5% increased gas mileage

[88turboii]

so ive been trying to figure out how to squeeze some extra mpg out of my TII.. trying to break 30mpg. currently running haltech e6k, 750cc primary injectors, RB 3" exhaust, cold intake, 60-1 hybrid turbo. cruising timing is about 27 deg, 0 split at 2500 rpm (60mph). If i keep it under 60 mph, i can get about 28mpg. if the headlights are up or a/c is on, ive found i lose about 1-2mpg each

my buddy mentioned an idea about the keeping secondary throttle plates closed during cruise and i figured i would try it out. Well if you look at the throttle body works, it does not take much to movement to start cracking open the secondary plates, so normally during cruise, they are cracked open. so i reinstalled the 3rd set of throttle plates, which are commonly removed for the TB mod, and i can rig them open or closed with a spring. this will throw your fuel p/w off during boost (map based tune), but its ok in cruise because i just run in closed loop fuel

during my last trip (5 hours driving one way), on the way out i rigged them closed, so only using primary throttle plate. on the way back I rigged them open, which would be how it works from factory fully warmed up. it made a difference of about 2mpg (28 vs 26 mpg), so about 5% gas savings. unfortunately, i had my lights for both trips, so still havnt broken 30mpg

anyway, so the next thing i was going to do is wire in a solenoid to control whether the valves are open or closed. this would be connected to an ECON button in the cabin, similar to modern hybrids. Im thinking you could also wire this to your standalone ECU to trigger dual maps and boost settings

[RETed]

Can't you just tweak your closed-loop parameters?


-Ted

[Pete_89T2]

Interesting thread... I'm currently running the stock ECU with the Rtek v2.x mod, and was toying with the idea of leaning out the off-boost/cruise part of the map a little bit to improve highway cruise MPGs. Basically looking to play with only those cells in the RPM range of 2000 - 3200, at MAPs deep in vacuum below 10 inHg (i.e., steady cruise, light load conditions). Unfortunately, I don't have anything to monitor AFRs or EGTs, so I'm a bit hesitant to screw with this. Is there any real danger here, presuming I do this intelligently, i.e., small adjustment increments followed by road testing to assess results?

My "baseline" highway MPG average to date is pretty good as-is. I got just shy of 25MPG (24.9), measured on my way back from DGRR/Mitty last year. That was over a 10 hour trip, average speeds between 70~75MPH. As for lights & A/C, lights were on for maybe the 1st 2 hours (dawn/early AM), and A/C was on for maybe the last 7 hours of the trip; it started getting hot by mid-morning. At that time, car was set up with 720cc injectors in all 4 holes, Supra TT fuel pump, BNR stage 1 turbo, Bonez DP into the stock main cat and Borla catback exhaust. Otherwise basically stock (stock S5 boost control, working air pump, emissions, stock fan).

Since then, I got her registered as a historic (no more sniff test), so I yanked the main cat & air pump/emissions gear and replaced it with the RB 2.5" DP and pre-silencer combo that bolts up to the Borla catback. Another trip to Dave at Speed1 is in order for some dyno tuning, since new the setup is now able to consistently hit 8psi boost with the stock boost control solenoid doing its thing. I was previously limited to about 6psi, due to the aging main cat.

[NoDOHC]

I would look into leaning the fuel mix in vacuum.

You should have a wideband on the car anyway with a turbo, it can give you advance warning before you damage the engine if something in the fuel supply fails.

In the summer my 4-port gets mid-30s on the highway (although I have never logged a tank over 33.6 MPG - due largely to my right foot having a problem on certain onramps, etc.) In the winter it gets about 30 MPG highway.

In my opinion, your timing in vacuum is 20 degrees retarded. However, you should read my transmission failure posting before you advance yours.

At 2500 RPM I run 30 degrees BTDC at 0kPA vacuum (WOT). I run 60 degrees BTDC at -61.3kPA (weird spot for a data point, I have a 3-bar MAP sensor) - the curve is linear between these two points. Porting is conservative, no changes to stock port timing, idles at -52 kPA, 14.2:1 AFR, 5 deg BTDC @ 900 rpm.

