I have been kicking around an idea of deactivating some cylinders on my tired '91 sedan in order to get even better gas mileage, but the lack of power could be bad in some instances for sure. So, I was just brainstorming a much better idea with a friend: Why not make a VTEC actuated change from 4 cylinders down to 2? This way you could turn your cylinders on and off as you please, no problem. (Yeah this sort of thing has been experimented with at honda.) What I want to do is go for a *budget* cylinders-on-demand system in the 4th gen.

You might say, "How the F is that gonna work?" Well, here is the run down:

• Buy a B16 SiRI dohc vtec motor.
• Swap motor into 4th gen.
• Get custom grind on cam. *
• Wire in a switch that deactivates the VTEC for the whole motor and turns off the fuel injectors to two cylinders.

*The asterisk here is because I haven't explained the way it will work. vtec works in a weird way with the 'rocker' for the high lift lobe essentially flapping in the breeze until oil pressure is let when the solenoid is activated (everybody already knows, I know). To get this to work I would grind the vtec lobes on all cylinders until they are identical in lift/duration to the stock lobes. This would give the same wimpy power whether or not Vtec is on or not. Then on the cylinders I would like to be deactivated with a switch, I would regrind the normal lobes down to just a circular lobe with no lift/duration WHATSOEVER. Then attach the vtec activation wire to a switch along with the two fuel injector wires from the cylinders you want to deactivate. This would let you flip the cylinders on and off.



I want to check to see if anyone sees a major flaw in this plan before I actually start, and maybe get advice on where to get the grinds done.

 

 

i'm not sure what your goal is too, I regularly get 35-40+mpg in my 177whp ls vtec in my eg hatch with an ls trans if i'm nice to it and drive it right. this is without getting too crazy too and i have performance 16' tires

 

You will need a programmable ECM, your MAP sensor is going to read the wrong vacuum and deliver too much fuel to the 2 running cylinders, your O2 sensor is going to show too much oxygen and deliver more fuel, your ECM will show misfires and go into Limp Mode. Even if you could MOD everything, I would rotate which cylinders are shut off to give even wear and tear. Its going to be way too much work to save a little gas even on smooth level highways.

 

Thanks for the comments.

PirateMcFred, there would be two differences in the cam lobes, yeah the vtec lobe would be ground down to stock level, but that isn't the thing that shuts off air to the cylinder. I was suggesting grinding the vtec lobe down to the wimpy stock mode just to have the good gas mileage of the stock grind. The real difference is that the two outer 'low rpm' lobes would be ground down to a circle/no lift/no duration. This is what would shut off the valves when vtec is not triggered (the middle lobe would still have a grind but it would just be flapping, doing nothing, the outer two lobes would have zero lift, doing nothing).

About the air pressure reading, yes I didn't think of that. The fuel levels at various pressures would have to be changed when in 2 cylinder mode (vtec off). I am unsure of how the ecu handles vtec but I assume that when it is activated it also triggers a new fuel map for the new cam grind. This would hopefully allow for a different control of the fuel levels on only 2 cylinders (vtec off) and working on 4 (with vtec on).

I am not sure about the O2 sensor, there would be no clean air pumped through the 'off' cylinders so there shouldn't be any extra oxygen in there (assuming you have a good fuel map for the 2 cylinder mode).

Aside:
Your suggestion of rotating which cylinders is interesting. Is there a way to turn on only vtec for one cam and not the other? This would allow you to shut off 2 cylinders with one cam, and shut off the other two cylinders with the other cam (but there would be a big difference, only either intake or exhaust valves would be shut off at any given time. I didn't think this was a good idea because the engine would be either pushing off the air, then closing off or sucking in and closing off with no full closure of the cylinder. I am betting that a full seal of all valves will be more efficient and the only way I can see to do that is with dohc. This is why I figured a dohc would be better than a sohc. The sohc you could only shut off the intake side using my theoretical method since it only has vtec on the intake valves.) I was considering merely doing this with a D15b sohc, but was leery of the process that would be happneing. The exhaust stroke would be pushing out air, but then the cylinder being shut off for the intake stroke would tend to draw a vacuum in the 'off' cylinder and then try to pull air past the piston rings and that would be bad. That type of problem would only happen in a situation when only intakes are shut off, or if exhausts are shut off - i.e. sohc or with a separated activation of each cam on a dohc (rotating cylinders setup).

