Jump to content


Photo

Mpi Not Switch To Closed Loop


Best Answer Ruckus , 07 May 2018 - 02:48 PM

Yes I’ve been testing everything today.
And I have a result, it was the ECU and the old CAM sensor.

I had a tatty ECU from somewhere which I dropped in today and managed to recode the immobiliser and bingo Cam sensor sync and runs smooth.

So it was 3 things. Lambda and Cam sensor and the ECU.

I tested the resistance on the old cam sensor and it was nothing. So maybe this damaged the ECU in some way?
I don’t know but it’s running smoothly now with all sensors reporting as expected. Go to the full post


  • Please log in to reply
39 replies to this topic

#16 Ruckus

Ruckus

    Ice Ice Baby!!!

  • Members
  • PipPipPipPipPipPipPipPipPip
  • 2,071 posts
  • Location: Top of the hill in Kent
  • Local Club: SMOC

Posted 30 April 2018 - 09:56 PM

It is worth checking the sensor is getting 12v on the heater circuit

 

Checked and even replaced the relay unit.



#17 spiguy

spiguy

    Camshaft & Stage Two Head

  • Members
  • PipPipPipPipPipPipPipPip
  • 1,537 posts
  • Location: UnderTheCar

Posted 02 May 2018 - 06:12 AM

You will only get a reading from the O2 sensor when disconnected at the plug, if the O2 sensor is hot enough to operate. When connected, the O2 sensor receives 12V to power the heater. When driving, there is sufficient exhaust gas passing the sensor to heat it to an operational temperature. When idling, the sensor may not get hot enough due to only a small amount of exhaust gas acting on it hence the reason for the heater element. The correct way to measure the output directly is actually to tap into the wiring so that you can measure the output with the sensor still connected at the plug and so still getting power to its heater element. I would make sure that you do this before condemning the sensor as otherwise it may just be too cold.

 

If you disconnect the sensor then this will not stop the car going into closed loop, the ECU will just use an assumed value for the O2 level in the exhaust gas. I was of the understanding that the only reading that is used to determine when to switch to closed loop was the coolant temperature, however that is based on the SPi.  I believe that once passed 70 DegC (I may be incorrect on the exact value) the system moves to closed loop.  One thing which seems very odd is that the intake air temperature is reading 79DegC!  I don't know if the car will use limp mode in the case of a faulty input air temp sensor, I would have thought it would just ignore the value and use an assumed value (and throw a code). You could try to disconnect it.

 

I do wonder about your cam sensor as it is showing Input was low. The crank sensor should be enough to allow the engine to run but I would think the car may use limp mode in the absence of the cam sensor-  again I would expect a code if this is not providing a signal. Have you checked for fault codes?  Can you shed any more light on the cam sensor message in the results? Is this saying there is no signal?

 

At this point my guess is cam sensor. Input from FlyingScot on here would be helpful, he is very knowledgeable about injection minis and will be more familiar with the MPi unit.


Edited by spiguy, 02 May 2018 - 06:14 AM.


#18 Ruckus

Ruckus

    Ice Ice Baby!!!

  • Members
  • PipPipPipPipPipPipPipPipPip
  • 2,071 posts
  • Location: Top of the hill in Kent
  • Local Club: SMOC

Posted 02 May 2018 - 09:01 AM

The output i posted was after a 'fast paced' drive so i don't think that the lambda was too cold or the intake air temp sensor reading was wrong. The engine is a 1380 and has a SC stainless LCB so its hot back there! I have a spare Air temp sensor so i can swap that (In fact I basically have a spare everything now).

 

The Cam Shaft single reads as OK (the sensor has been changed for new) but the position reading stays at 0 degrees. I would have thought that due to the 0 reading there is no history so it shows as 'too low'.

 

I was told by my local MG/Rover specialist that until the car goes in to 'Close loop' operation it does not use the Cam sensor and uses a default map (limp mode).

 

I have a new Lambda sensor in hand now and i can try the original Cam sensor and see if that changes anything. 



#19 Sprocket

Sprocket

    Great on Injection faults

  • Members
  • PipPipPipPipPipPipPipPipPipPip
  • 7,266 posts
  • Location: Warrington
  • Local Club: Manchester Minis

Posted 02 May 2018 - 12:12 PM

Disconnecting the lambda sensor, the ECU will still read about 0.4v. With a good sensor connected exposed to free air instead of the exhaust and the engine running it will read low volts, about 0.2v. (Sensor body must be grounded to the engine). You will then also be able to feel the sensor tip get very hot. I offer caution here as it can get hot enough to leave a permanent mark!!

Closed loop kicks in very early during warm up, its more to do with ensuring condensation is burned off the sensor than anything else cold start warm up will go into closed loop and target lambda 1 all the way up to full warm.