Typical cruising vacuum (-45kPa) at 2500 rpm = 62MPH on my 0.72 overdrive and 4.1:1 diff. I am running about 52deg BTDC and injecting about 2.5ms duration with 550 cc/min injectors.

[RETed]

Originally Posted by Pete_89T2

Interesting thread... I'm currently running the stock ECU with the Rtek v2.x mod, and was toying with the idea of leaning out the off-boost/cruise part of the map a little bit to improve highway cruise MPGs. Basically looking to play with only those cells in the RPM range of 2000 - 3200, at MAPs deep in vacuum below 10 inHg (i.e., steady cruise, light load conditions). Unfortunately, I don't have anything to monitor AFRs or EGTs, so I'm a bit hesitant to screw with this. Is there any real danger here, presuming I do this intelligently, i.e., small adjustment increments followed by road testing to assess results?

No, you don't *need* to monitor AFR or EGT if you know what to look for...
It will help to have an EGT gauge though - you want everything to be burning at it's hottest temps - misfires will cause the EGT to drop.
No need for a wideband - I hate those things.
A narrow-band AFR gauge is "good enough" in my book, as it will tell you that you're at least LEAN and not RICH - it's cheap enough that it's no excuse not to have one, unless you already threw away your stock O2 sensor?

The idea is to lean the fuel out and advance as much as possible.
Realistically, you can usually only advance the timing up to 50 - 60 degrees of total advance.
Lean out the fuel up to the point where you can tolerate the "lean surge".

There is no danger in doing the above if you do go beyond the "ideal" range.
If you do lean it too much, the lean surge will get annoying - the car will buck noticably.
If you keep leaning past that point, the engine will stop firing.
It is a trick for emission to run ultra advance ignition timing to burn as much as the combustion as possible.
Just don't run more than 60-degrees, as this is the point where you start to fire into the other rotor - which is not that big of a deal since the prior rotor face is going into the exhaust cycle, but it's still not recommended.

This all should work if you're keeping to legal speeds.
This means that for most FC's, this should be under 3k - 3.5k RPM's.
Don't try to do this over 4kRPM's - we're talking about speeds approaching triple digits at this point for most FC's.
At this point, the engine is running fast enough to negate effects of early lean surge.
The engine is so smooth, you don't feel the typical lean surge, and the EGT's start to climb dangerously.
I guess a "disadvantage" of the rotary engine?
I've seen temps hitting 1050C and climbing...this is getting close to 2000F.
Sustained EGT's at this level will cause damage to the engine.


-Ted

[Pete_89T2]

Thanks Ted, good info there. Still have my O2 sensor hooked up, and with the Palm hooked up, the Rtek Pocketlogger software can datalog the O2 signal, and just about everything else that the stock ECU uses as an input (i.e., MAP, TPS, water temp, intake air temp, RPMs, etc.). Sounds like this may be worth the effort.

[t_g_farrell]

Jamin get a honda

[reese.leblanc]

Ted, can I assume that raising the timing to that level (i.e. 50*-60*) in the vacuum would be done in the 1500-3500 range? I would think that doing it in the 0-1000 ranges would significantly affect trying to set up idle, etc?

[RETed]

Originally Posted by reese.leblanc

Ted, can I assume that raising the timing to that level (i.e. 50*-60*) in the vacuum would be done in the 1500-3500 range?

Yes, and all the way up to redline...
Above 4kRPM's, it's more for high RPM decel.
Most of us run decel fuel-cut off, so there is little to zero fuel in the combustion chambers.
Ultra high advanced ignition timing is purely for emissions reasons and has nothing to do with power.

I would think that doing it in the 0-1000 ranges would significantly affect trying to set up idle, etc?

