Yet another rv8 rich idle headache..

Yet another rv8 rich idle headache..

Author
Discussion

Unlucky48

Original Poster:

71 posts

51 months

Tuesday 4th February 2020
quotequote all
Hi people,
I’m nearly at the pulling my hair out stage with my freshly built engine and am hoping some of you may have some ideas to help me fix it..

Firstly it’s not in a tvr but my much loved rrc lse.
Symptoms are it’s running rich at idle (1.8% co) from cold start to fully hot and I can’t seem to effect a change.
It starts/runs/ drives ok and I’ve managed to do 500 miles in it to run it in. once off idle it drives fine if a little lacklustre considering the spec but that’s probably the standard fuel map/ auto box combination.
I am certain the cam timing is spot on after treble checking it before the engine went into the car, piper vernier camchain set to cam makers timing.
Using my old fashioned ears it sounds and feels just a tiny bit off/ lumpy at idle but I can’t detect why.
Ignition system is all good, new lucas bits, magnecore leads and new bpr6es plugs (rather black..)
Fuel pressure as it should be and all ecu side tests seemed fine, the only exception being the afm overshooting to .58 volts with ignition on and taking five or so seconds to settle where it should, substituted it for a newer one which doesn’t misbehave.
I’ve had a look in rovergauge, no faults but once warm the stepper motor seems out of spec with it being only a few percent and at one point 0%, long term fuel trim is at max with original afm and came down to 75% with better one, still no change.
I am a total newbie to rovergauge but after looking at other people’s screenshots my fuelling seems to be really rich with both on the lambdas and with pulse widtharound 3.8 at warm idle?
I’ve yet to confirm both compression and vacuum tests.
I did just for the hell of it thinking this might be a vacuum/ inlet air leak try blowing smoke down the inlet.. of course I didn’t get a nice steady stream of smoke from anywhere useful but while blowing something inside the motor made a click kind of sound similar to a tin plate/ glass jar test button popping and again on releasing pressure and I can’t for the life of me think what might do that?...
any advice would be really welcome!

Classic Chim

12,424 posts

149 months

Tuesday 4th February 2020
quotequote all
Might have been throttle butterfly pushing shut. Where did you pump smoke into engine from.
My first thought is do you have the correct chip in the Ecu for the engine spec.
Is your ECU from a Landrover or Tvr.

Unlucky48

Original Poster:

71 posts

51 months

Tuesday 4th February 2020
quotequote all
Thanks classic chim,
I thought that too so second time I blocked off the stepper motor pipe, held butterfly open with same weird click sound present..
method was simple, laughable and wasn’t really expecting a result but an excuse to light up in frustration. I simply took off the hose clip for the air flow meter, formed a good seal around the rubber inlet pipe with my mouth and blew a face full of smoke down it..
I’ve had the ram housing back off to double check and can’t see what it could be or any obvious sign of a problem although the butterfly seems properly closed when closed as in when cleaning it off the car no carb cleaner would pass it so there’s no tiny gap as per spec.

Chip and ECU are correct for the car and original engine spec but not the new engine spec. I get that I’ll probably have to seek out an affordable genius to get the fuelling right later but I’d assumed it would still idle and run ok ish for now?
It’s now on bigger (than rr) valves, ported and flowed heads, tubular manifold &stainless system without cats but with lambdas (was used with the previous now coffee table engine)
and a real steel cyclone cam.
I’ve tried on on the original ecu with ‘correct’ blue resistor and green no cat resistor. Also tried a second identical ecu with no change and just for laughs a 92’ 3.9 ecu (using blue resistor)
When I tried the non cat resistor I couldn’t effect any change in emissions even if I pulled the maf trim down to 0.8v?
I’m thinking it’s an air leak (although the above with the non cat tune confuses me)but god knows where and having only just built it I’m sure everything is spot on but obvs something isn’t..

Classic Chim

12,424 posts

149 months

Tuesday 4th February 2020
quotequote all
An affordable genius who maps the Cux for a bespoke engine is what you’ll need at some point.
Luckily we TVR owners have the likes of Joolz from Kits & Classics ( look him up)
He is one of the few people who regularly MAP powerful RV engines running the Lucus CUX with excellent results.
I’m not much help here as I gave up trying to fathom the CUX and went for a different ecu on my rebuilt 450 engine.
Cost of rebuilds had me mindful I wanted the best timing and fuelling. Mine over fuelled and felt a bit rough at low revs often.

