Tuned 3.9EFI 14CUX running rich

[jepper]

Hello!

This is my first post here, but I've been reading this forum for a long time.

I have a problem with my engine. It's running a bit rich and has a little "searching" on idle. Also a few pop's now and then. The engine is a 3.9 EFI 14CUX from a Range Rover 1990 (Cat spec.) and is now placed in my Range Rover from 1973. I have just had it running in for 20minutes at 2000rpm and then changed the oil and filter. Now trying to solve the rich running. Exhaust fumes both smell and looks very rich.

I'm still running the cat ECU map (although cat less system) because I want to take advantage of the controlled fueling the oxygen sensors provide to make it as clean and economical as it can be. (yeah, I know... )

My thoughts is if the rater high increase in CR (from 8.13:1 -> 10:05:1) and/or the cam timing can confuse the ECU (or other parts) to make it run a bit rich? Or maybe the low backpressure in the exhaust system?

I'm planning to try to adjust base idle and maybe the AFM setting. But I think setting AFM doesn't count when running cat ECU map? Also have tested both on the lower and higher side of the fuel pressure to no difference. Disconnecting the AFM makes it run even worse.

I have bought new parts to test with:
* 3AM to a 5AM AFM.
* Idle air control valve.
* EFI Fuel pump.
* Fuel filter.
* Lambda Oxygen sensors.
* New injectors, 4-hole ones.
* Throttle position sensor (used)
* Champion spark plugs.
* Distributor cap, rotor and ignition leads, all new Lucas.
* ECU Coolant temp sensor.

Engine spec.
Originally a 36D 8.13:1 cR.
* High comp pistons with grinded valve pockets
* Heads grinded 0,5mm. New CR 10.05:1 (estimated) with tin gaskets.
* High torque cam kit with adjustable cam timing from RPI. Set at +4degrees for more low end torque.
* A little ported and port matched exhaust & inlet. Just to make ports look more tidy.
* New push rods, steel rockers and bridges. Preload set with shims.
* From original exhaust manifolds: 1,75" on each side down to one 3" all the way and two mufflers (straight through ones). All home made, stainless.
* Lambdas in each down pipe, at about the orginal place.
* Adjustable fuel pressure regulator.

Pictures of my Range Rover here:
http://picasaweb.google.com/rangeroverjep

[RangingRover]

From memory standard Rover efi systems seem not to run much different with no exhaust past the manifolds, probably due to running off the AFM. I don't think lack of backpressure is the problem in this case. On my alfa V6, removing the exhaust past the 2 downpipes makes it run extremely lean, and backfire out the intake, but that is a manifold pressure based aftermarket ECU.

I would suggest the AFM needs to be adjusted, as that is one of the main things that tells these cars how much fuel the engine needs, along with engine rpm. Once the ECU has the air flow figure and the rpm figure, it basically looks up a table in its memory, and uses the fueling number from that point. The oxy sensors then read what is coming out the exhaust, and makes adjustments from the table figure, HOWEVER their adjustment range is usually very limited (5% maybe?). If the AFM is way out, the oxy sensor will never be able to bring it back within its specified range.

Other thought - have you put larger injectors in than the ECU was programmed for?

[jepper]

The injectors was the standard Lucas black/green 1-hole ones.
I've changed them to Bosch 4-hole types but it made no difference. (Said to be a few % more efficient)
Atleast i know they aren't flooding the engine by leaking.

I'm going to se what the reading is at the AFM and then try to adjust it as you said. I'll post updates after the weekend in the workshop

Thank you for your reply.

[RangingRover]

The injectors was the standard Lucas black/green 1-hole ones.
I've changed them to Bosch 4-hole types but it made no difference. (Said to be a few % more efficient)
Atleast i know they aren't flooding the engine by leaking.

I'm going to se what the reading is at the AFM and then try to adjust it as you said. I'll post updates after the weekend in the workshop

Thank you for your reply.

The bosch 4 hole type (assuming you mean the short green ones as fitted to series 2 discovery) require a higher fuel pressure (3 bar I think) and have a slightly different flow rate to the green lucas items..... Personally I'd put the Lucas ones back in, with your fuel reg set at 2.5 bar until you get the rich running sorted out, as this way you get back to what the ECU was set up with originally.

[jepper]

Ok thanks!. That's good info!

I have been testing a few things today.
I started it again from cold and it ran very smooth and no black smoke. But after engine got warmer it started to smoke and smell more. I measured a CO rate at 8%. (CO limit for EFI's in Sweden is 0,3%)
Adjusted the AFM from orginially 309,7ohm to 320ohm. Made no difference, nor did the opposit direction. Should I adjust more?

