I'd like to get a discussion going on what the "right" part-load mixture (for a fully-warmed up motor) is for different engine configurations. Why? Because for D-Jet and other systems, this is the most "fundamental" mixture adjustment, one that all of the other conditions (e.g. warm-up, acceleration, full-load, overrun, etc.) are all affected by. It's also important from a fuel economy and emissions perspective, as part-load is where the engine spends the majority of its time in normal operation.

To define what I mean by "part-load mixture", I'm talking about the CO level as measured by a shop-quality analyzer on a fully warmed-up car either on a dyno or running on flat road, at a small, constant throttle angle, with engine speed between 2000 and 3000 rpm (pick 2500 rpm if you want an exact value). Air temperature around 70 deg. F, near sea level.

I recently (last 2K miles) had my 2.0L rebuilt. Only mod is Euro pistons, otherwise, it's mechanically bone stock, including the cam, outfitted with D-Jetronic and a Crane XR3000 ignition set to standard timing specs. I initially set it up with an NOS, unmolested Bosch MPS. What I found was that the part-load mixture under the conditions I described above was about 4.5%, quite rich.

I have an MPS that's in good condition that I have removed the epoxy protection from the full-load stop, so that I can remove it and adjust the main mixture screw. I did some experimentation and found from road testing (with my CO meter attached), that I could lower the part-load CO to 2.5% while maintaining good driveability. If I went any lower than this, I began to have lean mixture issues, such as poor on/off throttle response. I did a number of abrupt throttle openings from part-load, and didn't hear any evidence of pre-ignition (knocking), either.

The most obvious change I've noticed since this mixture adjustment is that my fuel economy has jumped to about 26 mpg in normal urban driving. I don't have oil or head temp gauges. I intend to do some oil temp measurements (thermocouple down the dipstick tube), but the car doesn't seem to be running hot at all. I also will yank the plugs the next time I get a chance to see what they look like after a run under part-load conditions (tips on the "right" way to do this would be appreciated).

My question to the group: is this a reasonable part-load mixture level? Is there a better method for determining the "right" part-load mixture? How do the guys out there with carbs, L-Jet, and aftermarket FI setups determine their correct part-load mixture? Is Jake posting here any more, I'd really like to hear how he does it.

First I don't use a CO2 sensor. I and most that could help you will use a wide band O2 meter and temp gauges.
For power I use 12.5 to 13: AFR (air fuel ratio) it sucks a little on mileage but the engine runs real cool.
Then I've tried 14:1 AFR and milage goes up but so does the temps, but still well within the safe zone. I get this result when I run closed loop as well.
15-16:1 AFR and you start to get into the danger zone with the temps and I've never ran it long enough like this to check mileage.

This is at part load, At WOT I want to see the 12.5-13:1 AFR figure.

For those of you who prefer A/F ratios to CO levels when talking about mixtures, here are some conversions:

CO / AFR

5.0% / 12.64
4.5% / 12.83
4.0% / 13.00
3.5% / 13.20
3.0% / 13.38
2.5% / 13.56
2.0% / 13.77
1.5% / 13.94
1.0% / 14.11

http://perfectpower.com/technical_info/afr.asp

The aftermarket FI is easy. I put a heated O2 sensor in and let the computer adjust the mixture after so many ignition events. I'm looking for .49 to .5 volts back from the O2. Ideal is supposed to be .47, but the cars runs much better at a very slightly rich setting. This equates to I think right around 13.5-1.

13.5:1 (or 2.5%) is looking pretty good so far.

13-13.2:1

I use a wide band sensor and a Gunson gastester and set my mixture as
follows.

IDLE 1.0% to 3.0% CO with a AFR of 14.1 -13.4:1, No load so engine runs cool

Cruise RPM 1.0% to 3.0% CO with a AFR of 14.7 - 13.4:1, Good gas milage

Power 4.0% CO with a AFR 13.0:1, Runs richer to keep engine cool

With these reading I pass the two speed idle test for inspection.

12.8-13.5 A 2056 with a cam, matched ports, and 9.5/1 C.R. Ran nice and crisp on NGK-7's with a sweet power band above 4.5K. Any leaner and you start to lose the power above 4K.

With a 044 ECU I like to set it up to 13.7-13.9 to one on a 2056 and a 9550 cam, 9 to 1 CR.

I'm not an FI guy so I have the slightest idea how to tweak it. I can say most carb guys fight with part load mixture. If you check out the carb threads here 90% of them deal with transition issues. Transition from the idle jets to the mains happens in the 2-3k range. It usually ends up as a compromise between vent size, jet size and ET design. As long as it doesn't lean pop or fall on it's face the carb guys are generally happy. To change the mixture at this RPM you have to either change vents, jets or ET design. Sometimes all three. It's not an easy tweak with one screw and it usually costs money and takes time. Some folks don't have the patience for it. Like you mentioned it does have a huge effect on MPG. I shoot for 13.5:1. If it dips briefly into the high 12's I am not concerned. If it starts moving up to the 14's I have more of an issue. I'd rather spend a little extra cash at the pump than burn up an engine. Remember that 14.7:1 is the stoichiometric ratio for gasoline. As you creep up there you take the chance of burning up an engine. If you have blended fuel, like I do up North, you may need to richen it up even more.

A bit richer than others here are reporting, but you're the guy with the most engines by far. When you're tuning an engine on the dyno for part-load, what tells you that you've hit the right mixture? Where do you first see problems if you were to go leaner, e.g. 13.5:1? Might be something I'm overlooking.

The closer you get to stoich your temps will increase, basically starting into heat soak. These temps are great for mileage, but they start to and will kill power. I don't know on d-jet but almost all aircooled VW were slightly rich as the excess fuel actually aids in the cooling of the heads. CR also plays an important roll in temps and tuning.
Modern cars can run lean because they are watercooled.

One of the oldschool methods for recognizing a lean condition is crisp on startup and a loss of power hot.

Mark, thanks. Given that temps are cooler here right now, I'm thinking an extra 0.5% richer (3.0%) might be the best compromise for power and economy, and that gets me closer to Jake's numbers.