know any way to really know how much additional benefit it will give without testing it, by doing it ....

my 2 bits.

ian

Calpod, IamIan
I said up front that it was a crazy idea, just let my fingers type while my brain was dreaming.:rolleyes: ;)