Also lowered fuel pressure to only 1 bar. It still did idle with an CO level of 8%. This should result in a lot less fuel. Looks like the ECU in some way wants the CO level at 8%

I have measured coolant sender again. Around 3000ohm at 17°C and 320ohm with hot engine. That's the same as the old one.
I'm going to check that the coolant sender wires really reaches ECU connector.

Maybe my ECU is broken?
It looks like it is not able to 100% control the fueling, or I just have to get close to the narrow band where Lambda control can take over.

[jepper]

Coolant & Fuel temp sensor wireing to ECU connector is ok.
Tune resistor is ok. Resistor is 3900ohm = Europe cat tune

I've also read that adjusting AFM has no effect when running cat ecu map?

I'm going for another ECU to test with. It's the only thing left to change besides ignition amplifier module. '
If a new ECU doesn't help, I'm going to run it without lambda control and manually adjust it to correct CO level.

I'll update any progress.

[jepper]

I did a big breakthrough tonight when I discovered that 12V current reached the oxygen sensors only when fuel pump DIDN'T run!
With ignition on and fuel pump was running the sensors got 0.0V
With ignition on and fuel pump had pressurized and turned off there was 12V at the sensors. It should be the opposite.

So, when engine is running (and fuel pump), no current reached the oxygen sensors = they didn't work.

How's that?
Prior running in the engine I had a bad fuel pump relay which I changed for another in the loom (Lucas). It looked the same but wasn't.
I have now fixed the standard relay (BOSCH) and now there is 12V at sensors WITH fuel pump running!

To bad I managed to buy a used ECU on ebay already but after all, spare parts is good to have in the future.

I'll update tomorrow when CO level is measured again.

[nottie]

Ok thanks!.

Adjusted the AFM from orginially 309,7ohm to 320ohm. .

Ok on The RPI web site it says from factory the resistance is set at 174ohms!
So yours is way over.
In summer i have and also others that leaning it out to about 90ohms seems to make them run better and also stops it running rich.

If you havent seen this site then have a read as there is some very good info on there.
http://www.v8engines.com/carbs-2.htm#top
Keep us updated on your engine as you have posted some great info.

[GRIMACE]

As nottie has mentioned briing the afm back down to 174ohms.

Nice work on the engine too, it would go pretty hard once everything is sorted.

[jepper]

Thank you both for the info about the AFM resistance! (and link).
The old 3AM AFM was set at 298ohm but it can surely have been adjusted before I got hold of it since the engine was in and old military Volvo. When the new one (5AM) arrived it was preset at 309,7, therefor I thought they should be set at around 300.

Now when oxygen sensors get 12V feed they give signals back to ECU as expected. I now have a voltage reading up to 1.1V over terminal 4 (black) and 23&24 (Shielded blue wires). The Oxygen sensors doesn't give a constant reading, but cross counts which ECU expects to see 8-10 of per second at 2000rpm.

I hooked up a oscilloscope to the sensors (one a time) and I could clearly see the signals reaching ECU. I filmed it and it can be seen here:

Video of Lambda signals at idle:
http://picasaweb.google.com/rangeroverj ... 6116228866

Video of Lambda signals at 2000rpm:
http://picasaweb.google.com/rangeroverj ... 6747871186

Although, now the oxygen sensors does their job, they both report a bit rich mixture constantly peaking at around 1V.
I've also measured the exhaust fumes again and they are now down to 1-2% CO. Still to much for a EFI system but low enough for the MOT of my old 73 RR.

It does still smoke, smell and search idle a bit. Fumes are a little blue. Since the engine is so "fresh" a part of the fumes can be oil since the piston rings hasn't worn in yet. (Has only run 1 hour)

Here is a video of the fumes (and sound):
http://picasaweb.google.com/rangeroverj ... 8565073474
in fact I couldn't hardly see any fumes on the video but there is for real

I'm going to look at the AFM on wednesday when I have taken a day free for some welding in the rear wheel housings.
Well, a good day today, some progress finally
I'll update any progress later this week.

[nottie]

Good to hear Jepper.
Grimace has dealt with RPI on a few occations and they are apparently really great to deal with and there info is awsome.
Cheers Nottie

[chimpboy]

Does anyone know what the difference is with the fuel maps for the different tune resistors?

I understand it goes:

Red wire, 180 Ohms, Australia, Rest of world.

