Rovergauge and fuel map woes.....
Discussion
Hi - I've been happily lurking on the fringes of the p/head forums, but I've hit a problem that has me completely flummoxed so I throw myself on your collective mercy.
1998 Chim 4.5, 39k miles, runs fine but is ferociously greedy on fuel. Every MOT since I've had her has failed on emissions but managed to squeeze through on the retest (sympathetic tester & lots of Cataclean). Based on documentation I inherited the last owner systematically changed out various sensor, including lambdas, so in hindsight I suspect that he had been chasing down the same problem. Anyway, last MOT failed again and I replaced MAF and got her nice and hot before retest and we sailed through. Since then curiosity has got the better of me and in anticipation of next MOT I've downloaded a copy of Rovergauge and made up a cable so I can investigate in a bit more detail.
So, first time up and Rover gauge tells me I'm running map 1, which Dr. Google tells me is an "Australia and rest of the world" non cat map, so unsurprisingly Rovergauge isn't showing any lambda trim data. I've checked the tune resistor which is 3890 Ohm, so should be European CAT map which is what I'd expect. I'm assuming that if Rovergauge is telling the truth the ECU is ignoring the lambda sensor inputs, which could explain the problem with exhaust emissions at test time and would possibly give rise to fuelling problems ?
Everything looks right, sensor and tune resistor wise, to my untrained eye, but for the life of me I can't figure out why the ECU is doing its own thing as regards the map. Any thoughts or suggestions gratefully received.
All this has taken a little bit of a back seat this weekend since the speedo died on Friday so I've spent a spirited few hours trying to get the dash top off - without doubt the worst job I've encountered so far.
MOT imminent, so speedo is top priority, but the mapping thing is bugging me.
Off for a bit of a lie down now....
1998 Chim 4.5, 39k miles, runs fine but is ferociously greedy on fuel. Every MOT since I've had her has failed on emissions but managed to squeeze through on the retest (sympathetic tester & lots of Cataclean). Based on documentation I inherited the last owner systematically changed out various sensor, including lambdas, so in hindsight I suspect that he had been chasing down the same problem. Anyway, last MOT failed again and I replaced MAF and got her nice and hot before retest and we sailed through. Since then curiosity has got the better of me and in anticipation of next MOT I've downloaded a copy of Rovergauge and made up a cable so I can investigate in a bit more detail.
So, first time up and Rover gauge tells me I'm running map 1, which Dr. Google tells me is an "Australia and rest of the world" non cat map, so unsurprisingly Rovergauge isn't showing any lambda trim data. I've checked the tune resistor which is 3890 Ohm, so should be European CAT map which is what I'd expect. I'm assuming that if Rovergauge is telling the truth the ECU is ignoring the lambda sensor inputs, which could explain the problem with exhaust emissions at test time and would possibly give rise to fuelling problems ?
Everything looks right, sensor and tune resistor wise, to my untrained eye, but for the life of me I can't figure out why the ECU is doing its own thing as regards the map. Any thoughts or suggestions gratefully received.
All this has taken a little bit of a back seat this weekend since the speedo died on Friday so I've spent a spirited few hours trying to get the dash top off - without doubt the worst job I've encountered so far.
MOT imminent, so speedo is top priority, but the mapping thing is bugging me.
Off for a bit of a lie down now....
If it helps, I had a similar issue recently - car appeared to be over fuelling and smelt like it was running rich, failed an MOT with very high hydro carbons and the O2 reading showing it was lean.
Tracked the fault down to a spark issue (faulty coil pack). Essentially the car was not igniting the fuel in one cylinder which was dumping unburnt fuel and oxygen in to the exhaust. The wide band lambda in my car was detecting the extra oxygen and was trying to richen the engine up by adding more fuel.
Tracked the fault down to a spark issue (faulty coil pack). Essentially the car was not igniting the fuel in one cylinder which was dumping unburnt fuel and oxygen in to the exhaust. The wide band lambda in my car was detecting the extra oxygen and was trying to richen the engine up by adding more fuel.
Bassfiend229hp said:
Is a working speedo even needed for an MOT?
I know that a working odometer isn't as mine has registered zero for the last six MOTs...
Phil
From the MOT Manual 2013 - Section 6.7 (updated 24th March 2016): I know that a working odometer isn't as mine has registered zero for the last six MOTs...
