Welcome to our Site for MG, Triumph and Austin-Healey Car Information.
MG MGB GT V8 Factory Originals Technical - EFI Drivability
|I posted several months ago and received advice on problems with surging during acceleration. To date I've not been able to get the FI working correctly.|
Any help would be appreciated. The system is a 1980 L-jetronic off of an 1880 US SD1.
Symptoms - Start warm-up and idle is fine. Pushing the accelerator, the car bucks and surges. CO at idle is set at 2.5% (Cat is installed). Mixture at cruise and idle must be nearly stoicometric, with the output of the oxygen sensor reading 360 to 450 millivolts. When the throttle pedal is pushed down, the O2 sensor output drops to 0-100 millivolts, indicating the mixture has gone very lean.
What I've checked so far - Throttle position indicator has correct voltage. Fuel pressure is 28 psig at idle, rising to 36-38 psig at WOT. Water temperature sensor has correct temp vs. resistance. Power to ECU is continuous. Signal from airflow meter shows voltage no discontinuities with changes in airflow. Output of the two O2 meters is within 50 millivolts of each other.
I would think that the airflow meter is the device that provides a signal to the ECU to enrich the mixture on acceleration. Is this airflow meter used on anything else that I could substitute for this one?
Rover were quite advanced for the 1880 SD1 and maybe the electronics have just given up!
Try CAR ELECTRONIC SERVICES, (ECU Diagnostics & Repair) for typical fault analysis. http://www.carelect.demon.co.uk/frame.htm
Iíll post a few comments extracted from their web site that may help.
The Lucas 4CU fuel injection ECU (fitted on Rover 3.5 vehicles up to late 1989) has proved to be extremely unreliable and virtually all engine idling/running problems are likely to be due to faults with the ECU. The airflow meter is the most reliable part of the system and should NEVER be opened/adjusted/cleaned or lubricated.
It is IMPOSSIBLE to adequately test relays, temperature sensors, idle control valves, MAP sensors. ALWAYS REPLACE these items before eliminating them from your decision making process.
VEHICLE HESITATES THEN SURGES FORWARD WHEN PULLING AWAY FROM STOP.
- Physically check rotation of throttle position sensor. Should be smooth.
- Check throttle sensor output at idle - should be between .3V and .36V. Adjust if necessary.
- REPLACE the water temperature sensor. (The one behind the easy one!)
- Check for sticking airflow meter flap. If any roughness or sticking, try applying WD40 to flap spindles from air inflow side and check for corrosion deposits. Do NOT remove plastic cover. If flap operation remains rough, airflow meter needs professional rebuild.
If above check out OK then the ECU is faulty.
Thanks for your reply.
Your referenced website says it all - to finally diagnose the problem requires the ability to substitute components to verify the faulty components.
It is interesting to note that CES says the AFM is robust and the ECU is often at fault. If other websites covering the L-jetronic are any indication, the AFM seems rather troublesome. I must admit I've taken mine apart to move the resistance track relative to the wiper in an attempt to get the wiper onto a fresh area of the resistance arc.
4CU ECUs and the airflow meter used on the SD-1 are in short supply in Arizona. My hope is one of the members of the local Triumph club has a TR8 that will let me try and substitute components.
Failing that, I'm going to put my carb back on and start looking for a hot wire system. If I could find an Electromotive or Halltech ECU, it would be even better.
Thanks for answering my post.
As there has been little response I will add my two pence. The air flow meter does measure the air entering the intake but its a bit slow. If the AFR is OK at steady speed and load then the meter is probably OK. If your problem is on sudden acceleration then look to the throttle pot for problems. The ECU should look at the rate of voltage change and workout additional fueling for the sudden change in air flow that is expected. So check the throttle pot output is continuous when you move the throttle.
If the problem is only at very low loads it may be worth disconnecting (no fuel) the cold start injector. If this is leaking you can compensate at idle with the screw on the air meter, but this biasses the rest of the fuel map so as soon as you increase air flow the rest of the map will be lean. This affect is small at moderate and high air flows though.