Green wire, 470 Ohms, UK and Europe - Non Catalyst

Yellow wire, 910 Ohms, Saudi - Non Catalyst

White wire, 3900 Ohms, USA and Europe - Catalyst

... and I believe the 180 Ohm is for a catless version..?

But if 180, 470, and 910 are all without cats how different are the fuel maps?

[Philip A]

Just a few things to check.
1 Check that you have the correct bank O2 sensor connected to the correct signal wire to the ECU. Its in the manual. After I fitted O2 sensors I found they rapidly cycle between 0.5 and 1 volt giving an AVERAGE of 14.7:1. I do not have an oscilliscope but used an LED DIY o2 output measure and my O2 output change was much faster than your oscilliscope at 2000RPM, but maybe that was an oscilliscope limitation. i also tried an analogue multimeter and it showed the same swings in output.

2 Have you checked that all the plug wires are correct, ie the timing order is correct? It sounds a bit like switched plug wires but who knows from the sound .

3 Have you checked the resistance of the ECU temp gauge.

4 Is the thermostat an 88C one. if under 80C the ECU will apply cold start correction.

5 The book says to have the MAF at 1.5 volts , and I think that this is so that the o2 sensors can take out fuel, but this is not critical.

6 there is a company in UK that makes a 14CUX code reader for a 100Euros or so. maybe you should get one and install the wires in the ECU plug.

The o2 sensors have a wide range of correction, so it is surprising that they cannot correct. perhaps install a non cat resistor and disconnect the o2 sensors and see what the CO is.
Regards Philip A

[jepper]

Just a few things to check.
1 Check that you have the correct bank O2 sensor connected to the correct signal wire to the ECU. Its in the manual. After I fitted O2 sensors I found they rapidly cycle between 0.5 and 1 volt giving an AVERAGE of 14.7:1. I do not have an oscilliscope but used an LED DIY o2 output measure and my O2 output change was much faster than your oscilliscope at 2000RPM, but maybe that was an oscilliscope limitation. i also tried an analogue multimeter and it showed the same swings in output.

I've read that the o2 output should be between 8-10 / second. But the same appeared with the new o2 sensors. As you say, I guess it could be a limitation with my oscilliscope. I've also had the sensors disconnected and measured between that and the ECU connector to get it to correct side.

2 Have you checked that all the plug wires are correct, ie the timing order is correct? It sounds a bit like switched plug wires but who knows from the sound .

I have also double checked this, they are correct. The sound in the video doesn't make the IRL sound justice! But it was running abit lumpy.

3 Have you checked the resistance of the ECU temp gauge.

Yes, it appear to have right values. around 300-400ohms with hot engine. (Same for fuel temp, but it's only used when doing a start I think, ECU knows if a warm or cold engine is started).

4 Is the thermostat an 88C one. if under 80C the ECU will apply cold start correction.

Good point! The thermostat is an 82°C. Maybe that is a little on the cold side? (I've ordered a new 88° just for sure).

5 The book says to have the MAF at 1.5 volts , and I think that this is so that the o2 sensors can take out fuel, but this is not critical.

Have tried to adjust the hotwire MAF down to 178ohm's (309std). That resulted in a very unresponsive throttle. I thought adjustments to the MAF had no effect when in closed loop, but apperantly it has

6 there is a company in UK that makes a 14CUX code reader for a 100Euros or so. maybe you should get one and install the wires in the ECU plug.

I already have the plug and also have been looking at these:
http://www.bespokeintelligentsystems.co.uk/Products.htm
It had surley been cheaper to buy one from the start...

The o2 sensors have a wide range of correction, so it is surprising that they cannot correct. perhaps install a non cat resistor and disconnect the o2 sensors and see what the CO is.

Last weekend I did a little switch and a few resistors that makes it easy to switch between "Non cat, Europe" - 470ohm and "Cat, Europe" - 3900ohm.
When I tried it in the "Non cat" setting, the engine ran very smooth with clear exhaust fumes. It was really as switching between day and night! I'm really looking forward to take a test drive soon :-)

I still don't know why the ECU struggles in the closed loop mode. I've tried another 14CUX ECU from 1991 and it does the same. Will follow up any progress!

Last, very big thanks for your thoughts and help!
Regards
John-Erik

[Philip A]

I still think you may have the o2 sensor wires mixed. Left is terminal 23 on the ECU plug and right is 24.

Did you add an earth to terminal 4 on the ECU plug? or maybe there is already one in your loom.

When I added o2 sensors to my non cat australian spec car there was no 23/24/or 4 wires in the ECU plug.

