Wide band O2 sensor on D-jet

[vandor]

A few months ago I bought a wide band oxygen sensor with a gauge to monitor how the D-jetronic FI is doing in my car. Previously I had adjusted the main mixture adjustment by leaning it out each time I drove it, and when it began to run poorly I richened it by 2 steps (one step for me was 1/6th of a turn). The car drove fine, but I had no idea where the mixture actually was.
I installed the wideband sensor, taped the gauge to the wiper and drove the car like that for a few weeks while I tinkered with the mixtures (both idle and main).
The first discovery was that the idle mixture screw did nothing. I installed the spare ECU, which had a working idle mix screw, and after adjusting it to ~14.5 the "low idle when hot" problem vanished. With the other ECU the mixture was off the peg lean at hot idle.
The 'main' adjustment was leaner than I had anticipated, with the gauge often showing mixtures in the 15s during cruising. I ended up richening it slightly.
I also learned that it is impossible to adjust it to be ideal in all conditions, as the mixture gets richer (13s) at light throttle and leaner (15s) with half throttle. It does go back to a rich condition close to full throttle.

I can see how a programmable FI could give gains in economy and output, but the stock system works well enough for me.

BTW, the injectors were cleaned 3 years ago when I rebuilt the engine, so their spray pattern should still be good.

[Graham]

Originally Posted by vandor

A few months ago I bought a wide band oxygen sensor with a gauge to monitor how the D-jetronic FI is doing in my car. Previously I had adjusted the main mixture adjustment by leaning it out each time I drove it, and when it began to run poorly I richened it by 2 steps (one step for me was 1/6th of a turn). The car drove fine, but I had no idea where the mixture actually was.
I installed the wideband sensor, taped the gauge to the wiper and drove the car like that for a few weeks while I tinkered with the mixtures (both idle and main).
The first discovery was that the idle mixture screw did nothing. I installed the spare ECU, which had a working idle mix screw, and after adjusting it to ~14.5 the "low idle when hot" problem vanished. With the other ECU the mixture was off the peg lean at hot idle.
The 'main' adjustment was leaner than I had anticipated, with the gauge often showing mixtures in the 15s during cruising. I ended up richening it slightly.
I also learned that it is impossible to adjust it to be ideal in all conditions, as the mixture gets richer (13s) at light throttle and leaner (15s) with half throttle. It does go back to a rich condition close to full throttle.

I can see how a programmable FI could give gains in economy and output, but the stock system works well enough for me.

BTW, the injectors were cleaned 3 years ago when I rebuilt the engine, so their spray pattern should still be good.

Glad to see that another djet owner making use of a wideband AFR gauge!

Can't imagine how mixture could be adjusted without some type of exhaust analyzers. You are right that the mixture does vary quite a bit. I see the rich startup when cold, then afr gradually rises to about 14. During hot starts, I see very lean numbers (but no trouble starting) and AFR gradually come back to about 14. I set idle on rich side.

BTW, I have my 350SL gauge permanently installed (where ash tray used to be) and with sensor in exhaust just before first muffler.

[alabbasi]

I use a tailpipe co meter for tuning my non catalyzed cars. I have been looking at getting an AFR gauge for a BMW 745i so that I can set it up properly.

[Mike D]

In the 15's is about right for a light "cruise" situation. You'll find the car idles best in the upper 13's but might not pass the emission standards for whichever state you register your car.

14.1 will give you a good compromise between emission standards and smooth idle.

[Graham]

Originally Posted by alabbasi

I use a tailpipe co meter for tuning my non catalyzed cars.

I never found a CO meter that worked and was priced for DIYers. What type of CO meter do you have?

One of the problems of measuring at tailpipe, is that it is hard to do measurements under load i.e. while driving car.

I am more than happy with my full time wideband AFR read out (AEM meter). Cost about $200 plus cost of welding bung to exhaust. Here it is at 1000rpm (I was checking AFR during a warm up from cold) https://dl.dropboxusercontent.com/u/54783344/sl%20engine%20002.AVI (may take a little time to load)

(At start up mixture is rich)

[alabbasi]

Graham, I own a couple. The latest purchase is a 5 gas analyzer. It probably cost several thousand dollars when it was new but I got it for a couple of hundred.



This old emissions stuff is now obsolete with OBD2. My local testing station still has a dyno to test 92-96 cars. They tell me that they may get 1-2 cars / month and every time it breaks down, it costs them several thousand dollars. They can't wait to be rid of it.

[Graham]

Originally Posted by alabbasi

Graham, I own a couple. The latest purchase is a 5 gas analyzer. It probably cost several thousand dollars when it was new but I got it for a couple of hundred.

Do you need a cylinder or two of calibration gas for that unit?

[alabbasi]

For the 5 gas yes, it came with a cylinder. The CO meter does not.

[vandor]

> In the 15's is about right for a light "cruise" situation.

It's in the 1314s in light cruise. It goes into 15s when I lightly accelerate.

>You'll find the car idles best in the upper 13's

Dunno, I tried to go richer but it did not seem to make much difference, that is why I set closer to stoichiometric.

[Mike D]

If it ain't broke........