In actuality, I doubt the ECU would even go into those ranges...
Really deep vacuum under 2kRPM is almost impossible to hit in most normally running engines.
The "best" vacuum is usually at idle.
(Most of us run anywhere from stock -5 degrees to up to +10 degrees total advance on idle.)
Cracking the throttle will cause the vacuum to head towards "0".
When idle speeds start to drop (500 RPM or less), vacuum heads to "0".
If you're running idle control, your idle control / BAC valve is going to bypass air when RPM's start dropping toward idle speed, so this hurts vacuum too.
I like to just "flatten" the map in those areas, just so the map looks pretty.
It's unlikely the ECU has to access those portions of the map in reality.


-Ted

[NoDOHC]

I don't share Ted's hate for wideband sensors, but I otherwise agree with his post above.
There is certainly a place for a wideband, however they certainly do not tell the whole story of the combustion process.

To idle at 200 RPM a stock 4-port rotary engine runs about -25 kPA manifold pressure. To 'idle' at 2500 RPM will run about -70 kPA manifold pressure. The 6-port is able to run at about -30 kPA at 200 RPM. Rotary engines don't run reliably below 200 RPM.

You can never access points on a MAP curve below the closed throttle MAP at each RPM. Mine cranks at 325 RPM and has -10 kPa Manifold pressure at that speed. I can't get a lower pressure (more vacuum) at 325 RPM without adjusting how low my car idles (closing the throttle plate further). Thus I only have access to 3 MAP points at 325 RPM (only 3 MAP points are valid operating parameters for my engine). My engine runs -90kPa at 6,000 RPM, closed throttle, thus I can access pretty much all the load points at that speed.

As Ted said, the BAC valve will not allow much vacuum below the idle setpoint. (Which is typically 750 rpm). The 4-port in my car has no BAC valve, thus it idles only at the speed set by the throttle plate. This amounts to about 550 rpm cold and 900 rpm warm, with a brief peak at 1,000 rpm as the oil gets warm but the mix is still rich.

To tune without a wideband, just lean the fuel up about 10% from stoic, (if you have a narrow-band sensor) or until the engine starts to surge (not really a mis-fire, but a slower torque variation, almost feels like driving into a headwind on a windy day).

If the weather is warm you should be able to break 30 MPG, I can barely get 30 in the winter, but it is usually well above that in the summer.

With vacuums deeper than about -55 kPA, the engine requires a richer mix (probably due to exhaust reversion). This is why most of us idle at -5 to 5 degrees BTDC to avoid running very rich at idle. This is unpleasant if you have to sit in traffic for very long. If you run more advanced timing, you will pull more vacuum, which will require a richer mix. My engine will run just fine at 16:1 at WOT at 1000 rpm (it is not tuned to do so, but it will), but would not even dream of idling at that AFR. You can idle at stoic if you run -5 deg BDTC, but you end up injecting more fuel than you do to idle at 14:1 at 5 deg BTDC.

[RETed]

Another area to look at if your ECU can do this is try negative ignition timing split - this is just a fancy way of saying: fire the trailings before the leadings.
There was this big thread on the Evil Forum (I think) talking about this...
I tried it with my Haltech E8, and it does allow you to run slightly leaner AFR's at idle and vacuum cruising.
I believe the numbers were you could gain another 5% - 10% in gas mileage doing this properly.

Now, the cavaet...
Although I was able to do this with my Haltech E8, it wasn't stable enough that I kept at it.
My set-up acted tempermental and sometimes felt like lean surge when cruising under heavy vacuum.
This annoyed me enough to give up trying to get that last few 1 - 2 mpg out of the engine.


-Ted

[NoDOHC]

Good call Ted, I forgot that I am running negative split at idle. I was able to run more vacuum without misfiring by running 5 degrees negative split at idle. I also run negative split over 6,500 rpm. You need to have relatively advanced ignition timing before you try it (at least 5 deg BTDC, negative split causes a lot of surge with -5 BTDC timing (the factory setting of 12-15 degrees of positive split is much better when retarded that far).

I have not tried negative split while cruising. I will give that a shot once spring comes.

I am not sure that your e6k will support negative split. I think not, actually.