Welcome to our resource for MG Car Information.



MG parts spares and accessories are available for MG T Series (TA, MG TB, MG TC, MG TD, MG TF), Magnette, MGA, Twin cam, MGB, MGBGT, MGC, MGC GT, MG Midget, Sprite and other MG models from British car spares company LBCarCo.

MG MGB GT V8 Factory Originals Technical - fuel injection starts, no idle

Everything is hooked up & I can start the car & keep ir running by pumping the accellerator, but I cannot get it to idle. Per the workshop manual & advice from a Rover mechanic, I tried to set the idle by blocking the hose to the stepper control & adjusting the screw under the cap on the mass air sensor, but no luck. I can only get it to idle by blocking the butterfly partially open. It is running very lean & will backfire at the air intake.

Amy ideas?
Jim Stuart

You are probably over this hurdle by now but look for leaks around the plenum and check the hotwire measurement - it is possible to 'wind' the pot totally off the scale (and this may be why its running lean) but it will come back (best done on a bench).

Next turn up the idle on the throttle housing.

Then confirm that you have not reversed the Fuel and water temp sensor connectors.

persevere - the end result is worth it.



To my eternal embarrasment (I will probably never live this down) I had exactly the same was caused by transposing the No1 injector connector on the harness with the water temp sensor one !!!

The give away (apart from the obvious colour difference of the connectors)was that I had 7 sooty plugs, and one pristine clean one.

Michael barnfather

I think you may be adjusting the wrong screw to set the idle. The butterfly adjusting screw is located on the throttle body and needs to be adjusted with the stepper valve closed (or the hose blocked so that air is not introduced from there). You may also need to verify the Throttle Position Sensor voltage output setting. IIRC the Hot Wire setting is done afterwards.

Edd Weninger

It could be the tps voltage, but I'm more willing to suspect the stepper motor is out of sync with the system. Here's the method I use to set it:

Code 48 - Stepper motor
Check base idle speed as follows:

First remove and clean the idle motor and the port it screws into. Clean the throttle body as well.
On the top of the throttle body you will see a hole (possibly covered by an anti tamper plug) for the base idle adjuster.
Remove the air bypass hose from the throttle body, which will cause the engine to speed up to 2500rpm or so. Unplug the connector to the idle motor after 5 seconds then reconnect the hose.
Squeeze the hose shut with needle nose vise grips and adjust the base idle using an allen wrench to give an idle speed of 6-700rpm. Get the lowest speed you can that gives smooth running and does not stall when blipping the throttle. Screw in for slower idle, out for faster idle.
Reconnect the idle stepper and remove the vise grip and you should be done.
In addition, refer to tests 15 and 16 of continuity test procedure. Check road speed sensor- refer to test 25 of continuity test procedure.

Here's the webpage I use for that information:

That guy is VERY knowledgeable, and I've solved many problems using information on his site. I'm going to connect my road speed sensor this year and see if that doesn't help with some of my frustrations with my idle, and stalling problem.



Justin's comments reminded me that I mis-spoke when I stated that the butterfly had an adjusting screw. It is really a butterfly bypass where some air is allowed around the butterfly during idle.

Justin, my setup has the opposite condition running without the speed sensor. Drops to ~1200 rpm at let off and eases down to ~800 rpm idle but never stalls. I'll bet there is a good useable middle ground there somewhere. I just stopped tweaking mine when I got to where I am now. In any case, keep us updated on the speed sensor install.
Edd Weninger

Hi guys,

I'm having flashbacks to the '03 British V8 Convention in Townsend, TN and Justin's parking lot frustrations with his EFI. :)

Shut up man! =) Most of my problems were caused by my using a fuel injection computer that wasn't happy at all. When I put my old computer back in she ran fine... sort of...