I’m not sure you’ll create enough air pressure to find air leaks without some decent kit.
There are a lot of CUX experts on here
Wait a bit and someone might come up with a plan smile


Unlucky48

Original Poster:

71 posts

51 months

Tuesday 4th February 2020
quotequote all
Wise words mr classic!
If I had the time and funds I’d probably enjoy getting stuck in to a different fuelling and spark system but I have all too many vehicles to divide my time between and was hoping that the cux would do a reasonable job especially as the car isn’t a tvr but a load hauling adventure car. The engine I removed due to split liner/ block was an incorrect for my lse 3.9 and ran just fine with mot happy emissions for 8 years running on the 4.2 with cat ecu set up although when I split it to keep the block I was shocked at how much carbon had built up in it!.
I did try to talk my daughter into letting me sell it and buy a chim but she said ‘how would we go camping in that?!’ We are planning on doing the nc500 soon and a run to the Le Mans classic if I can get it running properly.
Will wait in hope of any other ideas and look up the recommended cux genius for a later date

blitzracing

6,387 posts

220 months

Wednesday 5th February 2020
quotequote all
Loads to work on here;

1) Blue map- I think this is the low compression map for crap fuel, so if you are in the UK you would be better off with white for a catalysts engine or green for non cat assuming its now the 3.9. The Blue and Green maps idle CO is set by the screw on the side of the AFM, and you have a big range, you can run so weak the engine runs really lumpy, or so rich its like its on choke. There is no lambda feedback as the probes are ignored if you have them, and RoverGauge should show the MAF CO trim voltage- should be between 1.2 and 1.8 volts or so. (see below)

2) Stepper- should be about 30% on a warm and stable engine at tickover. If you have an air leak it may be shut (0%). To find an air leak , try and get a steady idle, and then unplug the stepper with the engine warm and running. This locks the stepper position so the air cant change. Now spray easy start around any suspects points and if the idle changes, then easy start is being drawn in through a gap.

3) Lamda control- Normally white map. Introducing lambda control makes life difficult as the ECU is constantly trying to correct the mixture even if there is a fault, and it can mask a lot. If you have Lambda probes- put a DC test meter on the lambda output directly, (Black and white wires) and set to the green map (or blue if thats really what you need). You should get between 0 volts (lean mixture) and up to 1.4 volts for a highly rich mixture. The game here is to see if you can get it to switch with the CO trim- it happens at 14.7:1 AFR and you should be able to to swing the mixture through this point with the CO trim screw, The RV8 does not appreciate running this lean at idle, Id expect a smoother engine at 1 to 1.2 volts.

Ive put lots here: http://www.g33.co.uk/pages/technical-fuel-injectio...





Edited by blitzracing on Wednesday 5th February 09:35

Unlucky48

Original Poster:

71 posts

51 months

Wednesday 5th February 2020
quotequote all
Hi Mark,
Thank you ever so much for that info. I’ll digest it later. On the 4.2 lr ecu it runs blue resistor for cats or yellow for Saudi/ not cat. As you say the 3.9 runs green or white I believe. I don’t have a yellow but have ordered the correct resistors so I can knock one up just to try it without the lambda control.
I’ll try the easy start idea. As much as I don’t want it to be an air leak it just seems odd that it seems so nearly closed yet Is running over rich..
I have a vacuum gauge on its way in the post too.
I have your g33 site to blame, in a good way for the current engine spec as it was reading it years ago that inspired me to build the current head spec rather than just slapping on a refreshed stock pair..

blitzracing

6,387 posts

220 months

Wednesday 5th February 2020
quotequote all
Confused- an air leak makes it run lean at idle, not rich. You will see this in RoverGauge , the long term trim will be adding fuel at anything up to 100% on a catalyst map for an air leak. The long term fuel trim is only active at idle on a hot engine, and takes about 2.5 mins to set itself, so if you disconnect the ECU it will loose its settings. It really does tell you a lot- whats yours sitting at adding or removing? If the engine is running rich, the long term trim will be trying to remove fuel, again up to 100%. If it is at 100% you have run out of adjustment range, and something is seriously wrong.

