Large negative fuel trims

Message
Author
rich r
Junior Member
Posts: 224
Joined: Tue Aug 26, 2014 2:48 pm
Location: South end of North Yorkshire

Re: Large negative fuel trims

#21 Post by rich r » Mon Jan 26, 2015 9:19 am

With it being an auto, it could well be hiding some issues by changing down earlier maybe. It does seem quite eager to drop a gear when you accelerate, far more so that any other auto I've driven. But I haven't driven any other Honda before so don't know if this is normal or not. Fuel economy isn't as good as I expected (on LPG or petrol), but that's obviously very subjective. I'll hopefully get chance to go out and see what the lambda sensors are saying this evening.
-
2015 Mitsubishi Outlander PHEV (electric/petrol)
1991 Mitsubishi Pajero 2.5 LWB (WVO)
2008 Volkswagen Caravelle (Diesel)
(previously 2004 Subaru Outback 3.0Rn auto (LPG))

LPGC
Installer
Posts: 3319
Joined: Wed Sep 29, 2010 6:01 pm
Location: Yorkshire
Contact:

Re: Large negative fuel trims

#22 Post by LPGC » Mon Jan 26, 2015 7:21 pm

Just going back to your question, ways to check the cat, can sometimes tell if the cat is blocked because the exhaust gas is much hotter than you'd expect.

CRV automatics do seem eager to kick down easily in my experience, at this point nothing seems unusual there.

What economy have you been getting?

Simon
Full time LPG installer
Servicing / Diagnostics / Repairs to all systems / DIY conversion kits supplied with thorough tech support
Mid Yorkshire
2 miles A1, 8 miles M62,
http://www.Lpgc.co.uk
Twitter https://twitter.com/AutogasSimon
07816237240

rich r
Junior Member
Posts: 224
Joined: Tue Aug 26, 2014 2:48 pm
Location: South end of North Yorkshire

Re: Large negative fuel trims

#23 Post by rich r » Tue Jan 27, 2015 8:32 am

About 20-22mpg on LPG doing the same commute (very much extra-urban, mostly 50-70mph A-roads with no town driving) that my Vectra 1.8 Dualfuel was getting closer to 40mpg on (around 350 miles on average to a 43 litre tank of LPG). On petrol it's around 25-28mpg. Now I know that the Vectra engine is a lot more efficient than the Honda, but given the CRV is supposed to do 37mpg extra-urban, I was hoping for a bit nearer 30mpg on LPG. This could of course be normal for a CRV auto, and the published figures aren't as realistic as Vauxhall's. Or maybe my driving style in the Vectra was more fuel efficient than it is in the CRV.

I didn't get chance to monitor the lambda sensors last night but will try tonight. I also want to put a mechanical pressure gauge on to see what the manifold vacuum actually is to rule out that MAP sensor being faulty. It's the way that once it's up to temperature that it all 'seems' normal that's a little odd. Could a lambda sensor that's contaminated or not heating properly maybe produce these symptoms by indicating overrich when it's not?
-
2015 Mitsubishi Outlander PHEV (electric/petrol)
1991 Mitsubishi Pajero 2.5 LWB (WVO)
2008 Volkswagen Caravelle (Diesel)
(previously 2004 Subaru Outback 3.0Rn auto (LPG))

rich r
Junior Member
Posts: 224
Joined: Tue Aug 26, 2014 2:48 pm
Location: South end of North Yorkshire

Re: Large negative fuel trims

#24 Post by rich r » Tue Jan 27, 2015 8:47 pm

OK - stuff to report :)

I set it monitoring Bank 1 Oxygen Sensor 1 and Bank 1 Oxygen Sensor 2, along with the STFT and let it idle. I noticed something very curious:

When cold, and having switched to closed loop mode, both lambda sensors are cycling either side of 0.5V. Sensor 1 is going between about 0.1V and 0.8V. Sensor 2 is going between 0.2V and 1.2V. STFT is around -20%:

Image

As it warmed up, the STFT reduced, lambda sensor 1 is cycling more slowly between rich and lean, but look what's happened to sensor 2:

Image

As it gets properly up to temperature (85C), the STFT sits close to 0%, sensor 1 is cycling between 0.1V and 0.8V - speeding up when I accelerate gently to 2000rpm (you can see where I increased the rpm on the picture). But sensor 2 has gone off the top of the scale - it's somewhere over 1.275V. Incidentally when I came off the accelerator it stalled. But after another couple of minutes it was happier for me to increase the revs and drop them again.