Car wont start without cam sensor, ECU needs to know where ignition event is on cylinder 1 for the injectors to trigger at the right time. Theres an article somewhere with Martin Theaker the designer of the MPI system talking about the development of the system. He showed the Rover development team how the system could work, but without the cam sensor it would only start the engine half the time.

From memory, I read some Rover technical documentation where its suggested that if the cam sensor was to fail while the engine was running, the engine would stay running. However if the cam sensor had failed and the engine was already stopped, it wouldnt start. This is because the ECU cannot determine which end of the engine cycle it detects TDC at. Spark events dont matter much as they are of the wasted spark type, its just the injection events, no cam sync no injection events.

Limp mode for IAT uses fixed value of 35degrees again from memory. CLT is similar but uses fixed value of 60degrees. Which is why its easy to start the engine hot than it is cold when one of the sensors is in limp mode. What initiates limp mode is either an open circuit or dead short or any reading that might be very very near the trip values (Siberian winter for example).

#20 pete l

pete l

    One Carb Or Two?

  • Members
  • PipPipPipPipPipPipPip
  • 1,355 posts
  • Location: East of France

Posted 02 May 2018 - 12:25 PM

Just read this 

 

Attached File  cam sensor.jpg   36.93K   17 downloads



#21 Ruckus

Ruckus

    Ice Ice Baby!!!

  • Members
  • PipPipPipPipPipPipPipPipPip
  • 2,071 posts
  • Location: Top of the hill in Kent
  • Local Club: SMOC

Posted 02 May 2018 - 03:06 PM

Sprocket: I can confirm that the engine keeps running if the Cam sensor is unplugged while running. I did this the other day while troubleshooting. VERY hot back there!

 

I haven't tried firing it up without the Cam sensor being already unplugged. I never have a problem starting the car, in fact it drives really nice using this default/limp mode.

 

 

pete: Interesting that makes me again worry about the Swifttune Cam it has...

 

 

So what am I'm looking at now? so much information but I'm none the wiser it seems... LOL



#22 Sprocket

Sprocket

    Great on Injection faults

  • Members
  • PipPipPipPipPipPipPipPipPipPip
  • 7,266 posts
  • Location: Warrington
  • Local Club: Manchester Minis

Posted 02 May 2018 - 05:51 PM

Funny story regarding the MPi camshaft sensor pickup😆 I use the MPi block for the 16v conversion, primarily for the cam position sensor. Now with the 16v conversion you take the MPI camshaft and machine it for a special 720 degree duration 😂 (take ALL the lobes off🤪) leaving the cam sensor pickup alone. It worked with the lobes, so it should work without the lobes right?🤔 nope not a chance. I spent two or three evenings swapping out sensors, wiring, even swapping cam and crank inputs but could not get that cam sensor to work without the other lobes on the cam. I ended up fabricating a bracket and picking up one of the screws on one of the cam sprockets, works every time.

Having seen that bulletin before, Id like to think over the last 22 years youd have heard someone making noise about it especially if the car was problematic for that long. That said, from my slightly different circumstance in experience above, I can see how it could be a problem, but is it really causing your open loop fueling problem? Im not so convinced.

#23 Sprocket

Sprocket

    Great on Injection faults

  • Members
  • PipPipPipPipPipPipPipPipPipPip
  • 7,266 posts
  • Location: Warrington
  • Local Club: Manchester Minis

Posted 02 May 2018 - 06:26 PM

Does the lambda sensor respond to any changes at all?

I would leave the old sensor in the exhaust and plug the new one into the loom. Sit the sensor so the metal body is in contact with the cylinder head, make sure its securly placed so it cant fall anywhere. Also keep it from touching anything that might melt. Start the engine and carefully feel the sensor warm up indicating the heater is working, if its not heating up stop the test, and resolve that issue. If it does heat up, then and only the see what voltage reading you get. It should show a low voltage in oxygen rich air indicating a lean condition. You could then try saturating the sensor with butane from a gas cigarette lighter. Ive never tried it myself, but it seems to be something some have used to test the sensor for a rich condition, just dont set the car or yourself on fire. Id be looking at what voltage the ECU reports on the diagnostic rather than measuring it with a multimeter as this will prove the entire circuit and that the sensor does in fact do something and your problem lies elsewhere

#24 Sprocket

Sprocket

    Great on Injection faults

  • Members
  • PipPipPipPipPipPipPipPipPipPip
  • 7,266 posts
  • Location: Warrington
  • Local Club: Manchester Minis

Posted 02 May 2018 - 06:50 PM

Re cam position at zero degrees. Id this not to do with variable valve timing which the Mini doesnt have?