Also check for air leaks between the throttle and the air meter. A small leak can be adjusted out to get the correct idle speed and AFR, but as soon as you increase air flow significantly you could go lean again. Hope this helps.
|Voltage from the throttle pot checked out with the car static, but I was still suspicious.|
I built a break-out box and wired it into the harness such that I could read the voltage going to the ECU from the throttle potentiometer and it was smooth and continuous with movement of the throttle. (Can also read the voltage going from the AFM, the O2 sensors and the resistance of the water temp sensor)
I think the throttle pot is common to some other Bosch parts, including Jags, which are plentiful around here. That is at least one substitution I can do.
I'll blank off the cold start injector and make sure it isn't contributing an idle fuel flow and masking the problem.
For inlet leaks, I fabricated a steel 'el' from a piece of large header tube and sealed it with new hose and clamps. I checked for leaks with propane - none found.
Thanks for the input - I'll let you know this weekend if it worked!
|I have a 1985 BMW 325E that does that as well. The only difference is that the surging only takes place if I'm not at full throttle. It has been to the dealer many times since new (it now has 255,000 miles on it) for this problem, I finally gave up. The only thing that works for me is to put in a bottle of of injector cleaner, that lasts for 3 or four tanks of fuel before the problem returns.|
|Michael S. Domanowski|
|One other thing I've observed. |
The car performs flawlessly when cold. I can put the car in fifth gear at 500rpm, stomp on the accelerator and it will accelerate smoothly.
Once it warms up, the same test will result in violent bucking and surging, with voltage from the O2 sensor dropping to less than .1millivolt, indicating a very lean condition.
|Phil, this may sound odd, but have you checked and rechecked the timing anf timing advance? Reason I ask is that I had problems with my truck when the timing was off, she ran and idles great cold, but when warm the idle would rise and it would run diffently with less power.|
The lean situation I cannot help with, like others have said it could be many things, such as bad potentioners etc..
|One thing to consider: does the car perform badly when it warms up because it warmed up or is it because the temperature sensor told the ECU it was warm and it made compensations that caused the poor running? If the latter, maybe fake out the ECU with a resistor in place of the sensor to tell the unit it is always cold.|
Went through the diagnostic process one more time.
When I substitued a 175 ohm resister for the water temp sender - the problem went away.
I'm off to find another sensor.
The words in your first response to my question now make sense-
"- REPLACE the water temperature sensor. (The one behind the easy one!)"
I took this out at one time and boiled it in water to check the resistance output and it checked out ok. Must be some other failure mode.
Thanks for the assistance,
|As you mentioned the O2 sensor I assume you mean it runs with one.|
Bear with me - this is just a theory. Perhapse its the sensor thats wrong. This may help Micheal as well.
On most cars the sensor is used when the engine is warm and NOT at WOT (or when the engine/cat would get to hot). When the engine is cold it uses a predefined injection based on air mass so its OK cold. When it gets above a certain water temperature it tries to use the O2 sensor, if the sensor has a fault it will make the fueling incorrect (but may still be driveable). If you then change the load quickly it could have a problem. At full throttle it doesn't use the sensor so its back to the predefined fueling. The sensor error means the voltage may appear OK but its actually a long way of the AFR you want. If the temperature sender doesn't solve it try running without the O2 sensor but I don't know how the calibration copes, hopefully it goes to predefined fueling not some limited operation.
|It actually has two O2 sensors, one for each bank of the engine.|
Typical millivoltage on one side is 450 mv with the other side in the 90 to 120 MV range. So one side is extreamly lean, like the computer isn't firing them consistently or one of the injectors is fouled and not puting any fuel in. Either of these two scenarios might explain the output of the O2 sensor. Of course the 02 sensor could be bad. I may have to trade positions and see if the low output follows the sensor or remains on that side of the engine.
Does anyone know what happens when the O2 sensors are disconnected?
I'm begining to suspect a bad fuel injector. The electrical circuit to the injectors is up to snuff resistance wise. It must be a temperature related effect when the injector gets hot.
I bought some noid lights today to make sure it isn't a signal problem. The next step will probably be to putt the injectors out and have them checked and cleaned.
