Checking ECU Fuel Mixture

I read about the voltage check using a diagnostic port by the ECU to help with determining fuel mixture at each bank.

I think I found the diagnostic port, but the wires don’t match. My guess is what I’ve been reading is Lucas related and different than Marelli.

Before I start digging through the wiring diagrams which is a bit of a chore for me, being electrically impaired, figured I’d create a topic for some quick guidance, comments, etc.


^^^ I think this is the diagnostic port to the ECU, but not entirely sure

The colors you have match my diagrams except one: your green/white should be green/brown…according to the diagrams I have.

I’d just put that down to an error in the diagrams and move forward.

Cheers
DD

1 Like

For clarity on my 89 Marelli,

B - ground
KB - battery
GW - A bank ECU adjustment amount
GR - B bank ECU adjustment amount

During checking these, the A bank was rich at 2.9v and B bank lean at 2.0v - measure with multimeter and not a fancy oscilloscope.

Due to kncoking of the 2 A bank injector, I’m thinking that’s why its rich. O2 sensors are new.

Once I replace/rebuild that injector, I’ll see where the readings are.

A pic is worth 1000 words…

1 Like

I might be wrong here, but being a Marelli car it would have a 16CU ECU, no?
You would need a 47uF capacitor across the leads to get correct readings.
Details at AJ6 Engineering website.
Also, the readings should oscillate, if not it means you are on open loop.

In open loop, my 16CU (1988 Lucas) reads 1.9V. So in closed loop you want them close to this.

FWIW, I’ve given up fiddling with this, and just leave ECU fuel mixture at 0 clicks, centered. I get different readings totally dependent on rpm and engine temperature. And I’ve noticed no difference in dialing it one way or another. Roger Bywater told me the ECU always corrects in closed loop, so if you set ECU richer, the O2 sensors will see this and correct leaner. I believe the only reason to set close to 1.9V is so there’s a smooth transition from closed to open loop, as when flooring it or shifting into P, and open to closed, when shifting out of park. This still confuses me, so I just leave it at 0. :slight_smile:

Just tossing something out…

Long-ish story cut short…

I am knowingly and temporarily running in open loop at all times. The open loop fueling map must be pretty agreeable to the engine, as it behaves very nicely…although probably not optimally.

Cheers
DD

Doug,

Running at open loop would continuously over fuel the car. If you’re ok with that, then I guess that’s that.

I’ve read that over the long haul that’s bad for the engine, something about bore wash, or something like that?

I would imagine your catalytic converter would be prematurely saturated as well?

That’s my understanding as well. But, apparently, not to the point where the engine objects in any way

I’m Ok with whatever the engine likes best…providing there’s not too great a price to pay :slight_smile: In any case, it’s a temporary thing.

That’s a thing, yes. But that would be gross over-fueling…think old carby engine with an automatic choke stuck closed. The mixture is so rich that it can’t possibly burn completely. The unburned gasoline washes down the cylinders.

I don’t have converters so not worry there.

To be clear I’m not advocating running in open loop as a practice. I was trying to point out that the difference between open and close loop fueling isn’t enough to cause any pronounced or, perhaps, even any discernable problems.

The feedback system trims the fuel mixture. The primary mixture controller remains the baro sensor in the ECU.

Cheers
DD

I am curious as to what the fuel economy is, if no significant change, probably not over fueling. I have a Rover V8 powered MGB that has the Lucas 14CUX fuel injection system, that is similar to the 16CU. I have installed oxygen sensors but no catalysts. I seem to get a stronger running engine when I run a fixed fuel map, but smoother and 10% better economy when running a catalyst map where the oxygen sensors are read. I have read that the Rover V8 makes more power and idles better when richer than 14.7 AFR, but can cruise much leaner. Problem is that the feedback locks it at 14.7. Jaguar may be the same.
If you just unplug the oxygen sensors on the V12, the ECM thinks it is running at 14.7, because there is a voltage divider with high value resistance that provides a mid point voltage value.

If you unplug the 02 sensors, the car will stay in open loop. It follows the preprogrammed fueling map.

1 Like

We will have to agree to disagree. Put a voltmeter on the sensor wire and unplug sensor; you will see what I mean. 16CU doesn’t know and provides middle signal. There is no monitor looking for signal change with purge, because ECM doesn’t control it.

I was wondering as well.

As it happens, I just filled-up today and got 14.5 mpg (USA) after 130 miles of 50-50 city/highway driving. IOW, the same as before under similar conditions.

Most my driving is city traffic so maybe I’ll check it again and see if there is any change. In any case, there is, at minimum, no tangible over-fueling.

Here, briefly, is the back story to my experiment.

For several years I have been running no cats, no oxygen sensors, and a non-feedback ECU. So far, so good. However, I’ve had some driveabilty issues crop up that defied the “usual” remedies. For giggles I installed my spare ECU, which is a USA/Lamda version. Since I have no oxygen sensors it cannot go into closed loop because there is simply no O2 signal. That’s how I understand it, at least.

Anyhow, the ECU swap solved (or masked?) my drivibility issues. And, after a few days, does not appear to be causing any problems.

Now I must decide on where to go from here. Add O2 sensors back in and continue using the feedback ECU. Or get the (apparently faulty) non-feedback ECU repaired. Or, do nothing.

If I send the faulty ECU for repair (weeks or even months) I’m pretty confident that using my spare ECU won’t result in anything bad happening.

Cheers
DD

I don’t follow, sorry.

And the “with purge” has me confused.

If the sensors are not there (unplugged or simply not present) how can closed-loop operation occur?

Cheers
DD

Sounds to be the same as my Rover V8 MGB; am running catalyst fuel map with oxygen sensors but no cats. Running fine. Hopefully you have bungs in the downpipes for sensors; I think dealing with the exhaust flanges is one of the more miserable jobs on an XJ-S. Wiring shouldn’t be hard by making an overlay harness, use heated oxygen sensors for sure.

It’s a pretty crude system. My understanding is that when a sensor signal is faulty, the ECU just continues to believe it, or as in the case of the oxygen sensor, has provision for a middle value if no input is present. Coolant temperature and runtime determine if open or closed loop and trim accordingly. There is no fault logging; that came later with the 26CU which added a “limp in” mode if faults were logged.
The 6CU/16CU use fixed fueling maps dependent on coolant temp and manifold vacuum with other sensors providing trim; 16CU added oxygen sensor trim when coolant temp was high enough. It’s not like a modern vehicle that has to see changing oxygen sensor signals to believe their signal. The easiest way for me to see which is which is to look at the purge system; if ECM controlled, then it is looking for an O2 change when the purge valve opens and will log a fault if not (fault logging type).

So perhaps it can be said that temp and run time have put the ECU in closed-loop mode but, lacking oxygen sensor input, no closed-looping is actually taking place?

Cheers
DD

No bungs so, yes, I’d have to drop the downpipes. But, sans converters, it isn’t quite so bad

All the wiring is there, just not in use. It’s taped-off. When I did the V12 converson I installed the entire harness out of the donor car, intact.

Cheers
DD

Well, can’t really operate closed loop if there are no oxygen sensors. But is running on all the other sensor inputs, so pseudo closed loop.

Yes, no, maybe.
The base fuel map on the ECU is adjustable, and if it’s properly adjusted the engine will run very close to stochiometric on open loop. If not it could be either rich either lean.

Coorrect, and that midle value is the Base Fuel Map.
And imy understanding is that Temp and Vacuum trims work the same in open loop as on closed loop.

2 Likes

Aristides said what I was trying to say, but better.

1 Like