For a while now my T16 has had a problem with high idling when warm and also pig rich mixtures. The rich mixtures made the car very foul tempered when cold, and a drunkard when warm :lol:
The high idle was less of a problem, but as I knew that something was wrong I really wanted to fix it.
When I turned up to Kodak, I wasn't expecting to do any work on the car but when I drove home later that day, both problems had been fixed. The reason for the rich running was straightforward, but the cause is still a mystery. I think I have unravelled the high idle cause, and as it's not something I've seen before I wanted to post it here.
None of the work I did on the car was anything new, or high-brow. Hopefully it might be of use to those with a similar problem. In particular, I'm hoping the photos will show some folk where stuff is in the engine bay.
Oh also I'm bored, and haven't used the digital camera for a while :lol:
Despite not intending to do any work on the car, I was encouraged to tinker about with the engine to solve the high-idle problem. I say high, it was about a steady 1200RPM on a warm engine.
I could only think of seven ways that the idle could be elevated: rich running; vacuum leak; idle-adjustment screw; dashpot wound in too far; throttle-stop screw; faulty auxiliary air valve or misadjusted throttle-cable. All of these I had adjusted before but with no improvement.
This is the throttle-body on a T16:
![Image]()
The idle-adjustment screw (centre of the picture) opens or closes an air bypass channel in the throttle body. Alex wondered if the channel had become blocked with crud, so I shot some carb cleaner into the channel.
![Image]()
To my amazement some of the carb cleaner exited the from solid metal on the other side of the throttle body.
It turns out that I'd been ham-fisted when previously trying to lower idle using the adjustment screw. I'd managed to puncture the throttle body using the end of the screw. In the photo below, you can just see cracked metal where the screw's pushed through:
![Image]()
This was potentially a vacuum leak. I pushed back the metal which seemed to real the hole. I'll do a proper and permanent repair soon using JB Weld. Be careful when winding in that adjustment screw!
Alex asked if perhaps maladjustment of the throttle-position switch was preventing the throttle-butterfly from fully closing. As I'd never adjusted the TPS, I didn't think this to be the cause of the high idle but given a lack of other options, I gave it a go.
The throttle-position switch (centre of picture):
![Image]()
![Image]()
The switch is adjusted by undoing two small bolts (7mm IIRC) allowing it to tilt and rise/lower. You can see each bolt in these pictures, they're either wide of the TPS:
![Image]()
![Image]()
Even though the switch was correctly switching the ECU between different engine maps, when unbolted it became clear that the switch was holding the throttle open very slightly. This allowed metered air into the engine, raising the idle speed.
The TPS informs the fuel injection ECU on the throttle's position. There are three positions: closed (idle), partially open (cruise) and wide-open.
Closed throttle causes the ECU to use a special idle map. Cruise means that fuelling's measured according to air metered by the AMM. WOT IIRC engages an RPM-only map. It is essential that the TPS is correctly adjusted so that the ECU can alter fuelling for different engine conditions.
The Bosch LH2.2 TPS is a switch, unlike later LH2.4 and Lucas cars that use a potentiometer. Because it's a switch, you can hear a quiet click as the throttle's opened to shift from idle to cruise and cruise to WOT.
A correctly adjusted TPS will switch from idle to cruise as soon as the throttle starts to open. Adjust the TPS so that when you rotate the throttle spindle, the switch clicks open immediately. Make sure it clicks again when the throttle's shut and that it doesn't hold open the throttle plate.
Adjusting the TPS is a PITA. Just when you have the thing in the correct position and tighten down the bolts, everything shifts slightly. Because only slight movement should be required to open the switch, tightening the bolts is just enough to cause maladjustment.
Because the throttle plate was now allowed to shut fully, I found that the throttle cable was too tight and that now was holding open the throttle. This of course then required further adjustment to the TPS.
I advise you to fully slacken off the throttle cable tension thing before you adjust the TPS:
![Image]()
![Image]()
The throttle cable should be adjusted so that it's only slack enough to allow the throttle plate to close.
Next I adjusted the throttle-stop screw:
![Image]()
Wind the screw down until it just touches the throttle stop and then wind it back 3/4 turn.
Since the dashpot is required mainly for emissions, I wound that fully back:
![Image]()
It's worth pointing out that I did almost everything in the wrong order, and consequently had to keep re-adjusting the TPS. To make the job easier, wind the throttle stop screw fully up; fully slacken the throttle cable; unwind the dashpot. Then adjust the TPS; the throttle stop; cable tension; dashpot.
I had suspected the auxiliary air valve, but this appears to be functioning correctly. I've cleaned it through with carb cleaner and WD40 on several occasions.
![Image]()
Onto the rich running. Two of the top suspects for rich running are: faulty NTC sensor and air mass meter.