My car does much the same As Ed's does. I let off the gas and it drops to about 1200 rpm's, then slowly feels its way back down to around 800. Occasionally it will start to rev up to around 1200, then back down again, and continue to do so until a) it settles at around 900 b) it stalls itself out. I still haven't figured that one out yet, but I'm working on it. I suspect it has something to do with my oxygen sensors, lack of road speed sensor, and my fuel temperature sensor might not be healthy either. Lots of little things that need to be played with before I'll know what the verdict is.


Ed, Justin,

Several of us over here have the same symptoms, we never fit Lambda or road speed sensors, Im not sure they are the problem, my system has been checked by the local Lucas/Bosch guy, he confirms all systems are OK, but when I knock the car out of gear quickly, revs rise to 1200, dropping to 800 idle.

I have a Golf GTI which occasionally suffers from similar problems .

Michael barnfather

Edds quite right, you've got the wrong screw. It's the one on the plenum (under a cap) between the air intake and throttle and you need to give it a few turns, almost until it whistles. It's an allan key. I think the one on the air meter is a variable resistor that's part of the hot wire wheatstone bridge set up. I'm not sure but I think it's used to adjust the mixture. Does any one know for sure? I wouldn't mind leaning out my mix a little.
I suspect the root cause of the problem is the phlenum shortening that we do.

Here again is my problem. It will start, run a few seconds, then die.

I have swapped in 3 different stepper motors, 1 from a known running car. I have tried 2 different throttle position sensors. All the wiring has been checked. I have no signifficant vacuum leaks. I have tried various methods of calibrating the stepper. The control or bleed screw has been uncovered & adjusted from completely closed to 6 turns out. The initial setting was 3/4 turn out from closed.

If I block the throttle part way open, the engine will start & run with significant rpm variations from 500 to 3500.

I am going to try a computer swap this weekend, but don't really think that is the problem. Is it possible that the fuel pump is not working coming off the start mode into idle? It obviously works for a few seconds when the ignition is turned on, & will work at partial throttle. Anyone familiar with how the circuit works?

Thanks to all who have responded.

Jim Stuart

Wasn't the fuel injection running before, or was it just the fuel pressure you tested out?



I'm going from memory here, I'm at work and don't have the manuals at hand. However, you should be able to get the car to run and idle without the stepper valve involved at all. Block off any opportunity for the stepper valve to admit air to the plenum. You can do that by crimping or plugging to hose. Don't rely on the various methods to switch power on and off to supposedly get the valve closed. I don't find them reliable.

You need to calibrate the starting position voltage and range of the TPS if you are interchanging them, you weren't clear whether you had done that.

Set the potentiometer on the hot-wire, the one you were adjusting originally, according to whatever specification you can find. (I'll look up something tomorrow if you can't find it.

Close the butterfly bypass grub screw completely and open it about 1 1/2 - 2 turns. Crank it and see what happens. Try different 1/2 adjustments of the bypass and see if you can get it to idle. Don't blip the throttle.

On start-up and base idle, the computer is just pulsing the injectors for a fixed time. The amount of gas injected needs to be supplied with enough air to burn. The base idle air is supplied solely through the bypass. The computer only gets serious when the butterfly is opened. The computer then determines the amount of opening via the TPS potentiometer and changes the injector pulse duration, so you have to get them both right from the start. Not like carbs which will work at just about anything.

I'm guessing your computer is OK if the motor starts. Idle depends on getting the right gas/air ratio. Getting it to run clean under load with power is a different issue. As is getting the computer to adjust the idle increase via the stepper.
Edd Weninger

unplug the fuel pump relay and connect the power to the fuel pump with a wire and a couple of spade connectors. That way the pump should run continiously without the engine going and you can hear if it's not doing this.
At least that will elliminate the pump as a cause and it's quick and easy to do. You might even want to dissconnect the fuel lines from the rail and run them into a bucket (or something) as I did to make sure that the pump was working OK and that I had connected them the right way around. It's easy enough to mix them up somewhere between the tank and the engine.

