My quest for better fuel milage
#101
Rotary Revolutionary
iTrader: (16)
Joined: May 2006
Posts: 3,881
Likes: 2
From: Jacksonville, Tampa & Tallahassee
Karack:
that sounds about like what I used to get before all the fuel system upgrades... Hopefully I'm on the way back. Guess I'll be looking into timing after I see what kind of effect these fuel reductions make.
Wms10th:
That is awful! You can do way better. What turbo u running?
Pvillknight7:
I have never tuned w/a haltech so I have no way to know. Is there no compensation for load at all? Doesn't seem like a very safe/efficient system if that is the case.
that sounds about like what I used to get before all the fuel system upgrades... Hopefully I'm on the way back. Guess I'll be looking into timing after I see what kind of effect these fuel reductions make.
Wms10th:
That is awful! You can do way better. What turbo u running?
Pvillknight7:
I have never tuned w/a haltech so I have no way to know. Is there no compensation for load at all? Doesn't seem like a very safe/efficient system if that is the case.
Last edited by sharingan 19; 09-15-11 at 03:27 PM. Reason: Responses
#102
Haltech E6k. When I adjust fuel during highway cruising for lean but still smooth operation the engine lean surges in parking lots because the same rpm and pressure are required but the load is different. I think that's what happens...so I compromise and run a little richer during highway cruising conditions.
Is there a way to compensate for this?
Is there a way to compensate for this?
Do you use the closed loop on the E6k? Even though the computer is "primitive", I've read it works well. I don't, but I run mine (map) rich at lower RPM <2500 and lean it out 3000-3500. Maybe a functioning closed loop will let you putt around in a parking lot easy, but lean you out nicely on a steady highway drive.
#103
Thanks for the tips. I want to try a few other things before I try closed loop. Wasn't trying to derail the fuel saving topic. Engine tuning theory is the same regardless of what ECU or ECU signal modifier thingy like the SAFC.
Apparently closed loop makes a big fuel saving difference. Is there an easier way to test for closed loop operation besides hooking up this closed-loop indicator light?
Apparently closed loop makes a big fuel saving difference. Is there an easier way to test for closed loop operation besides hooking up this closed-loop indicator light?
#104
Moderator
iTrader: (3)
Joined: Mar 2001
Posts: 31,197
Likes: 2,825
From: https://www2.mazda.com/en/100th/
i've done a few haltech cars, and the pulsewidths at idle are usually a little HIGHER at idle than cruise. i do usually end up with a richer idle 12.2-12.5 and cruise is in the mid 14's. i doubt it would make any difference in MPG
#105
#106
i idle at ~2.3 ms at 1000 rpm (1.9% duty cycle), cruise at ~4.0ms at 3000 rpm (10% duty cycle). idle is 13's AFR, cruise in closed loop. if you do the math, that comes out to about 0.36 gal/hr at idle and 2.1 gal/hr cruising, so quite a difference. If your decel fuel cut does not work, its better to throw it in neutral
#107
Do you have Zero Throttle enabled along with your narrow band? Tune for lean 11's then enable zero throttle, get that sorted, then enable your fuel correction and get that sorted. Then check when going from cruise to WOT to make sure you dong go lean when you get into it.
Haltech E6k. When I adjust fuel during highway cruising for lean but still smooth operation the engine lean surges in parking lots because the same rpm and pressure are required but the load is different. I think that's what happens...so I compromise and run a little richer during highway cruising conditions.
Is there a way to compensate for this?
Is there a way to compensate for this?
#108
Thread Starter
Rotary Enthusiast
iTrader: (25)
Joined: Jun 2005
Posts: 1,420
Likes: 3
From: clearwater, florida
Well I definitely think the best way to go is to ditch the stock ECU, and as of now, I plan to starting in December. I'll probably be going MS2, but I've still got much research to do, etc, before I get there. Thanks for all the input guys, this thread really has been a giant help!! Hope some other people benefit off this as well
#109
Moderator
iTrader: (3)
Joined: Mar 2001
Posts: 31,197
Likes: 2,825
From: https://www2.mazda.com/en/100th/
i idle at ~2.3 ms at 1000 rpm (1.9% duty cycle), cruise at ~4.0ms at 3000 rpm (10% duty cycle). idle is 13's AFR, cruise in closed loop. if you do the math, that comes out to about 0.36 gal/hr at idle and 2.1 gal/hr cruising, so quite a difference. If your decel fuel cut does not work, its better to throw it in neutral
granted my stock turbo 3 rotor might be different enough from your 13B-T to cause that difference....
