Ultragauge and Lean Burn - Estimate MPG?

[darcane]

Well, I tried a tank focusing on Lean Burn. UG indicated 40.7mpg, actual mileage 46.6mpg. That's down a little from previous tanks, but it was significantly colder (30-40°F cooler in the mornings).

I think I'll stick with it for highway cruising as it requires far less effort and has less impact to other drivers.

[brucepick]

Today is my first time viewing this thread - most interesting!

My HX is a '97, the generation before darcane's HX. I expect much of the lean burn control logic in darcane's HX is the same as on my car but I can't guarantee it.

Anyway, I have a 55 mile jaunt to work daily. For the highway, I've settled on a blend of Lean Burn and pulse-and-glide. Details...

I don't kill the engine (usually) since that requires 1.5 to 2 miles distance under power to make lean burn available again. I pulse the engine at the max throttle I can, without losing lean burn. I do that till it reaches my upper speed limit - somewhere around 70 mph. Then neutral coast till speed drops to my lower limit, around 50-55. Then back into gear, baby it just right to get it into lean burn, then max allowable throttle again.

I use this principle pretty much the same, regardless of whether I'm going up grade or down. Going up, it takes longer to reach maximum speed, and I may never reach it. Going down, it only takes a few seconds to reach max speed before I put it in neutral for a nice coast.

I have one coast over a mile, shortly before exiting the interstate. I do kill the engine for that coast, and then use a more conventional EOC with engine shutdown the rest of the way to work. No sense in preserving the lean burn capability if I'm getting off the highway anyway.

My MPGuino is very good at measuring fuel use even with a lean burn engine. It's nearly always within +/- 0.2 gallons when I fill up, often within 0.1 gal.

[darcane]

And what is you max allowable throttle with lean burn? I think I can get it up to about 30% TPS.

I may have to try this out.

[brucepick]

I've had max throttle at 32 TPS however in many situations I found the limit more like 30, and at lower rpms the limit was about 28 and I think sometimes even lower than that.

So I switched to using Load (LOD gauge on my ScanGauge) to manage Lean Burn. The limit stays right around 90, which I think is percent. Occasionally I will lose Lean Burn below 90 Load but it's a pretty good way to judge - in my experience. Better than TPS.

As you gain or lose speed with TPS held steady, Load will increase or decrease. But the limit of approx 90 stays pretty consistent. You get to step into the pedal deeper (or have to lift your foot some) depending on those Load changes, but at least you can see the change happening on the gauge. If using only TPS, the limit number changes but only experience tells you what the limit is at any given rpm. that's why I use Load now.

Yesterday I was able to free up one position on the gauge, so I can see Load and TPS simultaneously. I haven't had enough time behind the wheel since the change yet, but eventually I might get an idea of how the two readings can be used together to manage Lean Burn.

[darcane]

Thanks! I've starting trying that with my commute this morning. Seems better, but I won't know for sure until I fill up.

Any tips on getting it into LB quickly? After the coast, it takes me some time to entice it back into LB. I usually leave it under 18% TPS and slowly lower it if it won't engage.

I was getting some screwy numbers with load, so I switched to a different gauge. Maybe I'll try it again.

[darcane]

So, I started using % Engine Load again. The numbers still seem screwy to me.

5th gear, ~30% TPS, 2200rpm = 90% load?
Neutral, foot off throttle, 600rpm = 25% load?

25% load at idle, and 90% at 30% throttle? WOT puts it at 99%.

On my Silverado, idle was about 5-10% load. And opening the throttle to 30% at low rpms put me around 30-35% load. Are the numbers screwy or am I just used to seeing them on a much different engine?

[brucepick]

The load and throttle position sensor (TPS) numbers I've seen in my HX are basically same as yours. I have no Silverado experience to compare but it makes sense to me.

If you've seen those discussions of BSFC charts, the game for fuel efficiency is to have a small engine loaded kinda near max, not have a big engine choked off with high vacuum numbers.

First, load % is the pct of what the engine could do at that rpm, not vs flat out at red line rpm. At lower RPMs it can't suck in much air/fuel so by opening throttle even a bit, you soon let in almost as much air as it could possibly suck in. So TPS 30 or so will get you about 90% of max output (load) at about 2000-2200 rpm. OTOH, at 4000-6000 rpm it would be able to suck in so much air that TPS of 30-32 would actually limit it's output. So % load is reported as lower, for same TPS at higher RPMs.

About idle load figures. Similar logic but now you're down around 670 rpm. True, your foot is off the pedal so TPS is only 9 (on my car). However the idle air control holds the bypass valve partially open to maintain idle, so there is air flow even though TPS is closed (9). The approx 25% load you see is the result of that air flow vs what the tiny engine could do at that rpm if wide open. Apparently it calculates from air flow and not from throttle position.

Silverado has a much bigger engine, and likely has an idle speed that would provide more power at idle rpm, if wide open. At idle, it's probably running at higher vacuum (more air restriction) than your little civic engine. Just my guess though. Anyway, a 2x bigger engine running at an idle speed that is not half the Civic idle speed will show a smaller percent load at idle. But probably uses more fuel at idle than the Civic.

[darcane]

I am struggling with your technique, brucepick. My biggest issue is that it takes 3-15 seconds to engage LB when the conditions are met. So, instead of a 10 pulse, 15 second coast, it takes me up to 30 seconds to pulse (wait for LB then slower acceleration to maintain LB) for a 15 second coast. I seem to be using more fuel this way. I am guessing the reason for this delay may be either due to it being a California model or something with the 7th gen Civic.

Of course, adding to this is the colder weather, heavy rain, and abysmal traffic.

[brucepick]

darcane, are you using Marvel Mystery Oil in the fuel? I use 2 oz per 10 gallons. It definitely helps a lot.

I only use lean burn on the highway, where pulse and glide is less practical. And yes, you do have the delay before lean burn kicks in.

If you have an MPGuino, you can usually see the LB kick in sooner on the Guino than on a ScanGauge. Because the SG only updates about every two seconds, but the Guino updates 2x per second. My point is, it might not take quite as long as you think for LB to happen.

[darcane]

Originally Posted by brucepick

darcane, are you using Marvel Mystery Oil in the fuel? I use 2 oz per 10 gallons. It definitely helps a lot.

I only use lean burn on the highway, where pulse and glide is less practical. And yes, you do have the delay before lean burn kicks in.

If you have an MPGuino, you can usually see the LB kick in sooner on the Guino than on a ScanGauge. Because the SG only updates about every two seconds, but the Guino updates 2x per second. My point is, it might not take quite as long as you think for LB to happen.

No, but I use TC-W3 oil as a fuel additive which serves the same purposes (top end and fuel system lubrication, cleansing via detergents). I've used it long before I bought the HX.

My Ultragauge seems to update about once per second, maybe a little quicker. I've tried accelerating just before I am sure it is engaged (O2 sensor typically shows 0.015 before showing 0.000) and it isn't.