Umm , you are are using a roller vane high pressure fuel pump? Silly question but you never know.

Jim, as I understand it, at start, the pump relay is energised by 12V from the start relay. When the engine fires ,air is now sucked thru the airflow meter, which sends a signal to the ecu which then provides 12v for the pump to keep going. An other point to bare in mind is that the ballast resistor for the coil is also bypassed during cranking. I don't think this is affecting your problem though. The Morgan V8 website is full of EFI info. You might get some pointers from there.
Good luck. Barrie E
Barrie Egerton

Before trying to get the engine to run, I did test the fuel pump to determine that I had pressure. It is a high pressure Bosch pump from a Volve 240 turbo with max 150 lbs. Running the engine on part throttle, plenty of fuel pressure.

Ed, yours is the clearest explaination of how things work I have seen. I have never calibrated the TPS, nor do I have any info to do so. If you have teat values/procedures, it would be very helpful.

Peter, Barrie, I will try wiring the pump direct this weekend, as I am starting to think that I am not getting fuel once the engine starts. Before I try that, I may substitute the hot wire unit if that controls fuel after the engine starts.

Thanks all for the good ideas & suggestions.

Jim Stuart

If you've never adjusted the TPS sensor, then you have it wrong. I can almost guarantee it. The "sweet zone" is razor thin. Here's a link for a test procedure:

I have a dollar that says this fixes your problem. The fueling at idle is all wrong if the tps isn't set up perfectly.

I hope that works!

great site Justin, thanks


Several caveats. If you fiddled with the Air Mass adjustment, you need to calibrate that also. Justin is correct, the TPS setting should be very close to specification. Problem is, there are different specifications depending on the year of the EFI hardware and the ECU type. Some websites I found do not make this clear. I used the Haynes Service and Repair Manual for Range Rover 1970-Oct 1992 and used the specification for my 1989 system. There were some changes in 1990. If you can get a copy of it, it is worth having as it also has test parameters for all of the other components such as the temp sensors etc.

Let us know what you have and I can look up some numbers later. Good luck.
Edd Weninger

North American vehicles used the "14cu" ecu in 88,89, and probably part of 90. After that, the 14cux computer was introduced, and remained very stable until mid 95 when the 4.0 engine was intoduced. Any 14cux ecu will respond to the tps check on that website. The first site I posted early this week has a link to all the fault codes, and test procedures.


Just got back from the shop with no better results. I did hot wire the fuel pump & that had no affect, so I have sort of proved that the system is OK.

Yes, I have fiddled with the throttle pot, & no, I have never adjusted it, just removed & replaced. Just printed the RPI page on adjustment & will try that tomorrow. Somewhere I have a multimeter.

I have a harness from an '89 Rover, intake parts from various cars, several pots & stepper motors, but all runs with a 14 CUX computer. I have 2 air flow meters, both with the same model number, 1 is from a 1990, the other is unknown.

If this works, Justin & Ed & others will be due more than a $. It just may keep me out of St. E's for a bit longer. (local nut house- sorry, home for the mentally challanged)

More tomorrow.

Jim Stuart

Spent most of the day playing with the fuel injection with only slight improvement.

I calibrated the TPS, & you all were correct- it was way off. I also found the butterfly was not completely closed. I drilled & tapped the housing & installed a grub screw for adjustment, & got that corrected.

With the corrections, the engine would start & now it would run for 1-2 seconds before dying. Fiddled with the allen screw under the cap but could only get the engine to run for a few revs before running out of whatever.

If this was a carb, I would just increase the idle adjustment screw, or increase the float level. Obviously, I can do neither on the FI.
Jim Stuart

I agree - that's exactly what I would do - and it's surprising that it doesn't work.

In your shoes, I think I might hold the throttle open and run the engine up to working temp - see what happens with the sensor inputs. At the back of my mind here are problems with the sensors and/or leaks around the plenum or with the gasket ( you did use the latest compo didn't you ?)

