1 2 3 ... 5
EvanB (Forum Supporter)
EvanB (Forum Supporter) GRM+ Memberand MegaDork
1/10/21 10:54 a.m.

In the continuing saga of me being confused by the MINI I am taking a break and trying to think through it. It will still start, rev up to 1500 or so and sound good for 4-5 seconds then drop into a low idle and not take throttle. Since I have both oxygen sensors working now it will go into closed loop if I keep it running long enough and short term fuel trim goes to +30 and it idles well and takes throttle but breaks up when revving it, when it settles back down into an idle the STFT climbs back to +30 and it will idle fine. I let it run like this for about 30 minutes and no change to LTFT. 

It has a cheap aftermarket header, I know there are leaks in the header and cat flanges. The cat is also installed backwards so the rear o2 is in front of the cat. I am either going to take that off and weld up the cracks and turn the cat around or get an aftermarket OEM style header/cat. I'm not convinced that is where my issue lies though since I would think it would still run fine but exhaust leaks would make the sensors run lean and it would be way too rich, not just starting to run right after it adds a lot of fuel. 

Possibly jumped time? There aren't any bad noises from the timing chain area that I can hear and it isn't easy to verify the timing without pulling it all apart. There are no faults found for cam or crank signal or engine sync. 

There are no fault codes in the system since fixing both o2 sensors. The front one reads 0v or close to it when it is running, even when it is running well, the rear reads 0.1-.15v. 

I have fixed all the intake vacuum leaks and smoke tested to verify. 

Fuel pressure is correct with a gauge on the rail. 

 

Jesse Ransom (FFS)
Jesse Ransom (FFS) GRM+ Memberand UltimaDork
1/10/21 11:14 a.m.

I'm going to end up claiming a lack of coffee again, and folks are going to suspect there's more wrong with my diagnostic skills than trying them in the morning, but...

How on earth can the front O2 sensor be reading 0 when it's running well? I would have thought that running well would be nearly proof of not being that lean nonstop.

Moreover... If the rear sensor is in front of the cat, shouldn't it just read the same as the front sensor, since it's effectively in the same location?

I thought broadly that the front sensor was to manage fueling, and the rear to more or less verify cat functionality.

Are we sure the sensors and their attendant electronics are good?

I want it to be that the front sensor is having some kind of issue, the rear sensor is giving sane values but the ECU doesn't know about them because it's getting zeroes from the front one, but that still doesn't explain why it needs all that trim to run.

Pressure's good, but is it maintained under load and as the pump warms up? Does it have enough volume at that pressure?

EvanB (Forum Supporter)
EvanB (Forum Supporter) GRM+ Memberand MegaDork
1/10/21 11:22 a.m.

In reply to Jesse Ransom (FFS) :

Yes, the front o2 is confusing me as well, I figured once the STFT actually started doing something and it was running well that would pick back up but it was still reading 0 or very close. On first start when it runs well it reads a believable number (0.5 or similar, can't remember exact), then when the idle drops the sensor voltage very quickly drops to around 0.1 then less as it runs longer. 

It was extended to work with the aftermarket header, the initial fault code I had was traced to two of the solder splices rubbing through the heat shrink and contacting. I don't entirely trust the splices to be good. I am hesitant to buy a new o2 sensor and have to cut and extend the harness again to work with this header. I was hoping to find a stock manifold in the junkyard but they cut all the cats out which is understandable. 

I haven't left the fuel pressure gauge on long enough to see what it is doing when it finally runs well and I can rev it. 

bentwrench
bentwrench SuperDork
1/10/21 11:50 a.m.

Some O2 sensors can be ordered with long leads.  (Innovate?)

 

No upstream exhaust leaks will be tolerated......

EvanB (Forum Supporter)
EvanB (Forum Supporter) GRM+ Memberand MegaDork
1/12/21 4:50 p.m.

I got the exhaust manifold unbolted and the cat off. About half of the manifold bolts were finger tight...

The universal cat was not welded to the flanges well and there are many obvious air leaks that I marked. Unfortunately the orientation is wrong to put it in with the o2 sensor after the cat without the seam of the heat shield hitting the floorpan. 

This is the new o2 sensor that I put in, it certainly doesn't look like it's running extremely lean.

Now I need to decide if I want to get a new aftermarket (or try to find used) OEM style manifold/cat. Try to weld up the leaks in this one and leave the post-cat o2 before the cat. Or some other option. 

The front o2 sensor is a universal bosch that was spliced onto the original connector so that doesn't fill me with confidence in its quality. 

Jesse Ransom (FFS)
Jesse Ransom (FFS) GRM+ Memberand UltimaDork
1/12/21 6:37 p.m.

Any chance you could combine the operations of getting it welded up properly with reclocking it so you could turn it the right way 'round?

Sorta sounds like everything was in such catastrophic condition that just trying to get everything pointed the direction it's supposed to go with no holes in it would be a really good thing to achieve just to see whether it's the same or improved or still bad but with a more meaningful set of conditions.