Phil
Method of Inspection
1. Check that a speedometer is fitted.
2. Check the condition of the speedometer.
3. Check that the speedometer can be illuminated.
Reason for Rejection
1. Speedometer not fitted.
2. Speedometer incomplete, clearly inoperative or the dial glass broken or missing.
3. The speedometer cannot be illuminated.
Thanks Gents...
I've had the lid off the ECU and everything looks in good shape - EPROM seated ok. Tune resistance was measured at the back of the ECU plug between pins 5 & 27 so should be ok, but I'll check again tomorrow once I've got a new battery for my meter.
I logged some moving data with Rovergauge at the weekend so it's definitely using map 1.
As regards speedo problem, the road speed logged by Rovergauge seems to be doing what's expected - shows 0 at rest and then jumps to 35-ish when moving, so I'd assumed that the speed sensor was working and sending signal to the box of tricks behind the dash, which in turn was feeding the ECU the right info. Currently ferreting around for a scope so I can see what the speed sensor is sending to the box of tricks, and what the BOT is sending to the speedo (if I can figure out the wiring). Hoping speedo is ok and it's something a lot easier, cheaper, and quicker to replace. Seems to be a guaranteed MOT fail if speedo doesn't register.
I've had the lid off the ECU and everything looks in good shape - EPROM seated ok. Tune resistance was measured at the back of the ECU plug between pins 5 & 27 so should be ok, but I'll check again tomorrow once I've got a new battery for my meter.
I logged some moving data with Rovergauge at the weekend so it's definitely using map 1.
As regards speedo problem, the road speed logged by Rovergauge seems to be doing what's expected - shows 0 at rest and then jumps to 35-ish when moving, so I'd assumed that the speed sensor was working and sending signal to the box of tricks behind the dash, which in turn was feeding the ECU the right info. Currently ferreting around for a scope so I can see what the speed sensor is sending to the box of tricks, and what the BOT is sending to the speedo (if I can figure out the wiring). Hoping speedo is ok and it's something a lot easier, cheaper, and quicker to replace. Seems to be a guaranteed MOT fail if speedo doesn't register.
The speed traducer simply produces a very low level sine wave, that varies in amplitude and frequency according to the speed. The speedo calibration unit simply piggy backs onto the signal before it reaches the speedo, so you know you have a signal at that point as Rovergauge is showing the correct behavour. The speedo itself should work directly off the traducer output.
blitzracing said:
The speed traducer simply produces a very low level sine wave, that varies in amplitude and frequency according to the speed. The speedo calibration unit simply piggy backs onto the signal before it reaches the speedo, so you know you have a signal at that point as Rovergauge is showing the correct behavour. The speedo itself should work directly off the traducer output.
How clever is the calibration unit? If its only function is to tell the ECU the car is moving could it interpret a random or weak speed signal as 'car moving' whereas the speedo is saying the signal is not good enough so I'm not playing?Steve
Its not very clever- the input is a comparator so when you start to see a voltage swing across the +ve and -ve inputs, the comparator output switches and fires a timer circuit that makes the fake speed signal. As one side of the input is the ground plane, and the other the transducer output it is possible to get enough voltage float on the ground plane to make the comparator switch its output if it goes above the transducers + side- but the symptom is a speed signal with no car movement. The input really needed some resistors to put a bit of DC bias on the inputs so you dont end up with both inputs being near, but not quite 0 volts. The speedo input is designed to run directly off the very small transducer voltage.
Steve_D said:
Sadly 99% of that went over my head.
So, what I was asking is can the box give the ECU (and RoverGauge) the signal it needs when in fact the speed sensor signal was very poor and not good enough to drive the speedo.
Steve
it might... if you drive yours fast.. like 80 mph and speedo suddenly wakes up, that means the sensor is not close enough to the toothed wheel. It can be adjusted closer and will work. Check the clearence Clarence!So, what I was asking is can the box give the ECU (and RoverGauge) the signal it needs when in fact the speed sensor signal was very poor and not good enough to drive the speedo.
Steve
Over my head too, but lots of good info here.
I was thinking along the same lines, but without the encumbrance of any detailed technical knowledge. I'd wondered if the speed transducer is dodgy, but still producing a signal sufficient to drive the interface box to generate a signal to the ecu, but not a sufficiently good frequency input for the speedo to respond to. I've managed to track down a scope so I'll see if I can get a trace of input & output signals to see what's going on. Have to wait until the weekend though. The idea of a quick thrash to see if that will kick the speedo into life is also attractive, and has merit even if it does nothing but add to the body of knowledge. Probably end up replacing the sensor at the end of the day (or tweaking the clearance clarence), but it is better to travel hopefully, etc.