#110
Engine, Not Motor
iTrader: (1)
Joined: Feb 2001
Posts: 29,793
Likes: 119
From: London, Ontario, Canada
I was driving to work today thinking about this thread. As I happened to be driving the most fuel efficient gasoline production car ever created, my 2000 Honda Insight, I was thinking; why not apply some Insight engineering here? Here are a few quick examples.
First, rear fender skirts. The 1st generation Insight has a COD of 0.25 and part of the reason are the wheel wells.
It would be fairly easy to make up a set of rear fender skirts for the FC, which would eliminate the massive air scoop effect of the rear wheel wells. In fact, an FC with rear skirts was one of the design studies for the 2nd gen RX-7. Glad that it didn't go through because the car would look kind of funny, but in the case of fuel economy, they would make a considerable difference. Use rivnuts or other similar fastener to secure the panels to the wheel well lip for easy removal. Make a cardboard template and then recreate in fibreglass.
I see you have some of the factory aero pieces installed. You can go a step further and make the front lip around the front wheels much lower. Also widen it so that it is about as wide as the tire itself. This will funnel some of the air that would have slammed against the front facing tire around it and into the wheel well.
Looking at the Insight's wheel well, you can see that the wheel is set quite far inside. The front of the wheel is about 2" out from the wheel so that air can pass over the wheel instead of into the well. Notice the rear, where there is a significant scallop and lip to funnel air out of the well and along the body. Some fibreglass could bring the front of the well out a little, and then you could apply the same technique to smooth the rear.
Also, S4 phone dial wheels are probably the most aerodynamic. See how the Insight wheel is almost a total flat saucer? Sealing up the surface of the wheel provides far less drag and works with the fenders to channel air.
Another example of this are the windshield wipers:
Not a great picture but you can kind of see how the wipers are far below the "air line" of the hood. The hood also has a kickup at the rear to send air up and over the wipers and windshield. Another easy area to experiment with by using come cardboard and duct tape.
Finally, an idling engine is achieving 0 MPG. It's just sitting there, burning up fuel. Why not look to eliminate idle?
A very crude auto-stop setup could be made. Use a microcontroller such as Arduino or BASICStamp to read a few parameters and kill the engine when it's not needed. Measure TPS position, clutch switch, 1st gear switch (you'd need to add that), RPM and neutral switch (I think the FC has this?). It's all pretty simple to measure this stuff with very little in terms of support electronics. Most microcontrollers have several analog to digital channels which would do the job of reading the TPS (a simple 0-5V signal). Tach input is easy via an optocoupler (copy the Megasquirt tach input). Then the switches are just digital inputs.
The idea being that when you hit a red light, you push in the clutch, shift into neutral and then the microcontroller kills power to the injectors but switching on a NC relay. Engine shuts off. When you want to move again, you push the clutch and the micro switches the relay and cranks the engine via the stater.
Here's what the program would do....
Vehicle comes to a stop
If RPM < 1000 and TPS = IDLE and CLUTCH = pressed and GEAR = neutral Then Enable Injector Cut Pin
Loop and wait while checking inputs
If CLUTCH = pressed and RPM = 0 then
Disable Injector Cut Relay Pin
Enable Crank Relay Output Pin
Loop while cranking
If RPM > 500 then Disable Crank Relay and Exit Loop
If RPM < 500 then keep cranking
End Loop
End If
End loop
You kind of get the idea. You'd want a master enable/disable switch and it goes without saying that the car must be in good tune to start easily, and the starter must be in good shape. Probably not a useful system for ever little stop sign, but for sitting a few minutes at a red light, fuel savings could add up.
First, rear fender skirts. The 1st generation Insight has a COD of 0.25 and part of the reason are the wheel wells.
It would be fairly easy to make up a set of rear fender skirts for the FC, which would eliminate the massive air scoop effect of the rear wheel wells. In fact, an FC with rear skirts was one of the design studies for the 2nd gen RX-7. Glad that it didn't go through because the car would look kind of funny, but in the case of fuel economy, they would make a considerable difference. Use rivnuts or other similar fastener to secure the panels to the wheel well lip for easy removal. Make a cardboard template and then recreate in fibreglass.
