Hi - I own a 1998 RR with a 4.6L that I just completed an engine swap in with another 4.6L. Since dropping it it I had difficulty starting it and when it did with help of the accelerator it shut down quickly. Doing some research I learned that the MAF can be the cause of this so disconnected this after checking for spark and correct fuel pressure in the rail. Engine started with help on the accelerator and would stay running but very rough. It loves to accelerate and fires evenly during that but holding a given RPM it seems to be missing. I had used all the original sensors and beforehand had no such issues. I started to probe the harness after looking at manuals and made an interesting discovery. The Sensor ground which goes to many of the engine sensors and is typically a Red/Black wire coming from the ECU(Connector C507 Pin 36 is sitting at 5V with respect to Chassis ground. This is with the battery connected, the ignition on and all teh sensors connected. I then removed the connectors from the ECU and confirmed the 4 grounds that are supposed to be giving the ECU it's chassis ground - all good. I then checked for a connection between Pin 36 and some of the sensor grounds (with Red/Black coloring) and got good less than 1 Ohm) readings. It seems that the ECU is supposed to supply a ground to the sensors but is not doing so and one of the sensors is passing through a 5V reference through a coil or something. Can anyone confirm that indeed the ECU does supply a ground and whether that ground is there all the time or only when ignition is activated or at some point later when you got a closed loop condition. I have looked high and low for any information regarding this and all I have found is in the gemsbypoole.pdf where is states that this pin should be at 0V. Mine is not. I have a modified ECU that bypassed the imnmobilizer but this has been converted for years. I don't see how that would have a connection in this case. Any help appreciated... I also cannot find schematics for the GEMS8 ECU that would help me confirm this.

I guess a second possibility is that the Harness is supposed to furnish a chassis ground to the ECU at pin 36 and the sensors. I do not want to try that unless someone confirms for fear of damaging the ECU.

With the MAF disconnected it will start and idle but won't rev unless you constantly pump the throttle and you can get the revs up eventually as you have found. The GEMS SID simply says that pin 36 on C507 is sensor common but not what voltage it should have on it and as well as the MAF it is also common to the fuel temperature sensor, intake air temperature sensor, cam position sensor and throttle position sensor.

I had always assumed there would be a 5V reference voltage somewhere so, in the interests of curiosity, I've just checked mine. As I run a single point LPG system, I have a stand-alone fuel system that doesn't need a MAF (or most of the other sensors for that matter). Started her up on LPG, disconnected the plug to the MAF and checked the voltage on the Red/Black wire. Result was 0.04V with respect to battery negative so ground to all intents and purposes. After switching off and reconnecting everything I thought that maybe I should have checked with the ignition on but engine not running but I can try that if you think it might be a benefit. It might be worth checking the voltage on the other sensors to see if there is damage to the loom? When you changed the engine did the replacement come with the loom on it so it was just a case of plugging it in or did you swap the loom over? Did you remember to connect the main engine ground down underneath the alternator?

do you have access to a nano? Did you reset adaptive values yet? Nano may reveal some odd readings on the sensors, if you post the readings up

Thanks to both for the responses - Gilbertd - I am using the original harness from the old engine. I really appreciate you taking that reading - it is an indicator of something I feared. There is a separate 5V reference for the TPS on the ECU and mine is working. But I do not get the 0V reading off the Red/Black wire at any of the sensors you mentioned or at pin 36 of the ECU and as I feared I should be. One reading that would be very helpful to me would be if you pulled the red connector off your ECU then set switch to IGN (don't start) and take a reading from pin 36 to GND with a voltmeter - that would definitively tell me if the ECU is providing the ground. If you do, then turn off the ignition and tell me if you still get the ground reference or not. I want to understand if it is being switched within the ECU or a permanent Ground if you will. If, on the other hand you get a non-0V reading then try taking a reading from the harness pin 36 right there on the red connector to Chassis Ground. If you get 0 Ohms then the ground is being provided by the loom and I will need to check mine. However, my research thus far shows me that the ECU is providing that under normal conditions.

romanrob - I do not have a nano yet but am looking for a deal on ebay for one to come along...

I hope you get to the bottom of this quickly.

It’s probably fair enough to watch eBay for a Nanocom, but don’t waste too much time looking. Everyone who has one really likes it and probably won’t be looking to sell it. And the owner is likely to be a fairly diehard Land Rover guy, not planning getting out of the LR scene. As costly as they are, I’ve never heard of anyone regretting buying one.

I'm busy tomorrow afternoon but will have some spare time in the morning so I'll do some more checks for you and post them up.

So the results are in. With the red connector unplugged and ignition off, I get 0.00V at pin 36 wrt battery negative and a varying resistance that stabilised at around 16 Ohms which would suggest a cap on it. With ignition on I get no Check Engine light showing the ECU is not powered, I still get 0.00V but this time the varying resistance stabilised at around 1.7 kOhms.

However, when I put it back together I'm not sure the black connector C505, which I'd had to remove to get the red one out, was fully home so that may be a bit of a red herring. By then my fingers had lost all feeling due to the 2 degree temperature outside so I didn't feel inclined to take it apart again......

