Archives

So everyone knows you can do some crazy things in an Insight under "ideal" conditions such as those presented in the Insight marathon. But possibly, we could extend that envelope to accomodate a wider range of driving conditions. I'll explain.

Recently, I hit some backroads, driving the poor little Insight pretty hard on a nice day. I still manage to keep it well over 50mpg, however. :)

The most frustrating thing was the huge gaps in the powerband between gears. A shift at full tilt in first lets the RPMs drop to 3000! I wouldn't have thought this would be a problem, because the electric supplies a good bit of the torque, but from listening to the sound of the engine, I understood the problem.

The valves aren't all opening.

That's right, below 3500-4000 RPM, only one intake valve per cylinder is opening. It's part of the genius of the VTEC-E system. At low RPMs, one intake valve opens, inducing a large amount of swirling for the so called "stratified charge" effect. This allows the mixture to burn extremely lean while still producing enough power to cruise well. At higher RPMs and loads, the other intake valve will also open to allow a greater volume of air to enter the chamber, dramatically increasing power production.

However, it is well known that 4 valve engines have a much higher volumetric efficiency than 2 valve engines. A 4 valve engine with only one intake valve opening is *worse* in terms of VE than a 2 valve engine!


Let's discuss P&G specifically. While pulsing, we want to keep the RPMs low, load *fairly* high. This maximizes the efficiency of the engine. The problem is, in the RPM band we typically "pulse" the Insight, the VTEC system is *not* engaged. The very same system that allows insane lean-burn steady state coasting hampers VE during lower RPM pulsing.

Solution? The VTEC solenoid is activated by +12v. Use a momentary switch on a dual throw 12v relay to activate VTEC manually during pulses. This should allow high VE even during lower RPM pulses.

I may be way off base here. It could be that at those low engine speeds, having the second valve open gives no measurable benefit. Perhaps if it did, Honda would have used some more sensitive load-based VTEC activation rather than the stricter RPM activation.

I do know that under aggressive acceleration, the Insight is a dog until vtec engages - and the effect is very noticable, which causes me to think there is more torque and VE available lower. The ECU should compensate for the increased airflow at the lower RPM so long as it stays in closed loop (throttle not totally smashed during pulses).

The potential tangible benefits I see, if this works, is the ability to efficiently pulse over a wider range of gears and speeds.

If some of the real Insight P&G experts could toss out RPM bands, gears, and speeds they find most efficient, I'll get started with some testing.

You can try it and find out. Typically a 4-valve motor will have less TQ at low RPM than a similar 2-valve engine. There is a point though were the extra 2 valves really help. Intake and exhaust system tuning also have an effect.

It might have been wise for Honda to also have split the intake manifold and have a runner for each intake valve. A similar effect could have been included on the exhaust.

Let us know how that goes. Sounds like a neat project. Remember that more TQ will also have an effect of more fuel being injected. (more air flow = more fuel)

I believe one of the reasons that lean burn works on the Insight's engine is because of the massive swirl effect that happens due to only using one valve. If you take that away you won't have as even of a fuel mixture in the chamber and you may not ignite all of the vaporized gas. However, there may be a fix if you are willing to do a bit of work, and it sounds like you are. A guy from another forum I'm on likes to bias the intake ports on the 4 valve heads he works on. He adds material to one port, and removes some from the other. This increases swirl inside the chamber and allows a higher compression ratio. However, for your endeavour it may work enough to keep lean burn working effeciently. Of course, you'd still have to test it out to be sure...

I believe one of the reasons that lean burn works on the Insight's engine is because of the massive swirl effect that happens due to only using one valve. If you take that away you won't have as even of a fuel mixture in the chamber and you may not ignite all of the vaporized gas. However, there may be a fix if you are willing to do a bit of work, and it sounds like you are. A guy from another forum I'm on likes to bias the intake ports on the 4 valve heads he works on. He adds material to one port, and removes some from the other. This increases swirl inside the chamber and allows a higher compression ratio. However, for your endeavour it may work enough to keep lean burn working effeciently. Of course, you'd still have to test it out to be sure...

A lot of newer Honda heads are already biased this way - difference of swirl port (offset) vs tumble ports for higher flow rates. I know the D16Y8 head is this way (96-00 Civic EX). I haven't had a chance to crack off the intake manifold and look, but I'd be surprised if the Insight head isn't already ported that way.

Of course, cleaning out the cast marks and such (basic port work) could increase efficiency.

Hi Shifty:

___On my own Insight, P&G was not discovered in the modern sense of the word so lean burn is where it lived and died. With today’s P&G, opening up the Throttle that far will cause a lot of fuel thrown away due to high RPM friction and thus I would not go there if I were you. Watch the iFCD and see what happens at 2K, 2.5K, 3K and 3.5 on up. You start consuming fuel like a drunken fool and the accel is not that much more than a std. 2 + K pulse. Lean burn is thrown away during the pulse but at least it is a manageable 40 + mpg vs. what you are going to see at 3K + for V-TEC to come online.

___Good Luck

___Wayne

Well, the idea was to use VTEC to keep the valves open and perhaps gain a bit more efficiency at low RPMs (<3k) where VTEC is typically not available. I haven't get started with any tests yet (spent all week moving!) but I think the change in acceleration I have been feeling around 4k is simply the background charge being disabled, which is substantial at times. If this is the case, then the VTEC crossover is seamless, which would indicate that there is no benefit to be had by enabling it any sooner.

I've tuned quite a few "performance" based VTEC engines (B16, B18, D16Z6) and the normal method is to tune both cam profiles independently and then set the switchover at the point where there torque curves cross, producing the most power under the curve, and promoting a seamless crossover. This is undoubtedly what Honda intended when tuning the Insight, but it's usually pretty far off on other vehicles I've tuned.

More work to come.

I too think they would have tuned it to achieve maxium torque across the rpm range. However, if other engines you have played with have had gains from it your testing may prove quite interesting. I look forward to seeing your results.