If she runs smoothly at working temp - the injectors/ pump will all be good - is the mixture still too lean ?

.. and the ignition timing hasn't been disturbed ?


I'd be pretty frustrated by now I think, but the answer is probably just around the bend. Not that I can add much in the way of advice, but perhaps some encouragement is in order. It helps (or should) that you are using a system basically built for the engine, and you know, it would seem with the number of those out there that tuning aids would be available. Have you looked for a Rover EFI group on yahoo by any chance? I know that on my ford system a tuning aid (tweecer) was simply invaluable, as it allowed me to get into the processor, watch real time operation, create log files, and modify settings to get it to run better. I'd certainly think somebody has got to be making something similar for the Rover. Without that advantage it seems about all you can do is trick the sensors and such.

Anyway, I presume you've gone over all the obvious points (like the start/ign leads) several times but occasionally one gets tripped up. Recently on another forum a fellow had a very similar problem, only to find after weeks of trouble that he was trying to run the car with it in gear. What kept it from being obvious was that he had the car up on jackstands but with the emergency brake on and the tires touching the floor. When he finally gave it full throttle to start, it ran just fine... and filled his garage with tire smoke!! So despite all indications his efi system was not at fault. If only your case could be so simple. On mine I've had a persistent stumble at 2100 rpm which I now believe is the result of interference, possibly from the fan. Hopefully a new shielded timing sensor mount will cure it, if not a new wiring harness may be in order.

Aside from that, I would suspect the relationship between your start and run injector settings. If you can monitor your injector pulsewidth that would help nail it down. A crude tool to do that might be a "noid" light to show you what the injectors are doing. Also pulling one plug wire to watch the spark output could help. Sorry if these suggestions have been made already as I've had trouble keeping up lately, and best of luck. I know you'll figure it out eventually, but maybe some of this will help.

Jim Blackwood


It sound like it is getting too much fuel at idle (if it seems to run with the butterfly partially open). There should be a vacuum compensation on the fuel pressure regulator. Check and see if it is competent. And, are you sure the Stepper Idle Valve is blocked off and you have no other air leaks into the plenum? This sounds counter-intuitive but this may cause an odd manifold (plenum) pressure which could muck thing up.
Edd Weninger

I should have mentioned that the Fuel Temp Sensor and the Coolant Temp Sensor are also involved in setting the Idle pulse timing, I believe. Though not likely for failure assuming the car was running well before, might be worth checking.
Edd Weninger


Just to be clear on the symptoms:

It starts and dies. You can't get it to run for more than a few moments no matter what you do.

I don't know this system from Adam, but FI is FI is FI.

If these are the symptoms, I would want to put a fuel pressure guage on or close to the rail before the fuel pressure regulator. Is it possible that your fuel pressure regulator is not holding pressure? Is there a way to bench test the regulator?

For the impatient wishing to toss caution to the wind and see if this is a possibility, Does anyone see a problem with pinching the return line partially closed with a pliers momentarily to see if it continues to run for a longer period before dying? (you are holding more pressure in the rail this way)

Will it run with the gas cap off?
(some kind of vaccum problem) I doubt this problem a blcke vent would cause it to die so quickly though.

Good luck!

Brian C.

Brian Corrigan

Another thing which I'm sure you have checked but still worth mentioning - is the static ignition timing set correctly? It is not unknown for the timing mark on the crankshaft pulley to be in the wrong place especially if you are using a late serpentine engine with an earlier type pulley. Its best therefore to do a physical check that TDC on number 1 cylinder is marked where it should be.