And yeah... I'm very dubious of that O2 sensor between its description, appearance, and readings.

But that's all easy for me to say! I didn't have leave my keyboard or spend anything!

EvanB (Forum Supporter)
EvanB (Forum Supporter) GRM+ Memberand MegaDork
1/12/21 7:13 p.m.

Unfortunately I would have to put a new bung in it to get it to work. There just isn't a way to orient it to get the o2 in the right place and fit in the tunnel. 

I welded up all the seams. My current plan is to put it back in like it was (o2 before the cat) and put the new sensor in the front position and the bosch universal in the rear position. I'll unfortunately have to extend the harness on the new sensor but if I bought another new one I would have to extend it anyway. We'll see if that makes any difference. 

hoffmaw1
hoffmaw1 New Reader
1/13/21 9:03 a.m.

I know from personal experience these cars do not like universal O2 sensors precat. I put new Bosch ones pre and post cat and the car threw a code for the front sensor and ran a little rough. I replaced it with the recommended one, I think an NTK, and it has been fine and no more codes. It seems when it comes to certain sensors on these cars you need a specific one or it won't work. I will say I also had a significant leak before the back O2 sensor and it never threw a code.

EvanB (Forum Supporter)
EvanB (Forum Supporter) GRM+ Memberand MegaDork
1/14/21 4:53 p.m.

I welded up all the leaks in the cat flanges, used copper rtv on the flange gaskets, copper spray on the manifold gasket and put everything together. I went to the junkyard and snipped some o2 connectors from a mini and made an extension harness (and found out my torpedo heater can heat shrink terminals very quickly) to put the new NTK sensor in the #1 spot. 

Buttoned everything up and no change. It tries a little harder to smooth out the idle after start but ultimately the same result. 

After some research the fuel filter is on the opposite side of the tank from the fuel pump. I had previously thought it was in the same cannister unit as the fuel pump. I popped the cover off and it has definitely not been changed recently, if ever. New fuel filter is on order so I will try that when it shows up. 

Robbie (Forum Supporter)
Robbie (Forum Supporter) GRM+ Memberand MegaDork
1/14/21 5:18 p.m.

ok, random question. Is it POSSIBLE that the rear o2 is plugged into the front o2 harness location and the ecu is using the rear o2 physically as the front one?

Also, I normally relate throttle response issues to airflow sensors, or possibly throttle sensors, if it doesnt rev at all that doesnt seem like o2. 

Robbie (Forum Supporter)
Robbie (Forum Supporter) GRM+ Memberand MegaDork
1/14/21 5:21 p.m.

also, have you used an IR gun on the exhaust ports to see if one or two cylinders are the ones going lean? Maybe there are fuel injectors that are plugged up or paritally plugged. 

EvanB (Forum Supporter)
EvanB (Forum Supporter) GRM+ Memberand MegaDork
1/14/21 5:59 p.m.
Robbie (Forum Supporter) said:

ok, random question. Is it POSSIBLE that the rear o2 is plugged into the front o2 harness location and the ecu is using the rear o2 physically as the front one?

Also, I normally relate throttle response issues to airflow sensors, or possibly throttle sensors, if it doesnt rev at all that doesnt seem like o2. 

Not possible with the way the wiring harness is routed, the rear o2 plug is not close tot he front one. 

It uses two MAP sensor, no air flow sensor. Looking at live data and the sensor fault data there should be no issues (the map sensor have been replaced before I bought it). The throttle plate moves as it should with throttle pedal input, just has no affect on the running. 

Fuel injectors were also replaced while the previous owner threw the parts cannon at it, although with the quality of work I have seen so far anything is suspect. I don't have anything to measure exhaust port temps. 

EvanB (Forum Supporter)
EvanB (Forum Supporter) GRM+ Memberand MegaDork
1/15/21 5:09 p.m.

No change with a new fuel filter. The old one was very black.

I let it run long enough to go into closed loop then let it idle for about 45 minutes. When warming up and running poorly in open loop the o2 sensor voltage was 0.12 on the front and 0.19 on the rear. When it did go into closed loop finally and STFT up to +30 the o2 sensor voltage dropped to 0.04 on the front and 0.00 on the rear. No change to LTFT, I can drive it a bit at very light throttle but it dies under anything more than that. 

Pete. (l33t FS)
Pete. (l33t FS) GRM+ Memberand MegaDork
1/15/21 6:32 p.m.

Mark a line across the harmonic damper and run the engine for a minute or two.

 

Apparently a slipping damper is a common issue causing extreme lean running.  Like, evry iATN hit for lean problems turns out to be a bad damper.  I am not sure exactly what sorcery in the fuel calculation algorithms causes this, maybe it works like Volvo and uses throttle position for fuel calculation at low loads and the blower not turning means the throttle has to be open more.  But even that doesn't make sense because it should be running super rich, not lean.  Maybe has to do with how there are two MAP sensors?  I dunno.  I Ain't Tested Nothin' suggests it's probably a damper.

