55 replies to this topic

[madazv8]

So today's installment....

 

After talking to a mini specialist here in Sydney, which are thin on the ground, we went over my analysis so far.  He was really not convinced about valve guides being the cause.  I tend to agree as the fact there is zero smoke just didn't make it add up.  The head was refreshed recently also, and although the guides were not replaced they were checked for wear / free play.

 

Using my new gauge I blocked off the ECU vacuum pipe and rigged up a little t-piece.  1 end to the ECU, 1 end to the gauge, and 1 end to a syringe so I could manually pull a vacuum.  Started the car by inducing some vacuum with the syringe and then held it steady around 18 inHg.  Voila,  the car idles lovely. Connected MemsFCR and there were no fault codes and the O2 sensor was working as it should.  I seemed to be pulling pretty hard on the syringe to just get the right vacuum which was odd.  After about a minute I started to loose the idle and memsFCR was showing I was loosing vacuum,  so I pulled harder on the syringe.... then the MAP sensor just died.  Bugger. So looks like I destroyed the sensor with this test,  or maybe it was shot anyway and I just put it out of its misery. 

 

I don't see how a bad MAP sensor could make the vacuum dance around like described above... could it ?   Opened up the ECU, and the MAP sensor body is intact,  the rubber hose joining it to the nipple is not split.  I do note it seems to have been replaced already,  the solder joints are caked in dried flux which is not good.  I'll clean that up tomorrow and inspect under a microscope to make sure the flux hasn't eaten away any tracks. In the meantime I have ordered a new MAP sensor.  It is an easy replacement - only 3 pins to solder.  Sad part is it could take 6 weeks to get here. Not sure if this is going to be a fix,  or just get me back to square 1... time will tell.

 

Anyway here is a picture of the shocking solder job done by previous repair.

 

Attached Files

•  map.png   870.55K   3 downloads

[madazv8]

So an update on this saga. Car is running much better, but I would say still not perfect as reasons below...

 

After doing everything I could and still not getting to the bottom of this issue, I finally folded and admitted defeat and took the car to the mini specialist.  This was after replacing the MAP sensor I wasted as per above and on my other thread about replacing it.  It has been an interesting journey of learning and a few times could have led me down some crazy dead ends.

 

In summary,  the car would not idle well.  It was clear visually that it was throwing excessive fuel down the intake at idle.  Lambda was reading constant high, never went closed loop, and as a result it would only run in limp mode with bugger all power.

 

So.  A day after dropping the car off I get a TXT message with good news.  Car is running better. 

 

A spare ECU was plugged in,  no change.  ECU ruled out.

Confirmed no vacuum leaks.  I had gone over this countless times,  and was confident that was the case.

Checked fuel pressure.  Spot on. 

Measured valve lift with a dial gauge.  Seems somebody has fitted a high lift CAM ! We know the SPI doesn't like that ! and most likely why I have high MAP reading.

Un-plugged the coolant temp sensor and connected an original Rover item which he had on the shelf.

Place said coolant sensor in a mug of boiled water - immediate difference ! Car idles.

 

Frustratingly I had ruled out the coolant sensor as it was new, and it also seemed to be working according to MEMS apps.  The mechanic has had experience with recent replacement coolant sensors being bad.  Even though the readings on the MEMS apps seemed to look realistic,  somehow it does not marry up to the ECU.  My MAP reading is still too high,  and the car bogs a bit at mid range under hard throttle, but that is because of this camshaft.  I'll live with that for now.

 

So lessons learned:

 

When everything you read says the coolant temp sensor has the biggest effect.  Challenge your assumption that yours is ok. Especially if it is a replacement.

The Vacuum gauge  diagnosis I talked about earlier in this thread could have had me looking at CAM timing and valve guides etc.  These are written for carby engines.  Beware.

Don't pull stupid amount of vacuum on your MAP sensor,  expensive mistake.

 

Anyway,  It has been an education.  Thanks to all who took the time to offer wisdom.   I'm off for a 500km drive this weekend :)

[hungdynasty]

So an update on this saga. Car is running much better, but I would say still not perfect as reasons below...

 

After doing everything I could and still not getting to the bottom of this issue, I finally folded and admitted defeat and took the car to the mini specialist.  This was after replacing the MAP sensor I wasted as per above and on my other thread about replacing it.  It has been an interesting journey of learning and a few times could have led me down some crazy dead ends.

 

In summary,  the car would not idle well.  It was clear visually that it was throwing excessive fuel down the intake at idle.  Lambda was reading constant high, never went closed loop, and as a result it would only run in limp mode with bugger all power.

 

So.  A day after dropping the car off I get a TXT message with good news.  Car is running better. 

 

A spare ECU was plugged in,  no change.  ECU ruled out.

Confirmed no vacuum leaks.  I had gone over this countless times,  and was confident that was the case.

Checked fuel pressure.  Spot on. 

Measured valve lift with a dial gauge.  Seems somebody has fitted a high lift CAM ! We know the SPI doesn't like that ! and most likely why I have high MAP reading.

Un-plugged the coolant temp sensor and connected an original Rover item which he had on the shelf.

Place said coolant sensor in a mug of boiled water - immediate difference ! Car idles.

 

Frustratingly I had ruled out the coolant sensor as it was new, and it also seemed to be working according to MEMS apps.  The mechanic has had experience with recent replacement coolant sensors being bad.  Even though the readings on the MEMS apps seemed to look realistic,  somehow it does not marry up to the ECU.  My MAP reading is still too high,  and the car bogs a bit at mid range under hard throttle, but that is because of this camshaft.  I'll live with that for now.

