Mi16 (from Bx) Fuel Pump Relay Earthing Issues

[sport1901966 8]

I've been scratching my head trying to get the fuel pump to run off the ignition. Initially I assumed the worst and there was a major fault with my wiring.

 

Fortunately this isn't the case and on a (slightly) closer look I thought the relay was the issue - it was getting supply from the injection relay and was earthed, to a degree - multimeter showing 11.9v, so the only reason it wasn't activating, or so I thought, was a duff relay. I'll mention now that I had actually tested the relay by running a direct +ve feed and making a direct earth connection and I thought I remebered it working. After buying a new one, that definately works, the relay still wasn't energizing.

 

Back to work with the multimeter, there is definately over 12v (12.06ish) getting to the relay so the cause of the lower voltage must be excess resistance on the earth side of the relay. My concern is now where, checking CAPS showed that the earth side of the FP relay goes to the ECU pin 20, so I checked the resistance from the relay to the pin and there is none. I then thought maybe the ECU earths are the problem (pin 5,16 and 19), using the multimeter I checked these out and no resistance apparent. My worry is/was the ECU as the last link in the chain, but the fact any earth at all is being made shows the ECU is doing it's job so I ask - can anyone tell me what the hell is going on???

 

Cheers in advance

 

PS here is a diagram of the circuit in question (highlighted red) courtesy of CAPS

 

[sport1901966 8]

Has any had a similar problem, I really stumped with this one!.....

[welshpug 1,622]

are you getting spark fine?

[sport1901966 8]

Yep, the engine runs fine, I just have to bypass the relays with a direct +ve feed to the fuel pump to get fuel to the injectors. I was wondering whether my method of measuring resistances etc may be flawed - I have been touching a probe either side of the part of the circuit in question with the multimeter set on resistance. When the probes are in air it reads infinate resistance, if there is zero resistance i.e. a good connection between the probes then it reads 0 ohms.

[sport1901966 8]

Apologies for digging this up again after so long but I've come across more info.

 

It appears that the ECU only earths the Fuel Pump Relay in 2 situations:

1) For 1 second after the ignition is turned on

2) When it receives a signal from the Crank Position Sensor i.e. the engine is turning over, either by the starter or under normal running

 

I have the suspicion that it doesn't run in either situation so the problem must be common to both i.e. not the Crank Sensor

 

but I still wanted to confirm that the if the Crank Sensor isn't working the engine won't run full stop (with the fuel pump powered by alternative means of course). What I mean is, it would send a signal in a way that the ignition timing will be ok but the ECU still won't earth the Fuel Pump Relay?

 

Thanks,

Geoff

[welshpug 1,622]

yeah, if there's no TDC signal the ecu wont trigger the injectors or the coil.

[ade 4wd 11]

It sounds like the ECU is faulty because its not pulling pin 20 low to switch on the fuel pump relay.

 

Ade,

[sport1901966 8]

It sounds like the ECU is faulty because its not pulling pin 20 low to switch on the fuel pump relay.

 

Ade,

 

This was my initial concern and I'll have to take some more multimeter measurements but I was sure it was earthing the relay, just the quality of the earth wasn't good enough meaning there wasn't quite 12v through the relay. I intend to investigate this further tomorrow so will post up what i find!

[Gregg 0]

Voltage drop under load is the best way to test for resistance/poor connection. Basically you're mearsuring how much voltage is 'lost' when the circuit is working, positive lead at one end and negative at the other of the part of the circuit you want to test, a reading of any more than 0.5v could be considered excessive. Can be done on the live or earth side of the circuit.

[kyepan 291]

This was my initial concern and I'll have to take some more multimeter measurements but I was sure it was earthing the relay, just the quality of the earth wasn't good enough meaning there wasn't quite 12v through the relay. I intend to investigate this further tomorrow so will post up what i find!

Would be interested to hear what your results are.

 

Your information about the priming of the pump for one second when the ignition is turned on, was interesting, because i had been lead to believe that the MI16 ecu only switched on the pump when a cranking signal was received, (but the 8v did prime for a second). did you read that in caps?

 

I've been having some interesting situations where the pump takes time to get up to pressure and consequently the engine took many cranks to start, i am fairly sure it is the non return valve in the pump not maintaining pressure in the line.

 

Previously I'd noticed poor pump voltage, and really tried to clean up the fuel pump earth, and ran a new uninterrupted positive cable from the loom up near the battery (bx) to the pump itself, because the issue only manifested itself when the car was hot / cooling. Which caused me to surmise it was hot wire increased resistance that was causing the issue. However I didn't manage to make a jot of difference with that approach.

 

From what you've said it would appear that you have isolated the resistance to the ecu, i don't like thinking of the ecu as an object that is either duff or not, because of the following.

Integrated circuits are generally quite reliable, however these late 80's early 80's PCBs still contain plenty of discrete components such as capacitors and resistors, so a dry joint or failing component, may be the culprit here. Just want to say, my expertise in electronics is non existent beyond that comment.

 

 

J

[sport1901966 8]

Voltage drop under load is the best way to test for resistance/poor connection. Basically you're mearsuring how much voltage is 'lost' when the circuit is working, positive lead at one end and negative at the other of the part of the circuit you want to test, a reading of any more than 0.5v could be considered excessive. Can be done on the live or earth side of the circuit.