As regards the ECU thing, tune resister is definitely looking good at the ECU plug, so still flummoxed. Is there any harm in removing the resister to see if the ECU will actually respond to the change. From what I understand it should throw an error and fall back to map 0 if it sees open circuit between pins 5 & 27. Is this correct ?
I was thinking along the same lines, but without the encumbrance of any detailed technical knowledge. I'd wondered if the speed transducer is dodgy, but still producing a signal sufficient to drive the interface box to generate a signal to the ecu, but not a sufficiently good frequency input for the speedo to respond to. I've managed to track down a scope so I'll see if I can get a trace of input & output signals to see what's going on. Have to wait until the weekend though. The idea of a quick thrash to see if that will kick the speedo into life is also attractive, and has merit even if it does nothing but add to the body of knowledge. Probably end up replacing the sensor at the end of the day (or tweaking the clearance clarence), but it is better to travel hopefully, etc.
As regards the ECU thing, tune resister is definitely looking good at the ECU plug, so still flummoxed. Is there any harm in removing the resister to see if the ECU will actually respond to the change. From what I understand it should throw an error and fall back to map 0 if it sees open circuit between pins 5 & 27. Is this correct ?
On the tranducer front, you can scope it, but you need to jack the rear of the car off the ground and let the wheels spin in gear to get a reading- so its a fairly hairy way of going about things. The transducer is nothing more than a coil of wire with a magnet down the centre, so you can test it by doing a simple resistance check across the two connections to the transducer- I dont recollect the resistance, but Id think it to be quite low- in the hundreds of ohms. Water does get into them and corrodes the internal connections so they can fail and will be open circuit. I can understand a situation if the sensor output was low of it starting to work as the speed rises, say due to too larger an air gap, as the voltage will rise with RPM so you can reach the trigger threshold and everything bursts into life, but if the speedo calibration unit is firing correctly at low speed, yet the speedo is dead I think the transducer signal is good.
Edited by blitzracing on Wednesday 30th March 13:00
Limited progress to report on both ECU and speedo fronts due to work and domestic commitments....
Promised scope hasn't materialisd yet, but I've got the car up in the air and eventaully found the speed sensor and checked resistance as close to source as possible. Reading in the 100,000's Ohms so it looks like the sensor is goosed. I'll replace, but looks like a nightmare to get access.
For future reference, I had the lid off the little speedo interface box and the output to the speedo is linked directly to the input from the speed sensor - no signal clean up or other electronics involved. Incoming signal from the sensor is tapped off, processed by the TVR electronics, and stationary/moving signal sent to the ECU.
I've also disconnected battery, removed tune resistor, reconnected everything and gone for a spin. Rovergauge still tells me I'm running map 1. Now very perplexed about the whole thing. Comms to Rovergauge might be an issue since I made up the cable myself, but all other parameters are looking good on the Rovergauge log so seems unlikely it's a comms issue. Only leaves a faulty ECU unless anyone has any other ideas.
If I need to source a replacement ecu then it might just tip the balance towards an aftermarket system...
Promised scope hasn't materialisd yet, but I've got the car up in the air and eventaully found the speed sensor and checked resistance as close to source as possible. Reading in the 100,000's Ohms so it looks like the sensor is goosed. I'll replace, but looks like a nightmare to get access.
For future reference, I had the lid off the little speedo interface box and the output to the speedo is linked directly to the input from the speed sensor - no signal clean up or other electronics involved. Incoming signal from the sensor is tapped off, processed by the TVR electronics, and stationary/moving signal sent to the ECU.
I've also disconnected battery, removed tune resistor, reconnected everything and gone for a spin. Rovergauge still tells me I'm running map 1. Now very perplexed about the whole thing. Comms to Rovergauge might be an issue since I made up the cable myself, but all other parameters are looking good on the Rovergauge log so seems unlikely it's a comms issue. Only leaves a faulty ECU unless anyone has any other ideas.
If I need to source a replacement ecu then it might just tip the balance towards an aftermarket system...
Gassing Station | Chimaera | Top of Page | What's New | My Stuff