*edit:* To be clear, I don't expect to be able to rotate cylinders with a dohc while still having a full seal on the cylinders, I don't see an automatic way to have that happen. I could have two sets of cams and an extra ecu, then switch them out every 10k miles or something though, that could work. It would probably be possible to have it rotate cylinders by way of switch though if I only seal one side of the cylinder (using a dohc setup with cams that trigger separate from each other).

My goal would be 60mpg with this plus a few minor aero mods. Why? Well, it would be fun. Also, having hybrid gas mileage without buying a hybrid would be pretty dope.

 

plus aero is a big gas hog that people never think about. get way better mpg pushing an egg through the wind then trying to haul a box so to say.

 

with they money you spend on trying this. you could save alot more money by not doing this. if you so worried about mpg or gas price

 

It isn't about the money. Do you mod your car for the money in it? It is about fun, interest, etc.

 

I thought VTEC didnt work until you got up around 3000 RPM???

 

Thanks Fred, that is great advice. I wasn't really aware of being able to disable the activation of vtec cylinder by cylinder. And hell yes on the 0w20 oil, sounds like high mileage city right there.

I have been reading a little bit about a D15 (z1 or something) that is for 96 and up (maybe 95 and up) honda's that has a 3 stage vtec. This motor might make for even more interesting possibilities. Of course it is only sohc, but I might be willing to give the 'closed intakes only' setup a shot if the motor is cheap enough.

Aside: For those that don't know, there are people out there that can get 70-100mpgs with areo mods on their cars and with other subtle changes, along with changes in driving style (look up aero civic). The most radical of them being the pulse and glide method, where you accelerate up to say 75mph keeping it close to your most effcient engine rpm, then completely kill the engine and just glide down to like 55mph. Rinse and repeat until you get to your destination. The combination of this crazy driving and the mods gets really good mpg. I am thinking, why shut off the engine (which can be a little dangerous, no power brakes and stuff) when you could just disable 2 of the cylinders for the whole cruising time?


*edit* how about the cost of the grinding?

 

just unplug 2 injectors and if it gets too weak just shoot 100 shot nitrous end of story.

 

I'm not sure how diverse the tuning options are for vtec engagement on tunable ecu's, but don't forget vtec is also actuated (or deactivated) by engine temp and throttle position.

There are numerous car manufacturers out there that offer what you are talking about already, where the engine drops to half the cylinders or less during freeway cruising (my bosses Tahoe does this). It would probably behoove you to study them and see if you can figure out exactly how they are implementing it. I'm not sure you need to go as far as the whole vtec lobe concept, I think it just introduces more variables and complexities you don't need to deal with. I would think a good standalone EMS or some creativity with some kind of piggyback device and a tunable ecu you could just drop the number of cylinders under certain conditions (throttle position primarily I would assume).

In fact, I would think with a standalone EMS or tunable ecu with individual cylinder trims you could just put a fuel value of zero in the areas of the table you want to drop cylinders. At that point you could even increment it so that you drop only one cylinder under certain loads and then two or possibly even three under lighter loads. The only flaw is that I know usually values from neighboring cells are also read which could get a little interesting. It shouldn't hurt the spark plugs to keep firing and then you're not messing with the airflow properties either, it will just flow in and out like normal. Am I totally off base here?

 

If the air flows through it will screw with the o2 sensor readings. Maybe the dependence on the o2 sensor could be programmed out but I bet you lose a lot by pumping all that air through.

 

Another thing to take into consideration would be that once you block off the intake air. The piston is still moving and it will still require air to fill it. Where is the cylinder going to take the air from? Crank case, or probably end up sucking it past all 4 valves. Sucking in exhaust gasses, oil and air. This would probably also put more load/strain on the engine. In turn probably lowering your MPG in the long run.

I think if you are going to be successful at making this happen it is goign to be in the software/tuning of the ECU. Along with the placement of your O2 sensor. It's easy enough to move an O2 sensor so that it will always be reading from a cylinder that is in operation. And to disable the second O2 in an OBD1 system is easy enough as well.

It's all in the tune!

 

Ghost, yes that would be a problem if I used a sohc with the intakes continuously closed, and the exhaust valves still operating normally. It would try to suck air in on the suck stroke (but wouldn't get anything), then try to compress nothing, then try to bang nothing, BUT then it would expell whatever air was in there on the exhaust stroke (or maybe not even that, it might suck in exhaust gases due to the vacuum in the cylinder, which could be bad, but also, it might not matter). But you are right, that setup would put more vacuum in the cylinder and try to pull in oil/crankcase air etc.