At idle- most likely causes

For rich-

High AFM output
Faulty temp engine temp sensor
Wrong fuel pressure ( missing vac on pressure regulator)
Wrong map
Faulty lambda probe(s) or missing voltage to probes on cat maps
Misfire (seen as a lean mixture by the Lambda probes so the ECU adds fuel)
Faulty ECU

For lean at idle

Air leak
Wrong CO trim (on non cat maps)

Check for fault codes in Rovergauge- if the ECU cant get the probes to switch, it will throw an error code. This takes some time (as in several minutes ). If the short term trim is cycling, the ECU is controlling the mixture as it should. A point of note is the left and right side of the engine fuel differently due to the plenum feeding from one side, but the ECU can cope with this with different trim for left and right. Problem is if you alter the cam and valves, you alter the low speed gas flow significantly (make it worse) so the AFM output will be lower.




Edited by blitzracing on Wednesday 5th February 18:37

Unlucky48

Original Poster:

71 posts

51 months

Thursday 6th February 2020
quotequote all
I agree, confused myself hence hoping some good ph people may have some ideas.its only an engine and a few sensors but it has me beat at the moment.
The car drives ok and on its first run out for 250 miles returned 19 mpg so i doubt its over fuelling above idle but is idling very rich, 1.8% at idle and just over 1% at fast lambda test idle on an mot test machine.
The idle sounds and feels just a little off and unsteady, a partial misfire that I can’t find.

Roverguage shows temps working fine, maf is at around 4% linear and throttle pot at 4% absolute. Stepper at around 10% at cold idle and drops off to a few % or 0 once hot. That was the reason for me thinking air leak? Also the weird clack noise when I blew down the inlet and also only from the maf/ inlet pipe work while running it’s making a strange noisy metallic sucking noise that I don’t ever remember it making before that sounds almost mechanical (or possibly a leak?)
On checking the base idle with a meter I couldn’t get a steady 500 rpm, the meter kept swinging around 500/630, I guess because of the miss?
All sensors checked with a meter and read and scale ok.

Lambdas checked with a meter both are cycling voltage wise and show to cycle on roverguage. On hot idle they show an average of -25% short term trim but showed +100% long term? The only change that affected that was swapping the maf for a good one which showed the same short term trim but altered the long term trim to around+75% ?

spitfire4v8

3,991 posts

181 months

Thursday 6th February 2020
quotequote all
I'm struggling to think of a situation where you get lambda cycling but a rich mixture, other than the average co is weighted richer by a misfire causing the ecu to richen the mixture, but the occasional misfire is seen by the ecu as a weak mixture which then causes the lambdas to show the signs of correctly cycling but in fact are predominantly rich with occasional cycle to weak when the engine misfires.

1.8co is about 13.8afr so you cannot have 13.8afr AND lambda cycling unless there is something making the lambda show occasionally weak like a misfire. They certainly aren't cycling to zero volts because of correctly controlling the mixture around 14.7afr because the mixture isn't anywhere near there according to the co tester.

At this stage I would put your rovergauge away and go back to basic diagnosis.

Steve_D

13,746 posts

258 months

Thursday 6th February 2020
quotequote all
spitfire4v8 said:
..................At this stage I would put your rovergauge away and go back to basic diagnosis.
Such as getting a temperature gun on the exhaust manifold and looking for that misfire.

Steve

blitzracing

6,387 posts

220 months

Thursday 6th February 2020
quotequote all
Also- base idle... simply wind it down as low as the engine will reliably tick over, it should be stable, unless you have a misfire or mechanical issue. Then when you bring the stepper back into play it should add enough air to add a couple of hundred RPM, and 30% open would be a good figure. If its hitting zero on the stepper the base idle is too far open or an air leak. A base idle that is wound up to far makes a horrible hissing from the top of the plenum.