We started with the NTC sensor, by reading its resistance with the car at normal operating temperature. Although initial readings indicated open-circuit (faulty sensor) later readings showed correct operation.
The NTC sensor is mounted in the centre of the inlet manifold, in between runners for cylinders 2 and 3. The top looks a bit like a fuel injector, but the electrical connector is blue:
![Image]()
![Image]()
One of the diagnostic tricks suggested by Alex was to short-circuit the NTC sensor with the engine warm. If the sensor's faulty, the fuel injection ECU will default to a cold-start enrichment mode. This means the car runs pig rich all the time. Since the sensor's resistance is inversely proportional to engine temperature, short-circuiting it means the ECU sees a hot engine and will not engage cold-start enrichment: mixtures should return to normal.
Make sure the ignition's switched off before short-circuiting any engine sensors.
Using John's Gunson meter, we found that C02 levels at idle were 8.7%
when they should have been less than 3%. This reading stayed steady whether the NTC sensor was connected or shorted, so the conclusion was that the sensor was functioning correctly.
We then turned attention to the air mass meter. Using a pin-out diagram in the Bentley, Jon and I were able to check the meter's resistance. The reading was way above the correct which would give rich running, and explain the 8.7% C02!
Bentley specifies resistance of 380 ohms for the AMM, but this is for catalysed cars. My car does not have a catalyst so we tried setting the meter to 280 ohms.
The AMM is adjusted using a trim pot on the side of the meter. Jon commented that the most common failure mode of the meters is breakage of the trim pot. This would mean the meter does not respond to adjustment, and gets binned. However, the fault can be easily corrected using an inexpensive part, so don't be too quick to bin that AMM!
Jon also suggested blasting the meter's filament and thermocouple (heat sensor) with carb cleaner before refitting it to the car. Whilst the AMM's filament gets cleaned whenever the engine's shut off, the thermocouple does not. Muck stuck to the thermocouple will affect accuracy of the meter.
A prolonged and carefully aimed blast of carb cleaner soon removed all the dirt from the AMM's components.
We refitted the AMM and checked C02 levels. They were 1.5%, and we adjusted this up to 2.0%. Job done.
I might replace the NTC sensor and auxiliary air valve if they're inexpensive. After all, they've been working hard for 17 years!
Thanks to everyone who lent a hand at Kodak.
Sorry for the ramble. I warned you at the start it wouldn't be anything new Please post corrections as necessary.
__________________
Best Car Insurance | Auto Protection Today | FREE Trade-In Quote
The high idle was less of a problem, but as I knew that something was wrong I really wanted to fix it.
When I turned up to Kodak, I wasn't expecting to do any work on the car but when I drove home later that day, both problems had been fixed. The reason for the rich running was straightforward, but the cause is still a mystery. I think I have unravelled the high idle cause, and as it's not something I've seen before I wanted to post it here.
None of the work I did on the car was anything new, or high-brow. Hopefully it might be of use to those with a similar problem. In particular, I'm hoping the photos will show some folk where stuff is in the engine bay.
Oh also I'm bored, and haven't used the digital camera for a while :lol:
Despite not intending to do any work on the car, I was encouraged to tinker about with the engine to solve the high-idle problem. I say high, it was about a steady 1200RPM on a warm engine.
I could only think of seven ways that the idle could be elevated: rich running; vacuum leak; idle-adjustment screw; dashpot wound in too far; throttle-stop screw; faulty auxiliary air valve or misadjusted throttle-cable. All of these I had adjusted before but with no improvement.
This is the throttle-body on a T16:
The idle-adjustment screw (centre of the picture) opens or closes an air bypass channel in the throttle body. Alex wondered if the channel had become blocked with crud, so I shot some carb cleaner into the channel.
To my amazement some of the carb cleaner exited the from solid metal on the other side of the throttle body.
It turns out that I'd been ham-fisted when previously trying to lower idle using the adjustment screw. I'd managed to puncture the throttle body using the end of the screw. In the photo below, you can just see cracked metal where the screw's pushed through:
This was potentially a vacuum leak. I pushed back the metal which seemed to real the hole. I'll do a proper and permanent repair soon using JB Weld. Be careful when winding in that adjustment screw!
Alex asked if perhaps maladjustment of the throttle-position switch was preventing the throttle-butterfly from fully closing. As I'd never adjusted the TPS, I didn't think this to be the cause of the high idle but given a lack of other options, I gave it a go.
The throttle-position switch (centre of picture):
The switch is adjusted by undoing two small bolts (7mm IIRC) allowing it to tilt and rise/lower. You can see each bolt in these pictures, they're either wide of the TPS:
Even though the switch was correctly switching the ECU between different engine maps, when unbolted it became clear that the switch was holding the throttle open very slightly. This allowed metered air into the engine, raising the idle speed.