Image

When it's stable and up to temperature the MAP is around 41kPa, and when I increase the revs to 2000rpm it drops to 30kPa.

So does that look like something is very wrong with either the post-cat lambda sensor or the cat, or am I misinterpreting these results? Could a blocked or unwell catalytic converter cause this behaviour in the post-cat sensor - perhaps due to being too hot?
-
2015 Mitsubishi Outlander PHEV (electric/petrol)
1991 Mitsubishi Pajero 2.5 LWB (WVO)
2008 Volkswagen Caravelle (Diesel)
(previously 2004 Subaru Outback 3.0Rn auto (LPG))

rich r
Junior Member
Posts: 224
Joined: Tue Aug 26, 2014 2:48 pm
Location: South end of North Yorkshire

Re: Large negative fuel trims

#25 Post by rich r » Wed Jan 28, 2015 3:58 pm

I should maybe have done a bit more reading. So the post-cat sensor should only be oscillating when the cat isn't working, ie it's not using reacting the oxygen and carbon monoxide etc in the exhaust with the aid of the catalyst. If the cat is working then it should be a constant voltage - which maybe it is once it's warmed up, hence the voltage wandering off the top of the scale.

Or do the traces still look wrong?
-
2015 Mitsubishi Outlander PHEV (electric/petrol)
1991 Mitsubishi Pajero 2.5 LWB (WVO)
2008 Volkswagen Caravelle (Diesel)
(previously 2004 Subaru Outback 3.0Rn auto (LPG))

rich r
Junior Member
Posts: 224
Joined: Tue Aug 26, 2014 2:48 pm
Location: South end of North Yorkshire

Re: Large negative fuel trims

#26 Post by rich r » Mon Feb 02, 2015 9:29 am

I can't find an EGR valve on this engine (K20A4) - so I guess that's not the problem.

However, a bit of digging around on Honda Civic forums (same engine as the CRV obviously) seems to imply that exhaust valve clearances (or VSR) can cause high negative fuel trims, high MAP and stalling at idle. Given this engine has done almost 140,000 miles and had an LPG conversion without Flashlube (it had done about 15,000 miles on LPG before we bought it and fitted lube) - would this sound likely? The valve clearances and compressions are all good.

I'd rather not get the head off without being reasonably confident that VSR is the issue...
-
2015 Mitsubishi Outlander PHEV (electric/petrol)
1991 Mitsubishi Pajero 2.5 LWB (WVO)
2008 Volkswagen Caravelle (Diesel)
(previously 2004 Subaru Outback 3.0Rn auto (LPG))

LPGC
Installer
Posts: 3319
Joined: Wed Sep 29, 2010 6:01 pm
Location: Yorkshire
Contact:

Re: Large negative fuel trims

#27 Post by LPGC » Mon Feb 02, 2015 5:44 pm

I think these engines are well known for valve clearance issues... I also seem to remember Dai (Classicswede) posting on that topic on this forum a while ago.

Would make sense to have negative trims with valve clearance issues if the engine refers to map regards fueling around idle.. Clearance issues could mean the engine needs a higher map to run, the ECU will equate the higher map to a requirement for greater fueling but the lambda and trims will work to lean the mixture back to correct.... But you've checked the clearances. Hmm!

Instead of getting the head off to check for VSR (if that's the direction you're going in), maybe you could visit a garage that has a bore scope?

Simon
Full time LPG installer
Servicing / Diagnostics / Repairs to all systems / DIY conversion kits supplied with thorough tech support
Mid Yorkshire
2 miles A1, 8 miles M62,
http://www.Lpgc.co.uk
Twitter https://twitter.com/AutogasSimon
07816237240

rich r
Junior Member
Posts: 224
Joined: Tue Aug 26, 2014 2:48 pm
Location: South end of North Yorkshire

Re: Large negative fuel trims

#28 Post by rich r » Tue Feb 03, 2015 7:41 am

That might be what I need to do. I've checked the clearances twice now, and they're spot on still. Do those O2 sensor graphs look normal to you?

