36 replies to this topic

[orcadian]

Hi,
Just another thought - if plug colours 1 and 2 are different from 3 and 4, have you thought of swapping the two injectors over to see if the fault is still the same or transferred to the others. Perhaps you already had injector flow rates measured? I know you had them cleaned/checked.
Regards

[minimowta]

Hi,
Just another thought - if plug colours 1 and 2 are different from 3 and 4, have you thought of swapping the two injectors over to see if the fault is still the same or transferred to the others. Perhaps you already had injector flow rates measured? I know you had them cleaned/checked.
Regards

Thanks for the reply orcadian.

I'm actually getting plugs 1 and 4 different from 2 and 3. This is what prompted me to check the coil pack and the earths because of this "wasted spark" feature. As I understand it the same electronics fires 1 & 4 and then 2 & 3.

I checked plugs 1 & 2 this evening and actually they are now quite similar, No.1 being only slightly sootier than No.2 so maybe my cleaning of the contacts last weekend did some good. Pesky misfire is still there though!

[orcadian]

Hello again and apologies for throwing a spanner in the works - I thought I had carefully read the posts but read a non existent bit about the plug numbers - am I getting old? You have checked all sensors and earths - would I be right in thinking that it is NOT easily possible to borrow another ecu, unless you use the immobiliser unit and key fobs too? You said at the beginning that it has been on-going for 18 months - was it gradual or very sudden? Does crank position sensor get info from just 1 or 2 places on flywheel? I can't find any ref for that. Also cam position is measured - could that be an issue. Water temp, air temp, IACV and map sensor surely would play up on all pots.

Regards,

[minimowta]

Thanks for the reply.

I use the keyfob/imobiliser so I don't think an ecu swap is possible I'm afraid.

I checked the plugs from 1 & 2 yesterday and they were quite similar so I'm thinking the problem lies elswhere. Interesting what you say about the crank and cam sensors but wouldn't a problem there effect the engine throughout. Above 2k rpm it runs really sweet - I LOVE driving it then, the SW5 cam makes a subtle but noticeable difference and it is a real joy to drive (like a mini should be) This misfire is below 2k and most noticable if I hold the throttle at a constant 1.5k rpm. If I blip the throttle i.e floor it and instantly let go - this causes a miss/splutter before it catches and revs up then it "meets the pedal" on the way back and the revs drop to a near stall before recovering to tick over. Tick over sounds OK. That floor throttle-splutter-rev-near stall-recover sequence I just mentioned is not a long process, it takes about 2 seconds to go through.

I'm wondering if there may be a problem with the throttle position sensor, but I think this only has an effect if the throttle is changed. In a steady state (constant revs) situation I don't think this sensor plays a significant roll (sits back and waits to be shot down!)

Going back to the crank and cam sensors, these were taken out when the engine was rebuilt so if there was a mechanical problem with them they would have been replaced and I think they are OK electrically because they checked out OK on the diagnostics checks.

Cheers

[orcadian]

Hello again,
I agree with what you say about the TP sensor sitting and waiting for a changed input, but if it's a normal variable resistor (not been inside one) with a wiper like a volume control, then it's just possible that at the 1500 rpm position there is a bad contact which IS varying the signal. If you can't borrow one elsewhere, then I have a spare you could try but it would need sending from here in Orkney. It (TPS) must change resistance with angular position, so it could probably be tested with a multimeter - analogue preferably so you could see the needle jump if there is a fault at a particular place. I had a problem some years ago with the throttle pot on my XJS, but it misbehaved with no real pattern like yours.

Regards,

[minimat]

The injectors feed two cylinders which is why they are not a true mpi system i.e i injector per cylinder as per more modern cars so swapping them round will make no difference,getting them cleaned might though especialy if they have done a high milage. cylinders 2 and 3 will always be a bit richer than 1 and 4 simply because they run hotter being in the middle of the head and having more metal around them.
It does sound like you have exhausted most of the electrical possibilitys,i would still go back to basics and check compressions,even though it is a low milage enging you could still have a fault like a badly seated valve etc.
This will be a lot cheaper and easier than swapping expensive sensors etc!
Just noticed you have a pressure boost valve fitted,i would remove it as it could be doing more harm than good and see what happens.

Edited by minimat, 11 June 2010 - 05:13 PM.

[mpihornet]

The injectors feed two cylinders which is why they are not a true mpi system i.e i injector per cylinder as per more modern cars so swapping them round will make no difference,getting them cleaned might though especialy if they have done a high milage. cylinders 2 and 3 will always be a bit richer than 1 and 4 simply because they run hotter being in the middle of the head and having more metal around them.
It does sound like you have exhausted most of the electrical possibilitys,i would still go back to basics and check compressions,even though it is a low milage enging you could still have a fault like a badly seated valve etc.
This will be a lot cheaper and easier than swapping expensive sensors etc!
Just noticed you have a pressure boost valve fitted,i would remove it as it could be doing more harm than good and see what happens.