Geoff Richmond

Using Justins excelent RPI site,
I've been playing arround with the air flow meter in order to bump up CO content, which should make for more power. My motor is STD range rover and is set at .18% CO.
Now I don't own a CO analyser but if I use the CAT' converter instructions that should work. These instructions sugest aiming at between 1.5 to 1.95 V between the Blue brown air flow wire and redblack ground wire on the air flow meter. Problem is my loom has different colout codes. What I'e got is Brown orange(obviosly power) red, bluegreen and Blue red. Is it the two outer wires
peterozI've been pl

1 more thrash this weekend, with no improvement. I swapped ECU's to eliminate that element, and as expected, no change. The fuel pump is running with key on, stops after a few swconds, runs in the start mode & continues to run for a few moments as the engine dies, then stops. I feel this is not the problem.

Hope to find a new fuel regulator & install that next weekend. This seems like the final option.

Anyone out there with a new regulator for sale, plese contact me off list.

Thanks all for the help.

Jim Stuart

Jim - your first posting suggested the engine would run with the throttle open - is this no longer the case ?

Edd and I thought you might not have have the butterfly bypass setup. Can you confirm this is open ?

It is easy to get the two temp connectors reversed - again, can you confirm ?

There is a cruise control bleed at the back of some plenums - yours blocked off ?

The fuel line when disconnected at the motor produces a stream Mannekin Pis would be proud of ?

The FI harness is properly earthed to the chassis (not solely via the MGB gearbox strap ?

There is nothing else in the car which would make the voltage erratic ? - i.e. a drain with the ignition on - the coil ??


Engine will run erraticly with the throttle partially open. It hunts, & rmps change from 4-500 t0 2500. No change

Not sure what the butterfly bipass is. If you mean the adjustment screw on the upper housing under the cap, I am sure it is not correct, but I have adjusted it in half turn increments from closed to 6 turns out. The original setting was 1 1/2 turns out.

Connectors are correct. The Injectors are one color, the temp sender connector is slightly different.

Cruse control is blocked off.

Wiring is grounded to the motor. Motor & tranny are both grounded to the chassis with very large ground wires.

Have not disconnected the fuel line, but have verified the current to the pump & can hear it whine prior to starting. Have a new fuel pressure regulator comming, so during install, I will confirm fuel flow.

Thanks again for the suggestions.

Jim Stuart

Did you remove the plenum heater system from the bottom of the throttle body? (Two rubber hoses go into it, and it's held on by four tiny bolts)If you did, did you plug up those holes? If not, they're a major air leak. I had a coolant temperature sensor go out on me once (stupid thing melted while driving)Anyway, it ran pig rich. So much so that there was gas in my oil... craziness!

I'm not really sure there's a point to my rambling. I think I'm just rambling because I can. Oh yeah, try out that fuel pressure thing and see what happens. Just out of curiosity, You said that you had messed with the air flow sensor trying to set the idle, did you by chance go back to the website from rpi that I listed, and try to get the air flow sensor close to the spec listed there? If you opened up the idle speed screw all the way, it would make the car idle rather high. You could then slowly lower the speed from there. I went through this nightmare for a while. It was NOT fun.

You have my sympathy.

Just a thought.
Might there have been muck in the injector rail that has partially blocked the injectors?.
P.N. Sherman

The point of running it up to working temp might resolve PN Sherman's good point about muck or a faulty injector - justin's question about leaks apart - muck in the throttle bypass can't be overlooked either but the rev hunting shows the ECU is trying to compensate albeit unsuccessfully. Worth adding some injector cleaner to the fuel ?

Do you not have two temp. connectors ?

FYI anyone
Adjusting the the voltage across the air mass meter to 1.7V as per Justins ref' has made a BIG difference in power.


I have been working with this EFI system for several years now with Glenn Towery. We pulled our hair out for 3 weeks on a factory MGBGT V8 Sebring we converted to EFI that had the EXACT same symptoms. The issue turned out to be a lower intake manifold that was cracked between the velocity stack ports. The vacuum was leaking on the bottom of the intake where it was impossible to detect by traditional means.