Standing next to the car the other day I think I heard a leak in the exhaust near the front of the car - maybe it's sucking in air? Although I'd have though that would lead to it seeing too much oxygen in the exhaust, ie a weak mixture, not rich.
-
2015 Mitsubishi Outlander PHEV (electric/petrol)
1991 Mitsubishi Pajero 2.5 LWB (WVO)
2008 Volkswagen Caravelle (Diesel)
(previously 2004 Subaru Outback 3.0Rn auto (LPG))

LPGC
Installer
Posts: 3319
Joined: Wed Sep 29, 2010 6:01 pm
Location: Yorkshire
Contact:

Re: Large negative fuel trims

#29 Post by LPGC » Tue Feb 03, 2015 9:23 am

rich r wrote:Standing next to the car the other day I think I heard a leak in the exhaust near the front of the car - maybe it's sucking in air? Although I'd have though that would lead to it seeing too much oxygen in the exhaust, ie a weak mixture, not rich.
it would... The noise you're hearing may be normal, plastic manifolds and throttle bodies can be noisy, it's often possible to hear the air rushing in.

Have I (or anyone) previously mentioned checking the EGR valve if fitted?

The pre cat sensor graph looks normal. More difficult to say regards the post cat sensor as the info doesn't make sense - seems the post cat is reading a rich mixture under some circumstances where the pre cat always shows correct... If the mixture is rich post cat but correct before, this begs the question where is the extra fuel coming from? Exhaust valve stem seal issues would be one possible, cam timing / compression might be another, broken cat might be another, but the first question should be are the readings correct or is there some misinterpretation of the OBD data in your OBD software...

Regards cam timing, this is a Vtech engine? The effective cam profile is supposed to change with rpm.. Is the cam timing sticking in high rpm mode?

Simon
Full time LPG installer
Servicing / Diagnostics / Repairs to all systems / DIY conversion kits supplied with thorough tech support
Mid Yorkshire
2 miles A1, 8 miles M62,
http://www.Lpgc.co.uk
Twitter https://twitter.com/AutogasSimon
07816237240

rich r
Junior Member
Posts: 224
Joined: Tue Aug 26, 2014 2:48 pm
Location: South end of North Yorkshire

Re: Large negative fuel trims

#30 Post by rich r » Tue Feb 03, 2015 10:00 am

I couldn't see where the EGR valve is (was looking for something like a vacuum actuator on a pipe coming up from the exhaust manifold area), but I'll have another look this weekend. But yes, you did mention it and I want to check it isn't sticking.

It is Vtech, but that's all working properly. I had that all checked when I got the timing chain replaced.

Many thanks - some more things for me to look at anyway.
-
2015 Mitsubishi Outlander PHEV (electric/petrol)
1991 Mitsubishi Pajero 2.5 LWB (WVO)
2008 Volkswagen Caravelle (Diesel)
(previously 2004 Subaru Outback 3.0Rn auto (LPG))

rossko
Installer
Posts: 8094
Joined: Wed May 23, 2007 6:51 pm
Location: Wiltshire

Re: Large negative fuel trims

#31 Post by rossko » Tue Feb 03, 2015 7:31 pm

LPGC wrote:
rich r wrote:More difficult to say regards the post cat sensor as the info doesn't make sense - seems the post cat is reading a rich mixture under some circumstances where the pre cat always shows correct... If the mixture is rich post cat but correct before, this begs the question where is the extra fuel coming from?
It's not a mixture sensor. It is an oxygen sensor. Interpreting lack of oxygen as "extra fuel" is an error. When a cat is working, you would expect less oxygen in the output than in the input.
www.blazegas.com

rich r
Junior Member
Posts: 224
Joined: Tue Aug 26, 2014 2:48 pm
Location: South end of North Yorkshire

Re: Large negative fuel trims

#32 Post by rich r » Wed Feb 04, 2015 8:14 am

Thanks Rossko, that's the impression I got after posting the graphs then doing a bit more reading as I was confused. So it looks like the cat and sensors are working (as there's less oxygen in the exhaust after it once it's warmed up)- but maybe it could still be partially blocked?

I'm going to try to find the EGR valve this weekend (if it's not too cold), and also check the PCV system is working as that's been suggested. Also I'll try to spot any leaks in the exhaust under all the heatshields. If there is a leak in the cat section (which is where it sounds like there might be), then I guess I might as well look into replacing it which would hopefully resolve the issue if it is due to a partially blocked cat.

