Talk Morgan Forums Morgan Sports Cars Maintenance & Servicing Starting and running issue1991 Plus 4 M16 engine

Hi All

I’m new to the forum. My wife and I have a Plus 4 from 1991 with the M16 engine. We have an issue where we need to crank it to start first time (it runs for a second and dies otherwise), then once running cold it wants to die. Once warm it will run better, but leave it off for a few minutes and it need cranking again. If you stop the car and start again straight away it will fire up and continue to run. Even when warm there is a hesitation up in the revs, I can always find it at 2600rpm. I have already checked a lot out, the list below doesn’t include the things I checked before perhaps replacing them anyway and I might have missed some things off. The only item I haven’t managed to check is the fuel management ECU, I only recently realised that existed as I thought the programmed ignition system did all the work…

I have new injectors to try but I don’t think it will be that with what I have done with the existing ones. Does anyone have any other suggestions or know where I can get another fuel management ecu? It’s specifically the 14CU mkc10053 80203A.

- [x] Drained fuel, cleared the lines with new fuel and changed fuel filter.
- [x] Intake lead off of airflow sensor made no difference
- [x] Fitted new seals to injectors
- [x] Cleaned throttle plate
- [x] Fit new vac hoses
- [x] Changed distributer cap
- [x] Changed rotor arm
- [x] Clean crank position sensor
- [x] New crank position sensor
- [x] Crank position sensor checked via fast check tool at programmed ignition lead.
- [x] Check knock sensor via fast check tool at programmed ignition lead.
- [x] checked the fuses (taken each one out and put it back on)
- [x] changed the relays
- [x] Check fuel pressure
- [x] Check fuel pressure when cranking
- [x] Coolant temp sensor checked
- [x] Coolant temp sensor changed
- [x] Coolant temp sensor checked at Ecu plug
- [x] Coolant temp checked via fast check tool at programmed ignition lead.
- [x] Throttle position sensor checked with clean sweep on voltage
- [x] Cleaned injectors in a ultrasonic bath and cleaned through with carb cleaner (jet looks ok)
- [x] Engine cranked with fuel rail off and pointing into bottles to check operation.
- [x] Checked for air leaks using smoke machine. One tiny leak near throttle position sensor, not thought to be the issue.
- [x] New idle control valve fitted
- [x] Changed spark plugs.
- [x] Replaced coil
- [x] Checked HT leads
- [x] Changed HT leads
- [x] Check voltage drop to earth on engine
- [x] Voltage to coil from ignition checked
- [x] Programmed ignition swapped for another.
There isn’t a lambda sensor.

There is beers or another gift if someone can help me fix the issue.
Cheers
Ben

I see that you unplugged the MAF sensor, did you try cleaning or replacing it? Air flow sensor issues are a common cause of starting/idling problems.

MAFF electrically disconnected should allow the engine to adopt a fail safe running setting. check the distributer re wear and function vacuum etc Fuel lift pipe in the tank. excessive cranking could indicate fuel flowing back to tank, insert a one way valve. Should be able to plug a diagnostic tool in to see if you have any issues. How did you check fuel pump pressure? Injectors might look alright but need testing. You can do a rudimentary check by inserting each injector in turn into a plastic bottle and observing the spray pattern but if not confident get a fuel injection specialist to do it for you, remember its partially vaporised fuel so preferably outside!

Got called away! Do a compression check, dry then wet. Dry will give you a ball park figure and wet will ascertain if its piston issues ie. leak past. For wet I use about 10-15ml of paraffin or diesel. oil I find can give problems when getting it started later. Time I think to start checking it starting at a base level and then building from there.

Try

https://atcdrivetrain.co.uk/p/rover-820-fuel-injection-ecu-14cu-8.html

good luck, Ben regards John k

From what I understand the MAF is only for trimming for fuel and won’t have this impact.

I have a gauge mounted by the fuel rail so I can see pressure is good throughout. I ran the injectors with the fuel rail mounted off. It was difficult to see a pattern clearly but then I also connected them to a bottle of carb cleaner and activated them with a battery to see they sprayed nicely. I can still try the new injectors soon.

I haven’t done a compression test yet, it’s only on 40 odd thousand but I can borrow a compression tester.

I had contacted ATC and had a cost to check and or referb the ecu I have. I did contact them again after this and it seems they do have one they are working on that I could purchase… it’s of course more money but at least I would have one I know should be sorted, without just spending money on one I have which may be fine…

I would disconnect & block the pipe from the carbon canister. If the solenoid is staying open it will run rich.
Check aslo the vacuum pipe from the inlet manifold to the ECU

Ben these early ECUs are becoming rarer to the point that if you find one, buy it as a spare even if you don't need it, plus substituting will eliminate (or fix) the issue?

I thought the purpose of the MAF was to give the ECU a measure of how fast air was coming in. The TPS gives the signal as to how much the throttle is open (torque demand) so the ECU can then figure how how much fuel to add to get the best increase in revs to meet the desired torque indicated by the throttle setting. So I was thinking that if you've got an issue with the revs hesitating it might be that at those revs the ECU isn't getting a decent signal from the MAF sensor so it doesn't know how much fuel to add to get a clean burn.

But if this isn't how your car works then I'll wind my neck in, but overall I can't help thinking that malfunctions are much more likely to come from sensors operating in harsh environmental conditions on the ends of wires subject to flexing, abrasion and corrosion than a chunk of embedded software in an encapsulated box tucked out of the way somewhere? After all, the software can't change, it will still be running its algorithm on the data it is receiving from outside?