GEMS Mixture adjustment ?

[jagwit]

Someone brought a Disco1 fitted with a GEMS 4.6 to me, wanting the vehicle to be fitted with Megasquirt as I've done to mine. However, I did a "health check" on the engine and found the engine to be running consistently LEAN by about 2.5 AFR points! (Eg instead of running AFR 13 where it should, it was running 15.5 etc)

I have found the fuel pressure to be too low (2.1 bar) where it should be in the region 2.4 - 2.6 but I'm hoping that its possible to make SOME mixture adjustment via the ECU.

Can anyone tell me if it is possible to make any mixture adjustments via the GEMS ECU?

Thanks

Philip

[vougese39]

as the vehicle is a closed loop emission it is preordaned to how it runs by the ecu map there is no manual adjustment of fuel as the ecu is a generic it adpts to the increase in air flow to richen

[jagwit]

This engine did not and does not have O2 sensors. It is a low compression engine (as LR supplied in South AFrica due to our low quality fuel) and therefore runs in open loop.

[Ally V8]

I think I'd be looking at a low output from the airflow meter rather than trying to adjust the mixture first off.I think Testbook does allow trimming of mixture strength for open loop variants,but I've never used it as all the UK spec cars run Oxy sensors.If you can get some live data up,you should be seeing an airflow of 22-25Kg/hr at idle,engine hot all loads off.I'll have a look at the Testbook menu's when I get to work.

[jagwit]

Thanks Ally V8!

I have in the mean time realised that there is nothing wrong with the fuel pressure. I measured on the "tyre valve" on the fuel rail. This is AFTER the FPR. Before the FPR, pressure is close to spec evidenced by the pressure gauge needle rising when manifold pressure drops as the throttle is opened.

Do you perhaps have a graph of how the MAF translates airflow into an output voltage?

[Ally V8]

Sorry to take so long to reply,I dont have a graph of AFM voltage - Testbook just reads it as KG/Hr,which is easy to understand and there is no need to see the actual voltage.Testbook does have a menu that allows open loop tuning,it requests that an exhaust gas analyser be used etc.I checked this out out a 4.0ltr yesterday,but could not go beyond the initial menu as it was a UK spec closed loop car.

If you can get the car on Testbook,( I assume you have checked for inlet manifold leaks etc)it would be interesting to see if it will respond to mixture adjustment.

[jagwit]

Hi Ally V8

The owner is very busy at present, so I expect we will only get round to this some time next week if not later. Will provide feedback once its done.

BTW, I tried another used MAF last night and this made the engine run even more lean. This now categorically confirms to me that MAFs can go "off".

[vougese39]

are you sure it is on gems not hotwire? how did they get round the lambda reading which the ecu is programed to need to help mixture setting?

[jagwit]

are you sure it is on gems not hotwire? how did they get round the lambda reading which the ecu is programed to need to help mixture setting?

Unleaded fuel (a requirement for lambda sensors) was only "recently" introduced into South Africa. LR worked around this problem by supplying low compression, open loop engines in Disco and Range Rover to this market. There are a few high compression versions (also open loop) D1's running around - my nephew has one.

GEMS also uses a hotwire MAF I believe. Anyway this engine and its management system was pulled from a P38 and installed into this Disco1.

[FridgeFreezer]

A few questions:

1 - How are you measuing the AFR?

2 - Is 15.5 too lean relative to other 4.6 RR's you've measured or other vehicles?

3 - You state you measured the fuel pressure "before" and "after" the FPR. As far as I'm aware, most RV8's the FPR regulates the return to the tank, so the measurement before it should be correct and the measurement after it should be a low-pressure return back to the tank. I know later ones had a different setup, but I've not fiddled with one so don't know what the differences are.

4 - Why not just crack on and MegaSquirt it?

Low output from the AFM can also mean there's a vacuum leak downstream of it.

[jagwit]

A few questions:

1 - How are you measuing the AFR?

2 - Is 15.5 too lean relative to other 4.6 RR's you've measured or other vehicles?

3 - You state you measured the fuel pressure "before" and "after" the FPR. As far as I'm aware, most RV8's the FPR regulates the return to the tank, so the measurement before it should be correct and the measurement after it should be a low-pressure return back to the tank. I know later ones had a different setup, but I've not fiddled with one so don't know what the differences are.

4 - Why not just crack on and MegaSquirt it?

5 - Low output from the AFM can also mean there's a vacuum leak downstream of it.

1. LC-1 up the tailpipe;

2. 15.5 is definitely too lean for certain operating conditions like full throttle and idle but is perfect for others. I operate my 4.6 between 12.5 at full throttle (all rpm) and 16.5 at very light throttle (only certain rpm range);

3. That's right, the FPR regulates the amount of fuel being allowed back to the tank so that the pressure in the rail is around 2.1 bar at idle. When you remove the vacuum pipe off the FPR, pressure rises to about 2.6bar. Pressure also rises briefly when you stomp on the gas.

4. I have no time at present and I'm too busy finishing off my V12 Cobra and converting my V12 E-type to EFI (with Megasquirt). Besides, the solution is "so close" by replacing the MAF (or tuning GEMS via PC = the cheapest solution).

5. Correct, but not the case here. Did you read above that I tried another used MAF and that was even leaner?