Not too sure about mixture being heat related, its more to do with injector firing sequence and fuel charge robbing. 1-3-4-2 or more accurately 3-4 2-1 is the firing order, there are 2 injectors not 4 so as they each fire its 3 then 4 and 2 then 1. what this means is that 3 gets some of the fuel meant for 4 as and 2 gets some of the fuel meant for 1. This is why its easier to get bigger power from tbi than Mpi as TBI more accurately simulates a single point fuel source line an su which has constant fuel mixture which is good for siamese port heads. In the vizzard book its made clear that twin carbs on a cooper (like MPI) are not the best option for a series motors and he councils for big single throat carbs feeding into both ports. I think the temperature variance must be pretty close on the a series because its so short. In some straight sixes its number 6 that gets hot because its the last to get cold water.

The problem you are having is clearly electronic. if you had a problem with it before rebuild and it still exists, then its a sensor, or a wire rubbing somewhere and grounding and giving a dodgy reading. Inspect the harness at all the points where it touches the body or engine. THen consider replacing the cam sensor, crank sensor or throttle sensor, do them one at a time, not all at once.

IT cant hurt to do a compression check but the engine would be lumpy if it were week. YOu might have a tight valve but again its not consistant with symptoms. Its worth checking those things though as those tests are free and replacing items costs money.

[minimowta]

Right - done some more checking but it still aint right!

Checked the compression as minimat and mpihornet suggested (and I did take all the plugs out to do the test) and got 13.5bar across all 4 cylinders. I don't know what it is supposed to be but 13.5 sounds pretty impressive to me and, more importantly, they are all the same. A quick look at the plugs showed that 1 & 4 are still slightly sootier than 2 & 3 but still better than they were before I took the coilpack off.

I took the throttle position sensor off and plugged it into my digital multimeter and slowly rotated the core. The resistance increased steadily with no jumps or holes evident. I did this several times and really slowly and it appears to be working perfectly!

I also checked over the wiring looking for possible false earth contacts as mpihornet suggested and found nothing wrong. I moved any part of the loom that might be making contact with the bodywork so moved the loom about a bit which might have made a difference to a faulty connection but this made no difference to the chuffing misfire!

I'm really at a loss as to where to look next I know I said all the sensors have been checked OK but I'm wondering if there might be a fault which just slips under the radar. After a bit of research on these very informative threads I'm thinking of changing the sensors in the following order. What do you guys think?

Lambda sensor - I'm thinking the delayed response I'm getting when I floor the throttle and the stepped decay in the revs might be due to a dodgy lambda. I had problems in the past with the fan coming on at inappropriate times and this was put down to a faulty plug on the water temp sensor causing the car to go into "safe mode" which richens up the mixture. This might have sooted up the lambda sensor and reduced its ability to respond quickly to changes.

Camshaft sensor - only because others on these hallowed threads have had similar problems to me and changing this sensor fixed them!

MAP sensor - for same reason as camshaft sensor.

Any help or advise would be terrific before I start shelling out on sensors!

Cheers

[orcadian]

Hello again,
Glad you were able to test TP sensor, as you say, what's left? I always like to find the problem by testing rather than just replacing, especially if you can't borrow stuff. I have a spare Map sensor which I have not tested - but it wouldn't take long to swap onto my engine. It must again (like TP) change resistance but with varying pressure, so can you connect it to a large syringe (50ml plastic medical type) and draw the air slowly from it, causing partial vacuum while watching the resistance, as you did with TP? With a couple of cycles like that it should show a blip if there's a fault. Next the Cam sensor, (don't have spare one!) Surely that must only give digital info? as part of the cam passes a Hall effect type sensor it will suddenly switch from off to on and back. This might require an excitation voltage - does it have more than 2 terminals? Can't think of a way of testing Lambda without substitution, but have just trawled the net for possibilities (http://www.lambdapow...ic_index.asp#q9) Try the link. Do you have access to an oscilloscope or know someone with one? Another thought!! Do you have a timing light that you could hook up and tape to the bonnet to see, one plug at a time, if the sparks disappear when the 1500 rpm problem occurs?

Keep us all informed, we are bound to have something like this ourselves in the future! We need a file with all test parameters for each sensor and a procedure for testing these wonderful but annoying little beasties.

[minimowta]

Hello again,
Glad you were able to test TP sensor, as you say, what's left? I always like to find the problem by testing rather than just replacing, especially if you can't borrow stuff. I have a spare Map sensor which I have not tested - but it wouldn't take long to swap onto my engine. It must again (like TP) change resistance but with varying pressure, so can you connect it to a large syringe (50ml plastic medical type) and draw the air slowly from it, causing partial vacuum while watching the resistance, as you did with TP? With a couple of cycles like that it should show a blip if there's a fault. Next the Cam sensor, (don't have spare one!) Surely that must only give digital info? as part of the cam passes a Hall effect type sensor it will suddenly switch from off to on and back. This might require an excitation voltage - does it have more than 2 terminals? Can't think of a way of testing Lambda without substitution, but have just trawled the net for possibilities (http://www.lambdapow...ic_index.asp#q9) Try the link. Do you have access to an oscilloscope or know someone with one? Another thought!! Do you have a timing light that you could hook up and tape to the bonnet to see, one plug at a time, if the sparks disappear when the 1500 rpm problem occurs?

