Here's the update to my original thread on this subject:

Fun with D-Jet, a WB O2 Sensor, and a 2056

So I made the adjustments that I talked about, and it turned out pretty good - but had some driveability issues. I was getting a hiccup when shifting gears (only during full acceleration through the gears), and was having a hard time with consistent idle. So I decided that I had pushed it a little too far by having the full load diaphragm beginning to richen the mixture all the way out at 12-13 inHg (stock is 6 inHg).

So I dialed the point at which the diaphragm comes off of it's part-load stop down to about 9 inHg for a compromise between the two.

I also drilled a small hole in the end cap of the MPS so that I could access the small screw with the engine running. This turned out to be a good move as that screw is very sensitive - and removing the MPS and it's endcap to make every adjustment was too tedious.

Once I had the part-load stop set at ~9, there is no other adjustment necessary for the outer screw - so I could make all further adjustments with the MPS in the car and the engine running.

As I said originally, I leaned out the overall mixture. The results are below. You will notice a significant change in the mixture at steady speed - not nearly as rich as it was originally. And of course, the onset of richening the mixture now happens much earlier - which now makes sense. It didn't make sense to me that the motor would continue to lean out all the way to medium acceleration.

Still, this is d-jet, and the adjustments are limited. This has shown me how much of a compromise d-jet really is. My a/f ratio curve is far from ideal, but I think it is better than stock.

You will notice the dip in the curve at around 2 inHg. I didn't originally show this dip in my original graph, but the changes I made accentuated this dip. The engine goes richest at about 2 inHg, and then leans out a bit at full throttle (0 inHg). This dip is another shortcoming of d-jet and the MPS. I believe it is due to the diaphragm hitting the full load stop a little before full throttle. This is unfortunate - I wish the diaphragm had a little more travel and was a little stiffer - then I could get nearly exactly the curve that I would like.

But all in all it's pretty good. Drivability is excellent. Idle is better, throttle response from idle is better. Freeway fuel economy should be better since it runs at a full point leaner than before at steady speeds.

I won't be around today to respond to this thread (maybe a little tonight), but then I am out of town for a couple of weeks. Maybe I'll try a little more tweaking then, but I think I probably have it as good as it's going to get.

Demick

Fascinating stuff! I started down this road several years ago with an O2 sensor, but didn't give it near the time and thought that you'd applied. My experiments totally failed, whereas you've made great progress.

A couple of questions:

- What do you think would be a more ideal mixture curve?

- What tuning effects does the CO2 dial on the computer box have on all this?

Andrew

The dial on the box only adjusts mixture at idle, so it shouldn't really have any effect on those curves. That was great info! I've really got to buy one of those wide band O2 sensors and fiddle with my injection a bit.

Add instruments to the car and monitor CHT at the same time, add that into your spreadsheet. If you reaaly want to see whats going on do the same with EGT and then start playing with timing advance and see the changes.

I did this a few years ago when I first got the dyno and those numbers helped me attain my stand with the stock D jet arrangement- it was really scary what such a small adjustment could do!

What you are doing is really cool.

Andrew:

I think that my ideal curve would look something like what is below. Of course, this is just based on my limited knowledge of A/F ratios. The actual numbers might want to change a bit (like maybe the max A/F ratio would want to be 14:1, or maybe even 15:1, I don' t know), but I think the general shape would want to look like this.

As Mike said, the knob on the ECU only affects idle, so it has no effect on this curve. The knob changes the A/F ratio by about 1.5 points total. I find that my car prefers to idle very rich (about 12:1). By 13:1, the idle becomes erradic and the lean idle hunt begins. I'm not sure why it likes to idle so rich.

Jake - I have a CHT sensor, but it is mounted at the F.I. temperature sensor location - which I understand from you is not nearly as good as behind the spark plug. An EGT sensor would be very interesting though and probably be an excellent way to tune the engine. I wish I had bought your cam instead of the #73. But I had already bought it before I heard about yours.

Yes, the adjustments to the MPS are very sensitive. A 1/8 turn of the inner screw is sometimes enough to change the mixture by a full point! It is much more sensitive that I originally thought.

Demick

I gotta dig out my old dyno A/F graphs...you are going too lean at partload at 14to1. Factory specs were 13.7to1 so aim for that.
I totally agree that a small turn of the inner screw can result in big changes.

Geoff

Demick,

have you thought of taking the tests any further and datalog the throttle opening/air temperature, etc, etc with the LM-1??

Ok Demick, here's a A/F dyno chart on my 2.1l with a adjusted MPS at WOT.
I got it to be fairly flat with careful adjustment at a dyno shop.
I have been reading what you are doing and you should be able to flatten out that curve to be without that lean spike.

Geoff

Just got back from 12 days in Maui. So time to catch up.

Mike: No, I haven't planned on any further datalogging. Very little with d-jet is adjustable anyway, so I would really only be able to gather information - and not be able to do anything about it.

Geoff: Thanks for the graph - but your graph and mine are not comparable. Your graph is at WOT only. It's easy to get an A/F ratio at ~12.5:1 at WOT. What I was working on was the rest of the throttle range. There is no way that I know of to flatten out the curve that I showed (without taking the MPS apart and making mechanical changes to it).

Besides, while a flat curve across the rpm range as you showed is desirable, a flat curve across the load range as I showed would not really be desirable. I showed what sort of curve I think would be ideal a few posts back. Unfortunately, I don't think there is much way of getting there with d-jet.

Demick