Gilbertd - thanks again for your willingness to take live measurements - sometimes the manuals just come up a bit short - sounds like I owe you some pocket warmers :). So if I understand you correctly, you had both the small black and red connector out of the ECU - that would have removed both all official ground connections to the car (which are all in the smaller black connector) as well as the power to the ECU. I find it interesting that you still had only 16 ohms between pic 36 and chassis ground - it must have been some connection through the one remaining connected black connector on the ECU. I will try to repeat what you did on mine and see if I get similar results. Did you by chance take an ohm reading of pin 36 on the red connector (while disconnected from the ECU) to Chassis ground to see if there is anything in the harness at all? I hope to get my last gasket in the next day and then am crossing my fingers to have this all buttoned up by the weekend - we get Monday off for a holiday so an extra day to try to wrap this up! By the way, I am impressed with the folks on this Forum and their willingness to dig into the inner workings as it were of their cars... diehards as they say - I do plan on keeping my P38 for many a year as well...

No, the small black one was still attached, it was the large black one that I had to unplug as the wiring for that goes over the top of the red, centre, one and, like I say, I am not sure it was pushed fully home when I did the checks.

I did consider checking resistance to ground on the Red/Black wire while it wasn't connected to the ECU but thought that without unplugging all the other sensors, all I would be seeing would be the combined parallel resistances of the other sensors. All you would be seeing would be the static resistances of the cam sensor, MAF, IAT sensor, TPS, coolant temperature sensor and fuel temperature sensor. Now some of those may well have a 5V reference on the opposite side (TPS definitely does as it shows around 0.6V at closed throttle and near 5V at fully open) so if one has gone short circuit, that would show you 5V on the Red/Black. It might be worth checking and unplugging each sensor at a time to see if the 5V disappears.

Hi Gilbertd - thanks for the clarification on which plug you had removed. I just went and measured what you did and here are my results. Key off: Pin 36 to GND - Infinity (DC Volts: 0.006). IGN On: Pin 36 to GND - infinity (DC Volts: 0.006). As for the harness check - I was trying to ascertain whether the Red/Black wire was getting it's ground through the harness or one of the sensors - I checked my pin 36 on the red connector to GND and got Infinity. So from what I can ascertain, you are getting between 16ohms and 1.7k on Pin 36 of the ECU on the red connector and I am getting infinity. By the way I am using a fluke auto ranging meter in case that might cause a difference in readings. This tells me that whatever used to drive that pin in the ECU is no longer doing so which is a real mystery given that the ECU never left it's home during the engine swap and the harness as well as the sensors are all the original ones. It does mean that all my sensors have no reference ground which explains all the bizarre readings I have been getting. I am guessing no one else on this forum has had this happen or has any more insight on what Pin36 (Sensor ground) is supposed to do and why the system does not simply use chassis ground directly.

It shouldn't make a difference using an auto ranging meter. They can get confused at times, although I wouldn't expect a Fluke to, but better than the one I use for car stuff which has manual range settings. If I set it on 200 Ohms range and try to measure something that is greater than that, it just shows a 1. But it shows a 1 if it is open circuit too.....

I'm wondering if you could disconnect the Red/Black from the ECU and simply connect it to ground and see if that works?

Waiting on the front cover gasket - once that arrives (today I think) will get the engine back together and try that....

Hi all - wanted to bring closure to this issue. I was able to get the last of my miscellaneas items addressed on the engine and took the 'Sensor Ground' (Red/Black) wire cut it at the ECU and connected it to Chassis Ground. I then plugged in all the plugs into the ECU and placed the Key to the Ignition position. I placed a 100Ohm resistor between the ECU ground pin and Chassis ground and measured to see if there was any current being pulled through that pin. Since my pin seemed to be dead I was not surprised that there was no current draw at all. That pin seems to be severed from all internal circuitry in the ECU at this point. I then started the engine and had immediate success - the rebuild engine settled into a high idle and then settled down to a normal idle. IWhile it was running I disconnected the ground to the sensors and the engine immediately stalled. I have not run the car for about 30 min to let it smoke off from the rebuild and so far there has not been a check engine (MIL) light so all seems to be well. The million dollar question is still whether the ECU measures current draw through that sensor ground pin in order to perform an OBD2 function such as a failing sensor - I believe I have lost that capability at this point but I am happy to be back up and running. Thanks to all who assisted with the troubleshooting and I hope this allows someone else to keep their RR running a bit longer without an ECU replacement for this issue.

csduetto: I believe you have already eluded to this but (having poked around inside various ECUs) it is always worth remembering that they usually tend to supply 'Electronic' Earth rather than Chassis ground anyway.
An ECU 'Electronic Earth' is often 'stabilised' via caps. etc of course.

-What usually tends to cause these Eth. 'outputs' to fail is -depending on the sensor- 12V/5V connected directly to them -ie without the (resistive) sensor- for a period of time: Similarly the 12V/5V 'regulated' supplies from an ECU can also be burned out by a persistent 'hard earth' - but some ECUs provide a 'resistive' Electronic Eth. too.

-Another unusual 'quirk' of ECUs is that some provide "Earth Pulses" too, for instance fuel injectors with 12V ('permanently') on one side...

Back on the 'million dollar question' however ECUs - rather that measuring a 'current draw' (or lack of it) - usually monitor the lack of sensor output (and/or any intermittent / persistently off-map / etc values of that sensor voltage/signal (depending on the particular sensor/type under consideration, obviously)....