Keep us all informed, we are bound to have something like this ourselves in the future! We need a file with all test parameters for each sensor and a procedure for testing these wonderful but annoying little beasties.

Hi orcadian, thanks for the reply.

Never thought of using a timing light to check each plug in turn, I think that output may be analogue so may show up a weak spark aswell - Nice one!. I've got one tucked away somewhere from the old days when I had carb minis....... nostalgic memories of only having to top up the dashpot oil! ....... I'll defo give that a go when I can find it. Just remembered I also have a colourtune somewhere so maybe that will have another airing too.

Good idea about testing the MAP sensor. I'll get the stuff together to give that a go.

The crank sensor has two connections so must be digital as you say - magnetic switch or something?

I don't have access to an oscilloscope so unfortunatly can't follow that one up. Sprocket did a method for testing the lambda sensor but it involves messing about with the fuel line and I've got a bit of a phobia about that (and brakes as it happens) just not confident enough s'pose

Cheers

[mpihornet]

in the 80's and 90's lambda sensors were meant to be replaced every 30k miles. is that the case now? I see it has a heater, and thats to make it work from cold, old ones had just he sensing wires and were screwed if you ran too rich for any lentgh of time. but the heater should also blow of the soot. Soot will come off at 400c with enough air, and if the lambda sensor is zirconia based and its not very oxygen conductive below 600 degrees c and thats what the heater is for.....

If you did soot up the lambda sensor, then your cat may be hurting. If you take it out you should be able to see light through it. A slight increase in back pressure maybe causing your problem, but i doubt it, Im really clutching at straws with you. If your cat were plugged then flat out perf would suck and the wfo performance would really suffer, and your saying wfo performance is good so, maybe we can scratch that...., but you could take the cat off, give it a try for the cost of a cheap gasket (this is the part where i wish i had read that you decatted it already but i skipped through it....) although, a plugged cat is not consistent with good performance in high rev ranges, so Im thinking its unlikely. lambda sensor is about 50 squid at minispares. as its a limited lifetime item, (or used to be, i dont know anymore) it may be worth replacing as part of the "tuneup" rational anyway. What happens when you disconect the lambda sensor? does it still run? there will likely be a "no signal" dafe mode for it too, and what is that supposed to be me wonders????

Ps, which wire came off on which temp sensor, because my fan is running all the time and its the brown plug on the thermo housing that i clocked with a hammer whilst installing the mpi into the hornet... I dont want it to run in safe mode if its rich as I too don't want to have it plug the cat.

[minimowta]

A quick update after this evenings investigations.......

I took the MAP sensor off and checked the resistance at different suction pressures. There are 3 connectors on the sensor and I was getting 14.8kohms between the middle one and each of the outer ones and 2.9kohm between the two outer ones. No amount of suction made these resistances change I'm afraid I can't say what the pressures were but I would expect some difference.

Next I plugged in my timing light and checked cyls 2 and 4. Both were the same brightness and both never missed a beat even with the 1.5k rpm misfire.

Whilst the timing light was plugged in I shone it at the end of the crank - no timing marks of course but I could see the nut on the end "stationary" and it seemed to be advancing and retarding as I operated the throttle. At the 1.5k misfire it was all over the place. The question now is.....

Is my problem just a timing issue at 1.5k revs causing the mis or is the ecu changing the timing to try and correct another fault?

Cheers

[minimowta]

Ps, which wire came off on which temp sensor, because my fan is running all the time and its the brown plug on the thermo housing that i clocked with a hammer whilst installing the mpi into the hornet... I dont want it to run in safe mode if its rich as I too don't want to have it plug the cat.

It was the brown one on the thermostat connecting to the coolant water temp sensor. Apparantly the wiring to this causes problems aswell, there have been several threads on this subject. During one visit to the garage they found that the big black multiplug next to the relay on the bulkhead to the left of the air filter as you stand looking at the engine was corroded. They cleaned that with some contact cleaner and the fan has never come on since.

Cheers

[orcadian]

Hello again,
It sounds like you might be closing in on the problem - but just one question - when you shone the timing light at the bolt, did it go 'all over the place' on all cylinders at the 1500 rpm point or just one? Like you, I'm surprised that the Map test showed no resistance change - I will do the same test on my spare and report back. When I think more deeply about the 1500rpm bit and the Map sensor, you would really think that between say 1400 and 1600 rpm, there would be very little difference in absolute pressure, so the MAP sensor is surely not playing a part. Is there anyone who can properly test the ECU out there?

[Blatherskite]

Hi

There's a lot of complex testing going on here, but what about the cam? You mention a specific cam, and I assume this is not the standard cam? Some of the mini tuning books, also threads on the Forum, mention that some of the hairier cams can give poor idle and roughness at lower revs - could this be the problem here - especially as the ECU cannot b preprogrammed to suit special cam? Sounds like the motor runs like a dream at higher revs, so to be a bit brutal, is this actually a problem?