Thanks for your ideas and suggestions,
I am not familiar with the early system but if you disconnect the O2 sensors on the 14CUX system you will need to change the tune resistor - RW
|The original EFI system uses the same fueling for both banks. I would expect the same is true of the system with O2 sensors. If one O2 sensor/wiring is faulty then I assume it would bias all the fueling. In this scenario the car will start and warm up OK, even drive WOT OK, at part throttle it will still cruis OK (but may be lean), however if you open the throttle quickly it will probably go lean.|
Before a big strip down and injector checks you can do the cruise, ignition off coast down run, and check each spark plug for colour.
As far as I can establish if you disconnect the O2 sensors (both of them) you should run on the default fueling.
|One specific thing not yet mentioned is that the ballast resistor can be the source of problems. I have experienced this on a couple of cars where one bank of injectors has gone down either intermittently or for a period of time. |
On one occasion the problem was actually in the ballast resistor and was a heat related issue. The other was a make and break occuring on one of the power feed wires coming into the plug. You will see that there are two feed wires of a common colour code and eight others split into two rows of five. Check that you get the same normal system voltage reading on each of the two feed wires.
Read your comments with interest. A couple of conclusions/questions-
If indeed the ECU uses the same fueling for both banks of injectors and the O2 sensor is not bad, the only (??)
plausible explanation for the observed difference in output of the O2 sensor is that one of the injectors is not firing
or not firing consistently. (Verified the wiring with resistance test and noid light)
If the O2 sensor was bad, I might have diagnosed it by disconnecting both sensor s and go for a road test and observe
the sensor's output. I could have switched positions of the O2 sensors. If the low voltage followed the switched
sensors, then the O2 sensor is bad. If the low output voltage remained on the bank of injectors despite having changed
the O2 sensor's position, then it was likely a bad injector.
If one O2 sensor was bad, with the voltage dropping to low values indicating a lean condition to the electronics, would
the electronics compensate by richening the mixture, or would it return to a default-fueling schedule and ignore the O2
sensors? If it tried to richen the mixture, wouldn't it continue to drive ok, just the fuel consumption go up? If it
went to a default fuel schedule, wouldn't it do about the same thing - consume more fuel but still run ok? My symptoms
seem much more severe than a bad O2 sensor could produce. I'd always thought that an O2 sensor failed 'soft' and the
only way you know it is bad is mileage going away (or a fault code on later EFIs)
The questions /comments are probably moot, since I've already pulled the injectors and rigged up a fixture to flush and
test them. Before I reinstall them, I'll take them to the local fuel injection specialist who will test them all for
$25 and verify that they are good.
I will however, take your advice and pull the plugs and inspect them both now and if I continue to have problems once
the injectors are reinstalled. But if an injector isn't firing any fuel into the cylinder, wouldn't that be
undetectable on the plug? If the O2 sensor is bad, then one would think the plugs would be dark with carbon on both
sides of the engine.
Thanks for your input. Any further comments that may help me clarify my diagnostic strategy are appreciated.
Trying to make a 22 year-old obsolete injection system work is a pain in the rear!
|Good point Roger.|
Did a bit more digging:-
Although all 8 injectors are pulsed the same there are 2 drivers in the ECU so they are operated in sets of 4. I don't know if this is 4 per bank but it does meen that if one driver is failing or there is a circuit fault on one set you will have the AFR difference.
The O2 sensor output is compared to the base fuel value and treated as an ajustment. Can't find a reference to how 2 sensors are treated. Also it seems I was wrong in stating the throttle sensor was used in transients. Aparently they rely on the overshoot of the air flow meter to supply enrichment after initial enleanment.
Don't forget that the 'rich' O2 sensor could be the faulty part thus biasing the mixture lean.
If you didn't drive the car and key_off while driving you wont get a good reading on the plug colour. Especialy if the engine was running OK at idle.
If you have a little gadget called EGR it could be the culprit. If you do, take it off and compress the diafram (spelling) using your fingers to overcome the spring pressure and while holding it in the compressed position place your finger over the vacuum nipple to stop any air from entering and watch to see if the diafram moves at all when you release the pressure from the spring, while keepingg your finger on the vacuum nipple. If it does move, there is an internal leak in the unit. Replace it and everything should be ok.
This thread was discussed between 03/02/2002 and 22/02/2002
This thread is from the archives. Join the live MG MGB GT V8 Factory Originals Technical BBS now