The dohc vtec setup would block off all valves and eventually would just have some air in there getting compressed and decompressed, etc. But there wouldn't be a process that would create a vacuum, which is good! The dohc vtec is just more expensive to go out and buy so it is worth thinking about the sohc at least. Not to mention, if the sohc was viable and did work, it would be easy for others to copy it because they are actually in north american cars. A dohc setup is a lot harder to come by, especially on the cheap.

 

OK I see what your saying about closing off the valves.

Now are you talking about running this engine in sort of like a reverse VTEC? Instead of having VTEC engage later in the RPM and add lift to the valves. Your gong to have VTEC engaged at the start and then disengage at cruising speed to close them on certain cylinders?

You can have VTEC engage as long as there is enough oil pressure to slide the locking pin over. This would enable you to run your engine on the VTEC lobes. Then when you want to shut down two cylinders. You have the primary lobes of those cylinders ground down. That way when the VTEC rockers disengage you will only be running off the other two cylinders with the primary lobes still intact..... Does that even make sense????

 

Forget about the valve train, let the engine suck in air with no fuel and exhaust the unburned air, just move the O2 sensors to all 4 header pipes and turn off the ones actually not in use. Now you have a Cat Convertor issue of too much O2 to operate properly, now this also may be an easy issue to fix, no Cat or vent off the unburnt exhaust to atmosphere. I am thinking no Cat, 2 cylinders alone may not create enough heat to LIGHT OFF the Cat since 2 cool streams of air are coming in between hot streams. Would this qualify as a ULEV???

There is proof of the theory that a sealed cylinder does go further into a vacuum (lower than atmosphere) if the intake valves do not operate, this is known to suck in oil from blowby, piston rings need compression from a working cylinder to properly seal between the pistons and cylinder walls. With valve overlap, the exhaust valve starts to open before BDC so there is a time when exhaust gases would be sucked in due to this added vacuum condition creating carbon build up over time, no heat in the cylinder to burn it off and a chance of an exhaust valve getting enough build up to prevent a complete seal of the valve and seat. This is one reason why you would need to rotate which two cylinders will work and which will rest, at least on a daily basis I would think.

If you are worried about emissions I would try a two working header pipe to one Cat (double cat system), maybe two hot cylinders might LIGHT OFF the Cat??? I would think it would have to be closer to the head and not five feet down the exhaust section.

 

Ghost, there are lots of ways you could have the vtec engage. I was just planning on a switch, sort of as a proof that it will work. Later of course you could try to make it automatic based on throttle position, rpm, stuff like that. The details of the electronics of it all seem pretty easy to me but maybe I am being naive. I was just planning on a switch that disconnects the power to the injectors, shutting them off. And the same switch would turn vtec off. Single throw triple pole is how they term it I think. I would think that this would work fine as long as there are no air flow/pressure/knock sensors that screw stuff up. The details would have to be worked out but if anything some of them could go into a sort of manual mode on 2 cylinders, with a dial or something haha.

Duane I think doing exhaust work is so much crappier than doing engine work. The method that I am leaning toward would merely change out the cam, maybe take out the vtec pins in the cylinders that are always on, then doing some wiring - that's it (probably not though, there might have to be some programming). Not to mention, the system will run the same as normal when everything is turned on and there shouldn't be any weird programming issues there, just straight forward I believe, which would save a headache. There might be headaches if you start turning sensors off or moving them around. I am not really sure though. But, the main reason I don't want to just turn fuel off and let it pump air is because that isn't how the big guns do it (gm, honda, etc, the ones that have this already working). They seal off the cylinders completely and there probably is a efficiency reason for why they do it that way.

 

You are correct, this is for steady highway driving only, need to pass, need four cylinders, need to get off the highway, need everything stock working again.

This is why BMW went to electronic solenoid operated valves, easy to turn them off and on. Remember though, these newer systems are much lower compression now, they have learned how to burn a very lean AFR these days, that is not your engine.

Good luck brother, I will stay out of this one but I will continue to watch from the side lines, my idea was to put two motorcycles in the back minus front forks and such, how does that mess with your mind, actually five bikes in the back of a van with five drive wheels to the ground but that was not for good gas milage. Like you, I can put any one of the five in "N" or kill any one and go down to two for highway usage very similar to your idea, so I feel you.

