Tracing Range Rover P38 MAF Output Problem

I am new to posting on this forum but have spent many hours reading other people’s threads and gaining knowledge from it. My problem is with a 1997 P38 with a GEMS engine that runs like a bag of nuts and bolts. I am an electronics engineer and have test equipment and a P Scan Diagnostic tool, so I am able to approach the problems in a logical way (when the whisky clears from my brain). The output of the P Scan shows:-

Live update interval 602 ms

Engine speed 633 rpm

Fuel System 1 Status Closed Loop

Fuel System 1 Status Closed Loop

Calculated Load 34%

Engine Temperature 99 deg C

Short Term Fuel Trim 1 -16%

Long Term Fuel Trim 1 11%

Short Term Fuel Trim 2 -17%

Long Term Fuel Trim 2 11%

Error Status 0 MIL OFF

Vehicle Speed 0 km/hr

Ignition Timing Advance 24 deg

Air Intake Temperature 27 deg

Air Flow Rate 0.04 g/s

Throttle Position 11%

Commanded Air Status 4

Oxygen Sensor Location 17

02 Sensor Output Bank 1 0.645

02 Sensor Output Bank 2 0.805

The engine ECU/ECM has been recently changed (I have tried three different ones) and has done about 50 miles so the long term fuel trims are probably reflective of the previous engine that it was used on. The air temperature and O2 sensors were changed last year. My main approach is to get the Air Flow Rate right (should be about 5.5 g/s) before looking at anything else. Much is written about the MAF sensor on this forum but the MAF does not appear to be the problem here. I have now accumulated 4 MAF sensors and the one currently on the car gives an output of 1.35 Volts. Rave states on page 351 that the output should be 0 - 5 Volts (1.4 V at idle) so there is nothing wrong with that MAF. Injecting a regulated voltage into the ECU/ECM and reading the P Scan gives:-

0 Volts dc = 0 g/s

1 Volts dc = 0.03 g/s

2 Volts dc = 0.16 g/s

3 Volts dc = 0.58 g/s

4 Volts dc = 1.38 g/s

5 Volts dc = 2.19 g/s

So, the full output of 5 Volts from the MAF could not reach the expected 5.5 g/s at idle and therefore eliminates the MAF as the problem. Checking the wiring harness from the MAF to the ECU/ECM shows that it is continuous (less than 2 Ohms) and there is no leakage. Continuity checks were made with an old ECU/ECM that was opened so that measurements could be taken at the PCB and they were again less than 2 Ohms showing that the plugs were mating properly.

It appears that something other than the MAF can affect the reading of Air Flow and it is probably related to the adaptive compensation.

My question is this. What other factors/inputs can affect the Air Flow Reading from the ECU/ECM? I know that other things are suspicious but I want to get the air flow right first.

ElectronJockey said:

My question is this. What other factors/inputs can affect the Air Flow Reading from the ECU/ECM? I know that other things are suspicious but I want to get the air flow right first.

Click to expand...

The top item is an air leak between the MAF and the inlet valves.
You need to reset the fuel trims before you do anything else

The car isn't obd compliant, so don't expect a generic reader to give accurate results

The PScan does read the ECM data and I have recorded good Air Flow Readings from it previously. It has a facility specifically for the GEMS engine. However, it looks as though I no longer have that problem to sort out. The old lady has now probably blown the head gasket and according to my garage and their prices she is not worth repairing. She has an oil leak in the front axle which is an expensive repair and the paintwork on the roof has gone very pale. So, three expensive jobs on an old car. Shame, I have had the car for 5 years and I love the old thing.

ElectronJockey said:

The PScan does read the ECM data and I have recorded good Air Flow Readings from it previously. It has a facility specifically for the GEMS engine. However, it looks as though I no longer have that problem to sort out. The old lady has now probably blown the head gasket and according to my garage and their prices she is not worth repairing. She has an oil leak in the front axle which is an expensive repair and the paintwork on the roof has gone very pale. So, three expensive jobs on an old car. Shame, I have had the car for 5 years and I love the old thing.

Click to expand...

Front axle leak could just be the nut securing the diff to the housing, they regularly come loose on one of mine, the axle seals are not a hard DIY job either.
As for the paint on the roof, no one can see it
I assume it's a V8?

Yes, it is a V8 4.6 ltr. I understand from the garage that part of the front axle has rusted badly and it does leak oil. They have put some silicon sealant on it a couple of times but it is starting to leak again. They quoted £1400 to have it re-worked some time ago and it is probably not worth not much more than that in good running condition. True, nobody much looks at the roof but I see it when it is parked outside and I look out of the bedroom window. Thank you for the comments Datatek it is a hard decision to scrap the old lady and any discussion is helpful. Getting back to my original query, I would love to know why the air flow reads so low on the P Scan. With an input of 5 Volts to the ECM the P Scan reads 2.19 g/s. I calculate that for it to read 5.5 g/s (the rate at tick over) the ECM input would have to be 12 Volts and for 27 g/s (the rate at 3000 rpm) the input would have to be nearer to 60 Volts! Looks like I need another car and s Nanocom!!