I see you have some of the factory aero pieces installed. You can go a step further and make the front lip around the front wheels much lower. Also widen it so that it is about as wide as the tire itself. This will funnel some of the air that would have slammed against the front facing tire around it and into the wheel well.
Looking at the Insight's wheel well, you can see that the wheel is set quite far inside. The front of the wheel is about 2" out from the wheel so that air can pass over the wheel instead of into the well. Notice the rear, where there is a significant scallop and lip to funnel air out of the well and along the body. Some fibreglass could bring the front of the well out a little, and then you could apply the same technique to smooth the rear.
Also, S4 phone dial wheels are probably the most aerodynamic. See how the Insight wheel is almost a total flat saucer? Sealing up the surface of the wheel provides far less drag and works with the fenders to channel air.
Another example of this are the windshield wipers:
Not a great picture but you can kind of see how the wipers are far below the "air line" of the hood. The hood also has a kickup at the rear to send air up and over the wipers and windshield. Another easy area to experiment with by using come cardboard and duct tape.
Finally, an idling engine is achieving 0 MPG. It's just sitting there, burning up fuel. Why not look to eliminate idle?
A very crude auto-stop setup could be made. Use a microcontroller such as Arduino or BASICStamp to read a few parameters and kill the engine when it's not needed. Measure TPS position, clutch switch, 1st gear switch (you'd need to add that), RPM and neutral switch (I think the FC has this?). It's all pretty simple to measure this stuff with very little in terms of support electronics. Most microcontrollers have several analog to digital channels which would do the job of reading the TPS (a simple 0-5V signal). Tach input is easy via an optocoupler (copy the Megasquirt tach input). Then the switches are just digital inputs.
The idea being that when you hit a red light, you push in the clutch, shift into neutral and then the microcontroller kills power to the injectors but switching on a NC relay. Engine shuts off. When you want to move again, you push the clutch and the micro switches the relay and cranks the engine via the stater.
Here's what the program would do....
Vehicle comes to a stop
If RPM < 1000 and TPS = IDLE and CLUTCH = pressed and GEAR = neutral Then Enable Injector Cut Pin
Loop and wait while checking inputs
If CLUTCH = pressed and RPM = 0 then
Disable Injector Cut Relay Pin
Enable Crank Relay Output Pin
Loop while cranking
If RPM > 500 then Disable Crank Relay and Exit Loop
If RPM < 500 then keep cranking
End Loop
End If
End loop
You kind of get the idea. You'd want a master enable/disable switch and it goes without saying that the car must be in good tune to start easily, and the starter must be in good shape. Probably not a useful system for ever little stop sign, but for sitting a few minutes at a red light, fuel savings could add up.
Last edited by Aaron Cake; 09-20-11 at 09:47 AM.
#111
I was driving to work today thinking about this thread. As I happened to be driving the most fuel efficient gasoline production car ever created, my 2000 Honda Insight, I was thinking; why not apply some Insight engineering here? Here are a few quick examples.
First, rear fender skirts...
First, rear fender skirts...
http://www.autonews.com/article/2011...110919843/1172
#112
Moderator
iTrader: (3)
Joined: Mar 2001
Posts: 31,197
Likes: 2,825
From: https://www2.mazda.com/en/100th/
The idea being that when you hit a red light, you push in the clutch, shift into neutral and then the microcontroller kills power to the injectors but switching on a NC relay. Engine shuts off. When you want to move again, you push the clutch and the micro switches the relay and cranks the engine via the stater.
the cranking MS on the stock ecu is something like 13ms, which is a lot more than the 3 or so it uses to idle....
it would probably be better to just take right turns...
#113
Engine, Not Motor
iTrader: (1)
Joined: Feb 2001
Posts: 29,793
Likes: 119
From: London, Ontario, Canada
What serendipity-I had just read this before you posted...
http://www.autonews.com/article/2011...110919843/1172
http://www.autonews.com/article/2011...110919843/1172
you'd have to really look at this, but usually idling takes less fuel than the cranking/start does. the only reason its becoming viable is because we have things like the hybrids that don't have to add fuel while cranking, or mazda's stop start thing which also doesn't add fuel in cranking.
the cranking MS on the stock ecu is something like 13ms, which is a lot more than the 3 or so it uses to idle....
it would probably be better to just take right turns...
the cranking MS on the stock ecu is something like 13ms, which is a lot more than the 3 or so it uses to idle....
it would probably be better to just take right turns...