 

Would I be right in thinking to do this just set the multimeter to voltage and probe either side of the part of the circuit in question?

 

Would be interested to hear what your results are.

 

Your information about the priming of the pump for one second when the ignition is turned on, was interesting, because i had been lead to believe that the MI16 ecu only switched on the pump when a cranking signal was received, (but the 8v did prime for a second). did you read that in caps?

 

I've been having some interesting situations where the pump takes time to get up to pressure and consequently the engine took many cranks to start, i am fairly sure it is the non return valve in the pump not maintaining pressure in the line.

 

Previously I'd noticed poor pump voltage, and really tried to clean up the fuel pump earth, and ran a new uninterrupted positive cable from the loom up near the battery (bx) to the pump itself, because the issue only manifested itself when the car was hot / cooling. Which caused me to surmise it was hot wire increased resistance that was causing the issue. However I didn't manage to make a jot of difference with that approach.

 

From what you've said it would appear that you have isolated the resistance to the ecu, i don't like thinking of the ecu as an object that is either duff or not, because of the following.

Integrated circuits are generally quite reliable, however these late 80's early 80's PCBs still contain plenty of discrete components such as capacitors and resistors, so a dry joint or failing component, may be the culprit here. Just want to say, my expertise in electronics is non existent beyond that comment.

 

 

J

 

Yep that info came from caps. I would agree your problem definately sounds like a mechanical problem, maybe the pressure reg on the fuel rail isn't doing it's job properly if it turns out the fuel pump itself is ok...

 

After todays brief investigation it seems like it could will be the ECU. I tried a test where I set the multimeter to measure resistance between the earth side of the fuel pump relay (pin 20 on the ecu) and the main ecu earth (pin 19) - shown in the first pic below. In the 2nd pic the resistance is shown with the ignition off, the last pic shows the resistance immediately after the ignition is turned on - infinite. This says to me the ECU is not earthing the relay at all.

 

 

 

 

It's worth mentioning the ECU seems to operate fine in all other ways. For the moment I have got around it by directly earthing the FP relay but this means the pump runs all the time the ignition is on which isn't ideal.

[kyepan 291]

more detailed info on motronic

 

also i've just taken an ecu apart there are a few resistors but not really as many as i expected. Seven capacitors, and a couple of the black ones, who's name escapes me.

[Gregg 0]

Would I be right in thinking to do this just set the multimeter to voltage and probe either side of the part of the circuit in question?

 

Correct. Very simple and very effective.

[sport1901966 8]

Should I be looking for a new ECU or is it possible to repair mine? What do people think?

[kyepan 291]

Should I be looking for a new ECU or is it possible to repair mine? What do people think?

not sure if anyone has taken one apart on here, second hand ecu's are reasonably available, do you have a two row or three row? (two rows of pins or three, i can't quite see from the photo)

 

Yep that info came from caps. I would agree your problem definately sounds like a mechanical problem, maybe the pressure reg on the fuel rail isn't doing it's job properly if it turns out the fuel pump itself is ok...

 

I've tested the system thoroughly and the pressure regulator is working fine.

 

If you want to test your's, you'll need a gauge with a couple of bits of hose and some mole grips, i bought mine from ebay, it's only cheap, and has an inlet and outlet, so it's in line.

 

I think the tests are listed on my "tale of woe" thread... but in short.

 

to test the general pressure on the feed side of system, from pump to rail,

 

Put the gauge inline on the inlet to the rail, this is usually the bit from the fuel filter to the drivers side end of the rail. Run the car up and check the pressure according to caps, open the throttle and check it manages to keep up.

 

turn the car off, pressure should remain in the line for quite some time (minutes)

 

if it does not, it might be either the non return valve in the pump or the regulator, but more likely the non return valve.

 

to test the non return valve you need two sets of mole grips, to isolate the rail, one after the regulator and one before the gauge, put both on simultaneously and then kill the engine.

 

first undo the one after the regulator, if the pressure falls away, the regulator is not holding pressure (and would probably not have passed the first test to check the fuel pressure).

 

remove the second set of grips to test the non return valve.

 

Unplug the vacuum take off to check if the pressure regulator is not leaking, give it a sniff to check if the diaphragm is leaking. Plug the vacuum back in again run the car up, and un plug it, you should see a 2lb drop (i think or gain i can't remember).

 

finally

Put the mole grips after the gauge and before the rail (but only briefly) and check that the pressure rises above standard and the fuel pump is good and strong.

 

 

This is how I isolated the non return valve as being the issue with my car.

 

J

[matthewm 1]

Seen as your direct earthing method works the pump, then I'd say yes to a new ECU - be nice if you could find one to try first to make sure though.

 

Interesting reading, broadens my understanding as I'm currently modifying my MI loom and moving the relays inside the car. Saw 'CAPS' mentioned in the thread, wondering if anyone could send me a link as I cant get anything on search, as I could do with a wiring diagram - for a 3 row ECU. Many thanks.

[sport1901966 8]

If you pm me your e-mail add I can send it to you, it only covers the ML4.1 (2-row) but still has a lot of useful info on how the system works.

 

It would make life much easier if I could rule it out by trying another but for now it will have to stay as it is till I find time!

 

For info on the 3-row try PMing 'veero' I think his username is, really helpful guy and he re-built a 3-row loom IIRC

Edited February 13, 2010 by sport1901966