Fine tuning EFI for mileage

[sniperx]

I've got my EFI setup dialed in pretty well. I have a 2110 motor with a 110 cam in it (up for debate) and 8:1 compression. I'm running a 3.88rp and 82nd 4th gear on 6.70 R15 tires in a 67 camper with reduction boxes. I got 27.6mpg over 84 miles with a 7% grade for 3 miles. I want more. I haven't messed with my fuel table, I'm just letting "learning" and closed loop tune for me at the moment. I'd like to maximize my timing advance, I'm running a crank fire system so its pretty well unlimited. I'd also like to work on going toward the lean side of peak at cruise. How do I begin and how can I tell when its too lean (since past stoich CHTs drop)? How can I tell a lean det from a timing det? I'm currently running under 350 CHT on #3 spark plug on the highway and on climbing.

Any help on this? I'm trying to get to 30+

[Steve Arndt]

I read every article online about ROP, LOP, and mixtures on the aircraft sites before I tried it.

I'm now running around 16:1 to 16.5:1 between 2000 and 3500 rpms at around 50 to 75 Kpa cruise map. Timing is ~ 10 more degrees in the high vacuum areas (50 to 75 kpa) compared to 92 Kpa barometric here. The timing really smooths out the response. Trying to run lean without the extra advance makes it hunt/surge and feel ill-tuned.

I was also able to pull quite a bit of fuel from the map in the low rpm free rev areas with the extra timing implemented.

edit:
The cruise CHT dropped a bunch. I will have some concrete numbers after this weekened.

[SUbuggy]

I read every article online about ROP, LOP, and mixtures on the aircraft sites before I tried it.

I'm now running around 16:1 to 16.5:1 between 2000 and 3500 rpms at around 50 to 75 Kpa cruise map. Timing is ~ 10 more degrees in the high vacuum areas (50 to 75 kpa) compared to 92 Kpa barometric here. The timing really smooths out the response. Trying to run lean without the extra advance makes it hunt/surge and feel ill-tuned.

I was also able to pull quite a bit of fuel from the map in the low rpm free rev areas with the extra timing implemented.

edit:
The cruise CHT dropped a bunch. I will have some concrete numbers after this weekened.
I read every article online about ROP, LOP, and mixtures on the aircraft sites before I tried it.

I'm now running around 16:1 to 16.5:1 between 2000 and 3500 rpms at around 50 to 75 Kpa cruise map. Timing is ~ 10 more degrees in the high vacuum areas (50 to 75 kpa) compared to 92 Kpa barometric here. The timing really smooths out the response. Trying to run lean without the extra advance makes it hunt/surge and feel ill-tuned.

I was also able to pull quite a bit of fuel from the map in the low rpm free rev areas with the extra timing implemented.

edit:
The cruise CHT dropped a bunch. I will have some concrete numbers after this weekened.

all sound advice. here's my .02

#1 what plug gap are you running? run the BIGGEST plug gap your igntion system will cope with. this is why the OEM's run .065+ gaps with very powerful coils, (it wasnt to be nice to the hi-perf guys) the excessively lean mixtures need a very powerful long duration spark not not just start combustion but sometimes to sustain it and keep misfires away.

#2. REPEATABLE data. find a smooth long section of highway to do this, maybe 2+ miles. log your AFR's, temps, igntion advance, throttle position and so forth. then when you make a change, go back to the EXACT same stretch of road, run the car under the same conditions and observe any changes.

#3.dont be afraid of detonation. when the motor is in high vacuum at higway conditions, there isnt enough cylinder density (load) to detonate. Misfires are the issue to worry about, basically the fuel burn is not complete and ignites the incoming fuel on the next stroke, manifested as backfire throught the INTAKE.

#4 you will have reached "ideal" or peak efficiency for your engine combination when you can pull the vehicle at a given repeatable speed in the same gear under the same conditions when you achieve more vacuum (lower KPA)in the manifold. think tune to the highest manifold vacuum without misfires.

[Steve Arndt]

With regard to #1:
I run .040 gap, four of the bad ass IGN1 coils, with dwell mapping. Base dwell is 3.0 ms, and I ramp that up based on increasing pressure. I could probably run a bit more dwell than I am now.

Regarding #4:
I noticed much better (lower, and smoother average reading) map values at idle when I implemented idle based timing control, and increased the high vacuum advance. I had to pull quite a bit of fuel after those changes to get back to the same AFR as before. So more vacuum in the manifold, using less fuel, for the same idle AFR of ~13.8. All wins I think.

[sniperx]

ALRIGHT!!! This is what I've been looking for. Simple, concise information based on real world experience and usage. I'll have to read this a few more times then look at my gear and see what I can translate. Are there any specific articles or directions I should look?


For fun question: What kind of mileage do you think is realistically achievable with this setup? I'm at 28 or so now. My baseline info was done with the grade, fully loaded camper (gear, 100+ pounds dog, full tank, 2 people) and 2 doors strapped to the roof rack. My number on that was 27.6..thats why I say 28 or so with a more normal condition.

[SUbuggy]

For fun question: What kind of mileage do you think is realistically achievable with this setup? I'm at 28 or so now. My baseline info was done with the grade, fully loaded camper (gear, 100+ pounds dog, full tank, 2 people) and 2 doors strapped to the roof rack. My number on that was 27.6..thats why I say 28 or so with a more normal condition.

impossible to say becasue who is to say how far your tune is off now? also how you drive makes a huge difference.

With regard to #1:
I run .040 gap, four of the bad ass IGN1 coils, with dwell mapping. Base dwell is 3.0 ms, and I ramp that up based on increasing pressure. I could probably run a bit more dwell than I am now.

just log coil duty cycle. if you are under 75% you should be fine. are you running sequential or WS?