EDIT I also have an idea for the everyday gas burner, to coast up to a stop, pop it in "N", the engine dies, light turns green, push the gas just a touch, engine starts, pop it in drive (starter automatically disengages) and go, pretty much like an electric car. Please understand, in Tokyo, you might sit through 4 lights before you get to go, all day long.

 

Haha, 5 1000cc bikes (or bigger if you want). Do it.

 

why not run the motor normal when using all 4 then when you shut off 2 cyl have the 2 that are still running be in vtec so you still get some drivable power?

 

That is a possibility too but I am assuming that the vtec lobes are less efficient. Maybe I am wrong on that. There is also a drivability issue there, what if you are at 2000rpm but still want to be on only 2 cylinders, do you want the vtec lobe then?

But maybe you are right, and it could be made to work easily if I leave the vtec lobe there on the normal cylinders. If I want to try the non vtec lobes I can just disable the pins (or whatever you do to disable the actuation of vtec on those cylinders).

 

lower RPM, VTEC cam would create less power than the non-VTEC cam.


I'm at work so I skimmed over the other posts, so forgive me if these points have been made.

I have been rolling this idea around in my head for a few years.
all though how to combat the load that the two dead cylinders requiring to compress the air on their normal compression cycle is the difficult part.
the easiest would be to have your special regrind cam, have the VTEC lobes have the exhaust lobes open at all times, and the intake lobe remain the same.
this will pull some fresh air and even more exhaust air into the cylinder on the intake stroke, but on compression it will force it out the exhaust.
The worry, that the hot (potentially still burning) gasses from the exhaust of the running cylinder ignites the intake charge before the intake valve closes. Even with the cylinders injector off, it could still ignite the fuel charge to another cylinder... resulting in a backfire, or a serious explosion in the intake. (was thinking of adding reed valves to the dead cylinders when I last pushed this around in my head)
this will also keep your MAP sensor and O2 working correctly for the other two cylinders.

BUT the additional load on the running cylinders will require some additional monitoring to be sure you're keeping them happy. Individual EGTs would be required while you're tuning. Individual widebands would be good too.

some solutions, current OBD1 tuning programs have the ability to switch to secondary maps based on what ever criteria you set... you can also turn on an output based on the same criteria... say low RPM, TPS below a certain point, etc.
this creates your "low power" state trigger.
this also makes it so the ECU can control when you enter and exit this state. Less potential problems that way vs a manual switch.
You can also use a GSR ECU (P72 or add IAB hardware to another ECU) to have the additional higher power output, this can also work at lower RPM like what you'd need.


IMO this is WAY too much work, yes proving a point is great, keep in mind this sticks in my head as well... but honda made VTEC-E to do almost what you're talking about... http://www.b18c5eg.com/VTEC/sohcvtece.html
[ D15Z1 = OBD1(VX) - D16Y5 = OBD2a(HX) ]
With the stock ECU... I was seeing 50+ mpg highway cruising with my VX when I tried to keep my foot away from the firewall.

OR you can try this idea with the 3-stage VTEC concept... http://www.b18c5eg.com/VTEC/3stagevtec.html of the JDM D15B variant (there are at least three variants of the same engine code)

 

I still wonder as well if you'll have vacuum issues in the dead cylinders with both sets of valves closed. I wonder if you left the intake valves operating but kept the exhaust valves closed. You would have reversion during the compression and exhaust stroke, but it is clean air so it shouldn't matter. That should solve the issue of clean air travelling down the exhaust manifold. I hadn't thought about the cat before, but it is a good point; you will likely need a small tube header to keep the velocity and heat up. In this scenario you might consider shutting off the spark plugs in the dead cylinders, although with no fuel and no compression I don't think it would be an issue.

I don't think your idea of just putting vtec on a switch will work. There is more than just the oil pressure switch in the vtec system; the ecu has an entirely different set of maps for ignition timing and injector values. The only way it would work is if the ecu was being triggered off the oil pressure sensor, which I don't think it is.

 

If he is using a chipped and tuned VTEC ECU he can change or disable the perimeters that are required for VTEC engagement. Therefore, as has been stated before, it can be triggered by the ECU at any time it has been tuned for in the tuning software. As long as there is enough oil pressure to slide the locking pin over in the rockers it will engage.

As for the vacuum issues. I was skeptical until I read up on the Cadillac Northstar engine and some of the others that use variable cylinder management technology. They all have a mechanism that disables all the valves and the ECU disables the fuel and spark for the given cylinders. I'm still not sure if 2 cylinders will be enough to really make this a fuel saving endeavor though.

I'm in for the results though

Cheers