The TPS informs the fuel injection ECU on the throttle's position. There are three positions: closed (idle), partially open (cruise) and wide-open.
Closed throttle causes the ECU to use a special idle map. Cruise means that fuelling's measured according to air metered by the AMM. WOT IIRC engages an RPM-only map. It is essential that the TPS is correctly adjusted so that the ECU can alter fuelling for different engine conditions.
The Bosch LH2.2 TPS is a switch, unlike later LH2.4 and Lucas cars that use a potentiometer. Because it's a switch, you can hear a quiet click as the throttle's opened to shift from idle to cruise and cruise to WOT.
A correctly adjusted TPS will switch from idle to cruise as soon as the throttle starts to open. Adjust the TPS so that when you rotate the throttle spindle, the switch clicks open immediately. Make sure it clicks again when the throttle's shut and that it doesn't hold open the throttle plate.
Adjusting the TPS is a PITA. Just when you have the thing in the correct position and tighten down the bolts, everything shifts slightly. Because only slight movement should be required to open the switch, tightening the bolts is just enough to cause maladjustment.
Because the throttle plate was now allowed to shut fully, I found that the throttle cable was too tight and that now was holding open the throttle. This of course then required further adjustment to the TPS.
I advise you to fully slacken off the throttle cable tension thing before you adjust the TPS:
The throttle cable should be adjusted so that it's only slack enough to allow the throttle plate to close.
Next I adjusted the throttle-stop screw:
Wind the screw down until it just touches the throttle stop and then wind it back 3/4 turn.
Since the dashpot is required mainly for emissions, I wound that fully back:
It's worth pointing out that I did almost everything in the wrong order, and consequently had to keep re-adjusting the TPS. To make the job easier, wind the throttle stop screw fully up; fully slacken the throttle cable; unwind the dashpot. Then adjust the TPS; the throttle stop; cable tension; dashpot.
I had suspected the auxiliary air valve, but this appears to be functioning correctly. I've cleaned it through with carb cleaner and WD40 on several occasions.
Onto the rich running. Two of the top suspects for rich running are: faulty NTC sensor and air mass meter.
We started with the NTC sensor, by reading its resistance with the car at normal operating temperature. Although initial readings indicated open-circuit (faulty sensor) later readings showed correct operation.
The NTC sensor is mounted in the centre of the inlet manifold, in between runners for cylinders 2 and 3. The top looks a bit like a fuel injector, but the electrical connector is blue:
One of the diagnostic tricks suggested by Alex was to short-circuit the NTC sensor with the engine warm. If the sensor's faulty, the fuel injection ECU will default to a cold-start enrichment mode. This means the car runs pig rich all the time. Since the sensor's resistance is inversely proportional to engine temperature, short-circuiting it means the ECU sees a hot engine and will not engage cold-start enrichment: mixtures should return to normal.
Make sure the ignition's switched off before short-circuiting any engine sensors.
Using John's Gunson meter, we found that C02 levels at idle were 8.7%
when they should have been less than 3%. This reading stayed steady whether the NTC sensor was connected or shorted, so the conclusion was that the sensor was functioning correctly.We then turned attention to the air mass meter. Using a pin-out diagram in the Bentley, Jon and I were able to check the meter's resistance. The reading was way above the correct which would give rich running, and explain the 8.7% C02!
Bentley specifies resistance of 380 ohms for the AMM, but this is for catalysed cars. My car does not have a catalyst so we tried setting the meter to 280 ohms.
The AMM is adjusted using a trim pot on the side of the meter. Jon commented that the most common failure mode of the meters is breakage of the trim pot. This would mean the meter does not respond to adjustment, and gets binned. However, the fault can be easily corrected using an inexpensive part, so don't be too quick to bin that AMM!
Jon also suggested blasting the meter's filament and thermocouple (heat sensor) with carb cleaner before refitting it to the car. Whilst the AMM's filament gets cleaned whenever the engine's shut off, the thermocouple does not. Muck stuck to the thermocouple will affect accuracy of the meter.
A prolonged and carefully aimed blast of carb cleaner soon removed all the dirt from the AMM's components.
We refitted the AMM and checked C02 levels. They were 1.5%, and we adjusted this up to 2.0%. Job done.
I might replace the NTC sensor and auxiliary air valve if they're inexpensive. After all, they've been working hard for 17 years!
Thanks to everyone who lent a hand at Kodak.
Sorry for the ramble. I warned you at the start it wouldn't be anything new
Please post corrections as necessary.__________________
Best Car Insurance | Auto Protection Today | FREE Trade-In Quote