[quoteRegarding #4:
I noticed much better (lower, and smoother average reading) map values at idle when I implemented idle based timing control, and increased the high vacuum advance. I had to pull quite a bit of fuel after those changes to get back to the same AFR as before. So more vacuum in the manifold, using less fuel, for the same idle AFR of ~13.8. All wins I think.
][/quote]

exactly. any time you can use less RPM or MAP and achieve the same output (vehicle speed) you are operating MORE efficiently (fuel economy)

[sniperx]

From my reading the fly boys are running rich of peak to peak until cruising speed/altitude is reached. This is the equivalent of our straight smooth stretch of highway test. Once they reach this point they start pulling fuel until they hit max egt. That is peak. Then they pull a little more based on their experience to about 20-40 degrees drop after peak. These numbers are relative numbers of course, but the principle would be the same for our world.

My two hang-ups now are...timing and cells.

Cells. A plane pretty much runs at a fixed speed at altitude. On our ground based life we have other problems. For example small changes during an easy pass, accelerating after braking slightly, and etc. How do you blend your LOP cell or cells to the ones around them without reaching into stoich or the danger zone.

Timing. How is timing used throughout this idea?

I'm running tps based because my vacuum is running 80kpa at idle. The best I ever saw was 60kpa at very low idle. Everything is in good condition and the cam is thought to be closer to a 130 due to a misunderstanding in the build.

[sniperx]

Here is the answer to it all....


http://www.avweb.com/news/pelican/182084-1.html

[Max Welton]

Cells. A plane pretty much runs at a fixed speed at altitude. On our ground based life we have other problems. For example small changes during an easy pass, accelerating after braking slightly, and etc. How do you blend your LOP cell or cells to the ones around them without reaching into stoich or the danger zone.

I arranged my table indexing so that my lean and power AFRs have a steep transition, minimizing the area in around stoichio.

There are areas containing 14.7, but they are narrow and in my normal driving I don't spend much time in them. Usually my AFR just goes from 13.5 to ~15.9 in a nice step.

Max

[sniperx]

Any reason your closed throttle is so rich? Mine drops down to 5 for pulsewidth right there...which is almost off.

Do you feel the transition as a jolt or anything? Thats my concern and possibly falling in and out of LOP through small terrain changes or throttle changes.

Another thought. Would it be possible to use accelerator enrichment to help in passing while in a large lean of peak block?

I'm beginning to understand the fuel table as a sort of dart board in these cases. You set it for repeat-ability on the highway...then in acceleration, passing, hills, whatever...the chance of landing on the LOP cell again during those conditions is pretty slim. Of course if you change that from a cell to a block...your probability increases and you may hit a lean cell while accelerating or something.

One problem I've seen is some people...not here that I know of, are just repeating what they've read and leaving out key info. They say that running lean can cause a rough condition. According to the article above..these lean conditions are caused by an uneven mixture/charge caused by the design of their intake for airplanes. In a dual throttle body based on the weber pattern...you can't get much more even than that. The only thing that remains is the fuel charge...that is handled through balanced injectors. Most high end injectors are tested to be in tolerance anyway...the plane boys talk about hand tuning injectors...again to compensate for airflow problems. They can't just swap manifolds or something due to FAA rules and what not...something we don't have to worry about.

Another is the need for multiple EGT probes or CHT points. Again...in the weber pattern case...this isn't necessary for the same reasons. Same goes for burning up your motor due to a lean cylinder...shouldn't happen with good injectors and an accurately balanced setup. Of course non-dog house, missing tin, and a single throttle body throw all this out the window.

Heres some fun programming I thought of. If you hack into your software and change the IAC role from idle to lean tuner you could use it to run LOP during any long term cruise (constant throttle and RPM over 2 seconds would activate it). It would then lean out to a predetermined (by you through tuning EGT or CHT) AFR and keep it at that. This way it wouldn't be a highway only thing...but also a 45mph/35mph or any other constant point thing.

[Max Welton]

Any reason your closed throttle is so rich? Mine drops down to 5 for pulsewidth right there...which is almost off.

My "closed throttle" MAP is around 20 kPa if gravity is driving RPMs. Closed throttle at idle is around 48 Kpa. I found that when descending a steep hill in gear, I was getting a rough running feeling if I stayed lean at that point. There is very little flow at that time so I just gave it some fuel.

Do you feel the transition as a jolt or anything? Thats my concern and possibly falling in and out of LOP through small terrain changes or throttle changes.

I actually don't feel it at all when crossing the horizontal 66 kPs line. Not making much power there. Or maybe my butt isn't sensitive enough. I do notice that the head temp begins to come down after a minute or two in the lean box.

But if you'll notice, I have a little wet area in the upper left corner of my lean box. During a gentle (say, 30% throttle) acceleration through the gears, I tend to cross through that corner and if it goes lean there, I definitely feel it. If I accelerate hard, the RPM/MAP trace is all above the 66 kPa boundary. But I need a better solution to that spot. Because there is a bit of a ramp there, I can wind up around 14.7 in a climb taken in fourth gear between 2400 and 3100 rpms. I may reindex again so that I have more cells from 2000 to 3000 rpms.

Now, if the tables were, say, 100x100 my lean box could be much less rectangular and I could avoid ramps where I don't want them.

I'm continually fiddling with this. Just now I noticed that I have an indexing mistake at the lower edge of the lean box. I had intended to put that line at 20.0-20.1 kPs. All those 13.5s on the 20.0 kPa line should have been 15.9. Not a critical area of the map for me, but I'll fix it.

PS: Here's a recent log:

Max