I'm as sure as I can be that you have an air leak someplace. I would check by pressurizing the intake with compressed air. We used a blow gun in the brake master cyl. booster hose. Have an assistant blow air in while you feel all around and listen for escaping air. You may be surprised at where air will escape. Just be sure to block the air intake with duct tape while you look for the leaks.

All of the above suggestions are great but will be totally ineffective if you have air leaks. Above all don't be discouraged. The rewards are worth the effort.

Keep us posted,


Following on from John's posting above - alternatively; spray WD40 (or similar) around while you turn it over.

Jim which gasket did you use ?

I did remove the heater section & did not plug the bolt holes, as they did not look like they were through. This will be corrected.

I only have one temp sensor. There was a sensor on the thermostat housing, but that is the electric fan control & is not used in my application.

My check of the workshop manual indicates the 1.7V at the air flow meter is for cat equipped cars with the appropriate tune resistor. My setup is non-cat with the different resistor.

I can't believe that a vacuum leak would prevent the car from idleing when I can get it to run at part throttle, but I am more than willing to prove myself wrong. As far as I can tell, the only air leak I may have is the previously mentioned bolt holes to be plugged. At part throttle, the engine is definately hunting, so that may be the problem.

I am using a new composite valley gasket, my first time to use one. I did add RTV at the water ports- maybe I shouldn't have? I have used RTV to seal the intake mating surfaces, & frankly, it is a pain when you continually take them apart & reassemble. Any good
Stateside product recommendations would be helpful.

More trials on Sunday. Saturday, I take the GT with the Buick 300 to the paint shop. (it has a carb & runs great)

Thanks again to all. This thread goes to he shop with me.

Jim Stuart

There are two temperature sensors:
1) on the fuel rail
2) going down into the intake manifold for measuring coolant temperature.

You have them both right? A failure in either module will not help your cause. If you want, I'll take very close up pics of both sensors and e-mail them to you.


I concur with Justin. If you have the Hot Wire system and the 14cux box, it needs to have both temp sensors properly connected and, obviously, the correct values.

WRT mixture, can you figure out whether it is dying rich or lean when asked to idle? Install a new spark plug in a convenient cylinder, Run a bit as slow rpm as possible, let it die and see what the plug looks like. Unless you have a cool exhaust gas analyzer :)

Edd Weninger

As you knowI was playing around with the hot wire resistance and got more power with 1.7V (ignition on) across the two outer wires on the air flow meter. The down side was a slightly erratic idle until the engine warmed up. The orrigional setting wss 0.18V if this of use to you. It is around this Voltage that you get a rapid change for a slight adjustment on the screw.
This gets you a nice smooth idle, if the base idle has been set right.
If you set yours to this V you should be able to cross this factor off your list of possibilities.
Another thought, is there a break in the wire to the distributor?


Justin hit it with the vacuum leak from the heater plenum, & I found another bolt hole that was also leaking air.

I have adjusted the air flow meter to 1.7V & it will idle smoothly after a brief warmup. It does want to die when you blip the throttle, & the idle is a bit low at 500 RPMs, but I will get it out tomorrow & see if it will go around the block, & do some tweeking.

Thanks to all on the list who have helped. I have gotten quite an education & would neve solved the problem otherwise.

Jim Stuart

Jim Stuart

Do I still get that $1.00? :)


Yes, if I see you at the V8 Convention at Grand Rapids Aug 13/14. And, if you wish, you can drive the 325 hp MGBGT I will be bring.

Otherwise, send me a mailing address.

Got a chance to drive the FI car today, & while there is some sorting yet to do, it ran very well. Low end is better than before, a bit better accelleration, but a small loss in the upper rpm's. Some of this could be timing.

Tahanks again to all-

Jim Stuart

This thread was discussed between 01/03/2004 and 29/03/2004

MG MGB GT V8 Factory Originals Technical index

This thread is from the archive. The Live MG MGB GT V8 Factory Originals Technical BBS is active now.