 

So lessons learned:

 

When everything you read says the coolant temp sensor has the biggest effect.  Challenge your assumption that yours is ok. Especially if it is a replacement.

The Vacuum gauge  diagnosis I talked about earlier in this thread could have had me looking at CAM timing and valve guides etc.  These are written for carby engines.  Beware.

Don't pull stupid amount of vacuum on your MAP sensor,  expensive mistake.

 

Anyway,  It has been an education.  Thanks to all who took the time to offer wisdom.   I'm off for a 500km drive this weekend :)

I think the MEMS app just shows the data the ECU read from the temp sensor, we see what the ECU sees, do you mean the MEMs app shows normal temp sensor reading but the ECU read it as a fault? What exactly the problem is with the previous new sensor?

[madazv8]

I have the same confusion with this. When I looked at the MEMS app reading, everything looked normal with the coolant numbers from the sensor - it all looked normal and temp increased over time. That is why I assumed this was not the cause of any issues, and it was brand new !. In addition I had changed the inlet manifold and made sure it was flushed.   All I can say is,  when replaced with a working original factory sensor suddenly I have a working car.  I still have a slight over fueling issue (high MAP) due to the unknown high lift and overlap camshaft, but at least I can drive the car I bought at the start of this adventure. There is no history regarding who built the engine or what was in it,  thus why I had no idea about the camshaft. 

 

I agree that I expect MEMS app (or any other) would be telling what the ECU is seeing.  So I am baffled by this finding. 

Edited by madazv8, 28 May 2023 - 04:11 PM.

[andrewdjackson]

If your using the web.memsfcr.co.uk or the WebMemsFCR desktop app, I attempt to diagnose various faults from the running data. For example, it is expected that the coolant temperature will have reached 80C by a certain point in time. If the temperature isn't reach it flags a potential fault. Whilst an 88C thermostat is fitted, the ECU deems 80C to be the engine operating temperature.

As pointed out elsewhere in this thread, the app and the fault codes are guidance not absolute. The app is relaying what the ECU is reading from the sensors, one bad reading doean't mean it's faulty, look for trends and patterns. The ECU relies on those sensors to make adjustments via actuators which it assumes are working 100%. It's easy to forget that it's working on an old petrol engine that has wear and basic engine principles still apply.

 

I'm always open to ideas on how to improve the diagnostics, so do please post me any suggestions.

[genpop]

"when replaced with a working original factory sensor suddenly I have a working car."

What brand was the not working sensor.If the working curve is different from original ones you see the right temperature but you havent.In adition,

helpful is always a csv logfile.

[madazv8]

If your using the web.memsfcr.co.uk or the WebMemsFCR desktop app, I attempt to diagnose various faults from the running data. For example, it is expected that the coolant temperature will have reached 80C by a certain point in time. If the temperature isn't reach it flags a potential fault. Whilst an 88C thermostat is fitted, the ECU deems 80C to be the engine operating temperature.

As pointed out elsewhere in this thread, the app and the fault codes are guidance not absolute. The app is relaying what the ECU is reading from the sensors, one bad reading doean't mean it's faulty, look for trends and patterns. The ECU relies on those sensors to make adjustments via actuators which it assumes are working 100%. It's easy to forget that it's working on an old petrol engine that has wear and basic engine principles still apply.

 

I'm always open to ideas on how to improve the diagnostics, so do please post me any suggestions.

I hope I didn't suggest that your app was wrong,  it has been invaluable to me. Thank you so much for giving the community such a great tool.  I saw temperature readings increasing on the sensor so assumed all was good - I never verified if they were true readings.  My bad.

[madazv8]

"when replaced with a working original factory sensor suddenly I have a working car."

What brand was the not working sensor.If the working curve is different from original ones you see the right temperature but you havent.In adition,

helpful is always a csv logfile.

I am not going to go down the path of naming anybody.  It was purchased from a prominent parts place,  according to the mechanic this is not the first time he has seen this issue. Maybe I was unlucky. 

[pakgorgor]

Just read through the entire thread. I am having a similar issue on my 94 Spi. The car idles and runs fine for the most part but there is intermittent hesitation on acceleration. It has also developed an intermittent stalling (won't maintain idle), but would run fine for weeks until it happens again.

 

Plugged into MemsFCR,  MAP numbers are between 40-50, but lambda reads a constant 435mV regardless of engine temperature.

 

Things I have checked:

 

- No leaks on all vaccum lines (already converted to silicone).

- Fuel trap is clean.

- Lambda sensor relay bench tested and works.

- Coolant temperature sensor changed last year and it ran fine.

- New lambda sensor installed this year since the old one was from previous owner and who knows how old and I've had the car since 2011.

- No other faults from MemsFCR.

- With the ignition on, I am getting voltage from the relay to the lambda harness (blue/red and black wire).

- Wth the car running, lambda disconnected, using a multimeter to probe the two wires from the lambda sensor that would go to the grey and light green wires, I'm getting intermittent readings of around 100mV - 400mV while the engine warms up. Once warm, it stays around 300-400mV.

 

Is this a case of a bad lambda sensor? I didn't have MAP or stalling issues until I changed the lambda sensor this year.

[MatthewsDad]

I had intermittent hesitation on acceleration and the problem was two-fold which made diagnosis a challenge. In my case it was a failing coil (took a while before it finally packed up), which I replaced and things improved but still not right. The MEMS app highlighted a problem with my lambda sensor. Replaced with a Bosch and the car's running better than it has in a long time.

Is there any chance your new sensor has a fault?

[genpop]

@pakgorgor

 

It is discussed 1000 times here in the forum.Do you have an original bosch sensor?