Subscriber-unavailabile
Subscriber-unavailabile HalfDork
1/15/21 7:16 p.m.

I'm no expert on Minis, but aren't these prone to ECU failures? I've seen lots on Craigslist cheap for that reason.

EvanB (Forum Supporter)
EvanB (Forum Supporter) GRM+ Memberand MegaDork
1/15/21 7:22 p.m.

In reply to Pete. (l33t FS) :

That's one of the early things I checked. I didn't mark it and run it but I did take the belt off and try to move the damper and it didn't slip or have much back and forth movement. I'll mark it and try that. The damper is a common problem to cause limp mode but the EML light is not on and limp mode is not activated when viewing the data. 

EvanB (Forum Supporter)
EvanB (Forum Supporter) GRM+ Memberand MegaDork
1/15/21 7:23 p.m.
Subscriber-unavailabile said:

I'm no expert on Minis, but aren't these prone to ECU failures? I've seen lots on Craigslist cheap for that reason.

The ECU doesn't report any issues when scanned with INPA (BMW diagnostic software). 

Dusterbd13-michael (Forum Supporter)
Dusterbd13-michael (Forum Supporter) MegaDork
1/15/21 7:31 p.m.

Could it be running so rich its reading lean? I know with megasquirt that was a thing. 

Jesse Ransom (FFS)
Jesse Ransom (FFS) GRM+ Memberand UltimaDork
1/15/21 11:57 p.m.

In reply to Dusterbd13-michael (Forum Supporter) :

I keep wondering that, but then it also sounds like it runs better as the short term fuel trim dials up. I'm at a loss.

Robbie (Forum Supporter)
Robbie (Forum Supporter) GRM+ Memberand MegaDork
1/16/21 9:40 a.m.

The very poor throttle response is still telling me the ecu doesn't see it opening. Or fuel is very limited.

Do you see the map sensors working in live data? Like are they reading high vacuum at idle and then dropping as you open the throttle? You did fuel pressure test, did you do fuel volume test? The challenge saab once had a ground issue on the fuel pump and it had just enough fuel to start and idle but it wouldn't rev up. No easy pressure test port though so we didn't test pump pressure, I assume it would have been low.

You've probably done this but looking at all 4 spark plugs should tell you if the injectors are working similarly at least.

Is it a return less fuel system? If so you may also be able to watch the live data on the fuel pump signal.

EvanB (Forum Supporter)
EvanB (Forum Supporter) GRM+ Memberand MegaDork
1/16/21 10:08 a.m.

TPS values and throttle request in live data seem good. The map sensors are reading normal values in live data at idle but I haven't checked to see how it changes with throttle application. When it is running poorly it is very low vacuum then once it finally gets into closed loop and idles well it is normal vacuum. 

I haven't tested fuel volume. It is a returnless system but I haven't found any data to view on the fuel system. 

I haven't looked at the plugs yet, I am planning on doing a compression test just to see what it is so I can look at the plugs then. 

EvanB (Forum Supporter)
EvanB (Forum Supporter) GRM+ Memberand MegaDork
1/17/21 7:29 a.m.

After doing some research the map sensor values should be close to 1 volt at idle vacuum and closer to 4volts under higher load/less vacuum. Key on engine off one is showing 1.38 v and the other (i think pre supercharger) is 3.8v. That seems like a large difference to me but i don't know if that is wrong or not. The values both move equally when it is started but still a few volts different between them. They are both aftermarket sensors, which i have rrad don't always have the best results.

I have some screenshots of sensor data that i can add later.

Pete. (l33t FS)
Pete. (l33t FS) GRM+ Memberand MegaDork
1/17/21 7:46 a.m.

Key on engine off, if they are the same sensor (meaning both 2 bar or whatever) then they should be reading the same.

 

Definitely sounds like you've found an issue.  I wonder if both sensors are getting an accurate ground or 5 volt reference.

 

A 1 bar sensor should read around 4.5v KOEO and a 2 bar should read around 2.5v.

EvanB (Forum Supporter)
EvanB (Forum Supporter) GRM+ Memberand MegaDork
1/17/21 8:40 a.m.

Well guess I didn't read far enough. The intake manifold sensor is 2.5 bar and the other one is 1.2 bar? 120kpa range.

Pete. (l33t FS)
Pete. (l33t FS) GRM+ Memberand MegaDork
1/17/21 8:49 a.m.

Huh.   It sounds like the pre supercharger unit is reading correctly but I have doubts for the other one.  A 2.5 bar sensor (assuming a 5vref)  would be 1.38v at about 75kpa give or take, not 94-95.

 

2.5 is a WEIRD range.  The only other time I've seen one of those is the onboard sensor for Megasquirts.

1 2 3 ... 5
Our Preferred Partners
ndcOikiIbzKn0gfCoD7tFVWoUbiIhCS3ABeK8KeLD3NJBydvWpSwj8zpui95K3bd