ElectronJockey said:

Yes, it is a V8 4.6 ltr. I understand from the garage that part of the front axle has rusted badly and it does leak oil. They have put some silicon sealant on it a couple of times but it is starting to leak again. They quoted £1400 to have it re-worked some time ago and it is probably not worth not much more than that in good running condition. True, nobody much looks at the roof but I see it when it is parked outside and I look out of the bedroom window. Thank you for the comments Datatek it is a hard decision to scrap the old lady and any discussion is helpful. Getting back to my original query, I would love to know why the air flow reads so low on the P Scan. With an input of 5 Volts to the ECM the P Scan reads 2.19 g/s. I calculate that for it to read 5.5 g/s (the rate at tick over) the ECM input would have to be 12 Volts and for 27 g/s (the rate at 3000 rpm) the input would have to be nearer to 60 Volts! Looks like I need another car and s Nanocom!!

Click to expand...

A replacement axle from a breaker is not that expensive. Never heard of one rusting through

£1400 to do front axle seals is quite excessive. All it takes is to pull the Hub from the swivel complete with CV & driveshaft. Replace the seal & re-install.

Head gaskets are a weekend job, and easily doable for home mechanic.

pwood999 said:

£1400 to do front axle seals is quite excessive. All it takes is to pull the Hub from the swivel complete with CV & driveshaft. Replace the seal & re-install.

Head gaskets are a weekend job, and easily doable for home mechanic.

Click to expand...

Think it's the diff pan rotted out that's the issue.?

pwood999 said:

Head gaskets are a weekend job, and easily doable for home mechanic.

Click to expand...

Easy to do, if that's the problem.
If i ever had to take the heads off a v8 again i think I'd go for studs as an upgrade, less likely to crack the block

kermit_rr said:

Easy to do, if that's the problem.
If i ever had to take the heads off a v8 again i think I'd go for studs as an upgrade, less likely to crack the block

Click to expand...

+1 on studs. Double the cost of the stretch bolts but so easy to use (and reuse if necessary). Felt so secure when putting Otto's heads back on.

If you do a head gasket, don't just do stage 1, stage 2, stage 3, leave between stage 1 and stage 2 overnight for everything to settle. No need to wait on stage 3 though - I did and it made no difference.

Biggest pain on the head gaskets was either getting the heatshields off (needed contortionist certificate and dremel with a selection of die grinder bits) or the middle steering link, which was crudded solid at the steering box end. Eventually got that off, cleaned everything up well and now it slips easily on and off the splines with no stress.

Roof - if it is just faded paintwork, see if somewhere that does vinyls / car wrapping could do a covering for it to restore / contrast the colour?

We had issues with Otto's MAF due to dirt buildup and similar. I sorted them by taking the MAF off, removing the filter screen, plugging the outlet ports in the sensor section with cotton wool, then standing it on end, intake upwards in a bowl and carefully pouring isopropyl alcohol into the intake tract. Left it to soak through . evaporate for an hour or so, removed the plugs and then sprayed a bit more isopropyl through. Back on the car and he was transformed.
Your MAF readings could just be down to build up of dirt on the elements?

Thank you all for the advice. I have laid the car up for a while and will not get rid of it if is at all possible. I see no point in spending lots of time or money on it as long as it has this weird air flow fault. The problem is not the MAF because I simulate the MAF by putting a known dc voltage into the ECU and the pscan indicates a very low air flow. So, the problem is not the MAF or air leaks or anything that affects the input voltage to the ECU, it has to be after the input to the ECU. The pscan has worked correctly before so the problem is most likely related to the ECU but without the schematic for the ECU and knowledge of its programming I cannot go further on that investigation. My first action (when I have the cash) is to buy a Nanocom and see if it agrees with the pscan on the air flow readings. Then to check all of the other parameters on the Nanocom to see if there is some other sensor, which is not showing up on the pscan, that is causing the problem. If I cannot fix this p38 then at least I will have a Nanocom ready for the next p38.

ElectronJockey said:

Thank you all for the advice. I have laid the car up for a while and will not get rid of it if is at all possible. I see no point in spending lots of time or money on it as long as it has this weird air flow fault. The problem is not the MAF because I simulate the MAF by putting a known dc voltage into the ECU and the pscan indicates a very low air flow. So, the problem is not the MAF or air leaks or anything that affects the input voltage to the ECU, it has to be after the input to the ECU. The pscan has worked correctly before so the problem is most likely related to the ECU but without the schematic for the ECU and knowledge of its programming I cannot go further on that investigation. My first action (when I have the cash) is to buy a Nanocom and see if it agrees with the pscan on the air flow readings. Then to check all of the other parameters on the Nanocom to see if there is some other sensor, which is not showing up on the pscan, that is causing the problem. If I cannot fix this p38 then at least I will have a Nanocom ready for the next p38.

Click to expand...

Is it petrol or diesel? I assume petrol as the diesel will run quite happily with the MAF unplugged.

It is petrol/lpg 4.6 ltr GEMS It will just about run with the MAF unplugged but it is likely to stall at junctions and is likely to cut out at any time making it rather unsafe to drive.

Datatek said:

Is it petrol or diesel? I assume petrol as the diesel will run quite happily with the MAF unplugged.

Click to expand...

1997 GEMS in post 1.

To echo Kermie, some sort of leak on the inlet side.