The issue is definitely only when it's at idle. If it (naughtily) keep my right foot on the accelerator slightly when holding the car stationary with the brakes so it stays at about 800rpm (rather than dropping to idle at 650-700rpm), then it's rock stable. And of course once it's up to temperature there are no detectable issues at all.
-
2015 Mitsubishi Outlander PHEV (electric/petrol)
1991 Mitsubishi Pajero 2.5 LWB (WVO)
2008 Volkswagen Caravelle (Diesel)
(previously 2004 Subaru Outback 3.0Rn auto (LPG))

LPGC
Installer
Posts: 3319
Joined: Wed Sep 29, 2010 6:01 pm
Location: Yorkshire
Contact:

Re: Large negative fuel trims

#33 Post by LPGC » Wed Feb 04, 2015 8:44 am

rossko wrote:
LPGC wrote:
rich r wrote:More difficult to say regards the post cat sensor as the info doesn't make sense - seems the post cat is reading a rich mixture under some circumstances where the pre cat always shows correct... If the mixture is rich post cat but correct before, this begs the question where is the extra fuel coming from?
It's not a mixture sensor. It is an oxygen sensor. Interpreting lack of oxygen as "extra fuel" is an error. When a cat is working, you would expect less oxygen in the output than in the input.
Yes you're right mate. What do you reckon to the graph of his post cat readings?

Simon
Full time LPG installer
Servicing / Diagnostics / Repairs to all systems / DIY conversion kits supplied with thorough tech support
Mid Yorkshire
2 miles A1, 8 miles M62,
http://www.Lpgc.co.uk
Twitter https://twitter.com/AutogasSimon
07816237240

rossko
Installer
Posts: 8094
Joined: Wed May 23, 2007 6:51 pm
Location: Wiltshire

Re: Large negative fuel trims

#34 Post by rossko » Wed Feb 04, 2015 9:49 am

I don't think the graphs help either way, save to show that the sensors still work.
www.blazegas.com

rich r
Junior Member
Posts: 224
Joined: Tue Aug 26, 2014 2:48 pm
Location: South end of North Yorkshire

Re: Large negative fuel trims

#35 Post by rich r » Tue Apr 21, 2015 7:22 pm

Just to bring this old thread to a satisfactory conclusion, should anyone come across similar symptoms and find it with a search in the future:

As suggested by several people, it was indeed a partially blocked catalytic converter. It passed the MOT fine, but then I replaced the cat to try to solve this stalling problem. It's done 140,000 miles and looked very rusty so I thought it was worthwhile changing anyway just because it means only half the exhaust system will get an advisory note for being very rusty:)

And now - no more stalling, the engine feels much smoother and more powerful. :) Fuel trims are stable even after 10 minutes of idling.
-
2015 Mitsubishi Outlander PHEV (electric/petrol)
1991 Mitsubishi Pajero 2.5 LWB (WVO)
2008 Volkswagen Caravelle (Diesel)
(previously 2004 Subaru Outback 3.0Rn auto (LPG))

classicswede
Installer
Posts: 2583
Joined: Fri Oct 21, 2005 8:56 pm
Location: North Wales
Contact:

Re: Large negative fuel trims

#36 Post by classicswede » Wed Apr 22, 2015 9:41 am

That does make sense. I've come across a lot of cats that have failed and blocked off the exhaust flow causing lack of power. They do not last forever. Did you replace the lambda sensors at the same time?
http://www.classicswede.com/
http://www.classicswede.co.uk/LPG/cat17 ... 10800.aspx

LPG installer Anglesey North Wales

DIY LPG Kits

07824887160

rich r
Junior Member
Posts: 224
Joined: Tue Aug 26, 2014 2:48 pm
Location: South end of North Yorkshire

Re: Large negative fuel trims

#37 Post by rich r » Wed Apr 22, 2015 10:50 am

I did the front lambda sensor, not the rear one. Mainly because the old one had rusted itself into the cat section so much that there was no way it was likely to come out alive :) From the graphs I posted before I'm sure that there was nothing actually wrong with the old sensor, but it wasn't coming out. The rear one though was no problem.

It's interesting that once warmed up it was working well enough to pass the MOT emissions test too, but when cold was causing enough of a restriction to cause the ECU to fail to regulate the fuelling properly at idle. I didn't have chance to chop the old one up to inspect it, so we'll just have to assume it was partially blocked based on the observed symptoms.
-
2015 Mitsubishi Outlander PHEV (electric/petrol)
1991 Mitsubishi Pajero 2.5 LWB (WVO)
2008 Volkswagen Caravelle (Diesel)
(previously 2004 Subaru Outback 3.0Rn auto (LPG))

Post Reply