#114
#115
I'd bet all the aero tweaks the OP has tried/considered don't together equal the gains from going to skinny wheels with low rolling resistance tires....
#117
Moderator
iTrader: (3)
Joined: Mar 2001
Posts: 31,197
Likes: 2,825
From: https://www2.mazda.com/en/100th/
It would only be useful when the idle period exceeds the amount of fuel needed to restart the engine. I guess one could calculate 2 seconds at 13ms vs. 3 minutes at 3ms. Actually, just typing that makes it fairly clear. The microcontroller could also play a trick by cutting the fuel pump out during this cranking to keep fuel pressure low. A hot engine starts on much less fuel than the stock ECU is dumping in.
#118
#119
Rotary Revolutionary
iTrader: (16)
Joined: May 2006
Posts: 3,881
Likes: 2
From: Jacksonville, Tampa & Tallahassee
I figured the insight was slicker than .25 cod, wasn't the s4 sport like .27 ?
Could be worse I suppose... I'm at 110 miles and still over a half tank since I filled up and tweaked the cruise map. All city driving and I've been drifting 3 times, lol.
Actually, I do see a very slight difference in that an owner with limited resources retrofitting modern design elements onto a 20+ year old car in an attempt to improve efficiency is admirable. Whereas a manufacturer with substantial resources attempting to (what's the opposite of retrofit? Lol) decade old designs in lieu of designing efficient vehicles is rather lame.
The wheels/tires are pretty narrow. When u say "low rolling resistance tire" are you referring to a specific compound or just over inflation?
No doubt true, but then again, I don't see sticking skirts on a 7 is much different than the Malibu in the sense that the RX wasn't designed for them either.
I'd bet all the aero tweaks the OP has tried/considered don't together equal the gains from going to skinny wheels with low rolling resistance tires....
I'd bet all the aero tweaks the OP has tried/considered don't together equal the gains from going to skinny wheels with low rolling resistance tires....
The wheels/tires are pretty narrow. When u say "low rolling resistance tire" are you referring to a specific compound or just over inflation?
#121
The S4 Aero Package drops the car from .31 to .29 [as an amusing side note, .29 is also the coefficient for the FD, Mazda3, Subaru SVX, 1st gen Prius, Versa, 90's Eclipse, 2011+ Audi A4/S4, 2008 Infiniti G35, etc.. ]
#122
Moderator
iTrader: (3)
Joined: Mar 2001
Posts: 31,197
Likes: 2,825
From: https://www2.mazda.com/en/100th/
try it, you don't need a fancy controller, just turn the key with your hand.
#125
Engine, Not Motor
iTrader: (1)
Joined: Feb 2001
Posts: 29,793
Likes: 119
From: London, Ontario, Canada
No doubt true, but then again, I don't see sticking skirts on a 7 is much different than the Malibu in the sense that the RX wasn't designed for them either.
I'd bet all the aero tweaks the OP has tried/considered don't together equal the gains from going to skinny wheels with low rolling resistance tires....
I'd bet all the aero tweaks the OP has tried/considered don't together equal the gains from going to skinny wheels with low rolling resistance tires....
And yes, low rolling resistance tires! That would make a huge difference, especially when over inflated. By upping the pressure on my Insight from the stock ~32 PSI to 50 PSI, I picked up an average of 8 MPG in the city and made it far easier to cruise along at 90 - 100 MPG on the highway. The downside? The handling of all LRR tires sucks donkey *****.
I know the math isn't that simple, but 13ms for about 1.5 - 2 seconds at a cranking speed of 250 RPM is a lot less fuel injected than say, 3 minutes at 3ms at 750 RPM.
But more to the point, does the stock ECU really use a fixed duty cycle to crank the engine regardless of temperature? I tune most standalones to crank at about 5 - 6 ms once the engine is at operating temperature and they start with the flick of the key.
Yep, or you could do that. But it's not as cool.