Cam position is at zero degrees because it occurs in the same place all the time.

There lots of info in there that is not relevant, such as oil temp, the Mini ECU does not use this input.

Remember, the MPI Mini ECU was adapted from the KV6 ECU which was designed to operate two banks of three cylinders. There is even more stuff in its brain thats not used at all.

#25 Ruckus

Ruckus

    Ice Ice Baby!!!

  • Members
  • PipPipPipPipPipPipPipPipPip
  • 2,071 posts
  • Location: Top of the hill in Kent
  • Local Club: SMOC

Posted 02 May 2018 - 08:11 PM

Sprocket i cant thank you enough for your insights in to this so far.

 

So did a bit on the car tonight (only get an hour of day light).
 

Things have changed for some reason... for the worst!

The car now idles very rough? and the cam shaft signal is now showing as invalid? I couldn't get the car up to full temperature (don't want to upset the neighbors too much) but in the warm up when it normally sounds good it sound lumpy all of a sudden.

I swapped the cam sensor for the original one i took off a few weeks ago and it still shows invalid.

 

Oh and i swapped the intake air temperature sensor just in case... 

 

cyLs40G.jpg

 

What am i looking at here an earthing problems somewhere? its a brand new loom ! 

 

 

I'll try and get under the car on Friday and try what you suggested with the new Lambda sensor.


Edited by Ruckus, 02 May 2018 - 08:16 PM.


#26 Sprocket

Sprocket

    Great on Injection faults

  • Members
  • PipPipPipPipPipPipPipPipPipPip
  • 7,266 posts
  • Location: Warrington
  • Local Club: Manchester Minis

Posted 02 May 2018 - 08:47 PM

Under the car? SC manifold? is the lambda sensor in the 'Y' section under the car? How have you lengthened the wires?

 

It certainly seems the cam sensor is not working for some reason and the cam position angle is more to do with angle since last synchronisation I really don't know.



#27 Ruckus

Ruckus

    Ice Ice Baby!!!

  • Members
  • PipPipPipPipPipPipPipPipPip
  • 2,071 posts
  • Location: Top of the hill in Kent
  • Local Club: SMOC

Posted 03 May 2018 - 06:16 AM

Yes it has the SC manifold, the wires are extended from the loom.
Plug cut then longer wires soldered in. Every link was heat strinked and then the whole extra part covered in a silicon heat sleve. Wires were checked for continuity several times.
But this extra bit was used in the old loom so it has the old plug on it, so I’m also suspicious of this.
I plan to take out the extra bit and go back to normal length with the new plug and use the higher branch mount point for the sensor.

#28 Ruckus

Ruckus

    Ice Ice Baby!!!

  • Members
  • PipPipPipPipPipPipPipPipPip
  • 2,071 posts
  • Location: Top of the hill in Kent
  • Local Club: SMOC

Posted 03 May 2018 - 06:17 AM

by the way the cam position does seem to flick between 0 and the 65535 never noted this before.

#29 Ruckus

Ruckus

    Ice Ice Baby!!!

  • Members
  • PipPipPipPipPipPipPipPipPip
  • 2,071 posts
  • Location: Top of the hill in Kent
  • Local Club: SMOC

Posted 06 May 2018 - 12:28 PM

tumblr_mno8155o7z1sq9agdo1_400.png

 

New Bosch Lambda and another new Cam Sensor and we seem to be working.

 

O2 is now reading and working. free air on the head as you said Spocket its reads 0.02 down to 0 

 

Installed it reads as you would expect 0.01 up to 0.9xx.

 

 

So it looks like when the car went in to the MG/Rover specialist the problem was the link to the Cam sensor AND the Lambda. So when they tried a different Lambda the Cam link was still creating a problem. Therefore they thought the Lambda was not the problem.

 

Why is the Cam sensor dropping off... looks like i have a little bit of an drive shaft oil leak which is flicking oil up and getting in to the plug connection.

I've covered the connection with a sleeve of heat shrink which i hope should keep it oil free until i can get the oil seal sorted.

 

I've not had the emission checked yet, but I just listening to the car running it sounds so much smoother now.

 

Success I think.


Edited by Ruckus, 06 May 2018 - 01:31 PM.


#30 Ruckus

Ruckus

    Ice Ice Baby!!!

  • Members
  • PipPipPipPipPipPipPipPipPip
  • 2,071 posts
  • Location: Top of the hill in Kent
  • Local Club: SMOC

Posted 06 May 2018 - 06:12 PM

Oh no, spoke too soon

Lambda working happily but the cam sensor keeps losing sync/single then it runs like crap again. It was fine earlier then a few hours later is not..

Checked the plugs etc all looks ok..

No idea again.




0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users