Unlucky48

Original Poster:

71 posts

51 months

Friday 7th February 2020
quotequote all
Thank you guys for your continued help
I’m stuck on a project at work for a few days but will go back through everything starting with the basics and see if I can find the issue.
I can’t tell if the slight misfire is engine related or just a result of it idling so rich, it doesn’t sound cyclic so I’d assumed it was the rich running. It’s chucking out so much co that it fills up the drain in my co meter and makes eyes water of anyone who dares get too close.. It starts on the button and there’s no sign of ignition hesitation above idle so I’m sure it’s not spark related.
I’m new to using roverguage having only bought the cable from mark last week but I just thought as an additional diagnostic tool it may help and I guess it has as I wouldn’t have known what position the stepper motor was working at without it.
Having thought about it I think I answered my own daft question about the strange noise though. While attempting the daft diy smoke test I didn’t block off the flame trap pipe so I’m fairly sure the noise was just me pressuring the valley gasket..
will update as soon as I have any news.

spitfire4v8

3,991 posts

181 months

Friday 7th February 2020
quotequote all
Your co level won't make your eyes sting .. it's the unburnt fuel that will do that ie high HC reading .. caused by misfires / poor combustion. Indeed your engine should run very smoothly at between 13 and 14 afr - the RV8 likes it richer rather than weaker.
Go back to basics .. ignition system, fuel system, compressions etc

blitzracing

6,387 posts

220 months

Friday 7th February 2020
quotequote all
Id drop all 8 plugs out and check the colours as a starting point, but have a closer look at your gas results. Lambda is a measure of excess oxygen in the exhaust, which is why it gets confused when you have a misfire. Have a read here to really see whats going on if you have full gas results:

https://www.crypton.co.za/Tto%20know/Emissions/exh...

Edited by blitzracing on Friday 7th February 18:26

Unlucky48

Original Poster:

71 posts

51 months

Monday 17th February 2020
quotequote all
Hi guys,
I found some time to start digging around and other than miserable bottom end news I haven’t discovered the cause of the misfire problem yet.
The plugs were all too lean with exception of no. 3&4 which were perfect colour..

A compression test and leak down test revealed the heads are good but too much leakage past rings in all cylinders, around 20%.. i took a gamble on new rings and a hone as oversize pistons aren’t available for the 4.2 and looks like I now need new stock cylinders.
Compression test showed 6 cyls at 150-160 psi and one high at 170, one low at 145. Not good

A vacuum test connected to a stub on the plenum shows 12 ins hg at idle and rising to 20 at 2500rpm but I’ve no idea what it should be?.
I rigged up the correct ohm tune resistor for non cat map to eliminate the lambdas and it ran much the same but with even more co & hc

I still need to test the inlet area for air leaks but don’t like the idea of spraying easy start everywhere so I’m going to pull the ram housing and make up an air fitting that I can ram down each trumpet to check for a leak that way.

Other than pulling it apart again for a bottom end any ideas?.

blitzracing

6,387 posts

220 months

Tuesday 18th February 2020
quotequote all
Can you actually alter the mixture readings on a non catalyst map with the CO trim screw? It has a huge range available.

Unlucky48

Original Poster:

71 posts

51 months

Tuesday 18th February 2020
quotequote all
I could affect a change in co, taking it down to .5v still had the co at 1.6% and eye watering hydrocarbons, turning it up to 1v or more sent co climbing.
Another strange thing I couldn’t fathom, while fiddling with the maf (also tried good substitute) engine idling I took the filter housing off so I could lift it up a little to get to the plug leads and every time I lifted it more than 10 degrees up or more it had an incremental negative effect on the idle and made it stutter. Wiring all seems fine and giving the maf wiring connector a good wiggle didn’t cause it.

I’m going to do the double check for inlet air leaks, put it back to non cat map and do a plug chop to see if it changes how the plugs look and swap dizzy just for the hell of although I’m sure it’s not an ignition problem as i’m pretty much out of ideas now.
When doing the previous checks I checked the exhaust header temps while warming up and the two cylinders with happier looking plugs were 80-100 degrees c cooler than the other six which were up near 400 degrees c

blitzracing

6,387 posts

220 months

Wednesday 19th February 2020
quotequote all
If moving the AFM does that, id plug in RoverGauge and set the AFM reading to absolute and see if the reading alters as you move it.

Unlucky48

Original Poster:

71 posts

51 months

Wednesday 19th February 2020
quotequote all
Good plan, thanks will try that.