You are also supposed to use shielded cables from the 02 sensors, but if your loom already has them this is not an issue.

Did you use NTK Titania o2 sensors? The more common Zirconia sensors will not work.

Regards Philip A

[Philip A]

BTW , the plug that was in my loom was not the plug required for the US spec diagnostic tool but another plug for a dealer diagnostic tool.
The correct plug has 4 wires.

Regards Philip A

[chimpboy]

I still think you may have the o2 sensor wires mixed. Left is terminal 23 on the ECU plug and right is 24.

Did you add an earth to terminal 4 on the ECU plug? or maybe there is already one in your loom.

When I added o2 sensors to my non cat australian spec car there was no 23/24/or 4 wires in the ECU plug.

You are also supposed to use shielded cables from the 02 sensors, but if your loom already has them this is not an issue.

Did you use NTK Titania o2 sensors? The more common Zirconia sensors will not work.

Regards Philip A

Philip do you have a write-up somewhere on adding the O2 sensors?

[Philip A]

I have an acrobat document with photos on the installation.
It was posted in Projects on AULRO a few years ago, but PM me your email address and I will send it to you.
Regards Philip A

[Philip A]

Jepper,

Any result with your 14CUX???
Regards Philip A

[jepper]

After a few weeks, doing cold and warm starts and a few houres of running time on the engine I must say the exhaust gases looks ok now but still smells a little rich. CO mixture between 1% and 2% in closed loop.

Two other problems I've noticed, which probably has existed from the first start, is:

* With cold engine the idle is steady around 750rpm for around 10 seconds. Idle then drops to 4-500rpm for just a second and then back to steady idle again another 10 seconds. This repeats until engine is warm. There is no change if I disconnect the idle stepper valve so this has to be something the ECU does to the injectors, that is the only thing left the ECU can control. It's a little hard to drive when cold as you loose a little amount of power suddenly. It doesn't feel like a missfire, only as letting of the throttle a bit.

* When engine is warm, idle is way to high, around 1500rpm. After around 30seconds standing still, it sometimes lowers engine revs to slightly high idle, 800-900rpm. often it just hangs on at 1500rpm. The Idle stepper valve is brand new. If I clamp the hose to the valve it lowers to OK revs, base idle = the valve doesn't adjust because the ECU don't want to for some strange reason. I've read that the ECU will raise idle if missfire is detected. But I don't think it missfires. Sounds really good without any vibrations. On engine turn off, a "buzz" can be heard from the vavle fully open. The old valve did the same. If I disconnect the road speed sensor (also brand new), idle is even higher.

These two problems still exist after changing these parts:

* Used ECU from a 1991 14CUX Range Rover.
* New lambda sensors (NTK Titania)
* New Idle Stepper valve
* New Air Mass Meter (5AM) and hose
* New Vehicle Road Speed Sensor
* Reconditioned and bench flowed injectors.
* Used Throttle position sensor (measuerad ok, smooth over the whole range)
* New Genuine coil, sparkplugs, rotor, cap and HT leads.
* New fuel pressure regulator (adjustable with manometer)
* New fuel tank, pump, filter and all lines.
* New vacuum hoses and T-coupling.
* New colant temperature sensor
* Custom built stainless exhaust of "free flow" type all the way with lambdas in the original places.

Other things I've tried without success:

* (as in earlier post) Made a switch that I can select between "Europe non cat" 470ohm or "Europe cat" 3900 ohm with.
* Setting ignition timing from late to early.
* Electrically isolated Air Flow Meter from ground.
* Adjusted fuel pressure from very low to a little bit over 36psi.
* Reset ECU (unlpug battery with a main switch)
* 510ohm resistor to ground to set Manual gearbox (LT95) for the ECU.
* Searched for vacuum leaks around plenum with "start spray".
* Measured fuel temperature sender ok.
* Adjusted Air Mass Meter

Nothing has really made any difference so I'm starting to think it can be the difference spec on the engine, but why?
Cam shaft timing, compression ratio. Should that effect the ECU in this way? I can't understand how?
In my earlier post I did say there was a difference to the better when selecting "Europe Non Cat", but I can't relly notice that longer. Maybe it was imagination.
Still to check is earth/ground wire from terminal 4 on ECU and to adjust Air Mass Sensor when running "Europe Non Cat".
I have now made an simple Idle Stepper Valve substitute which I can adjust with my old choke cable to set idle manually while doing some more research.
I'm also thinking about going for megasquirt insted of the 14CUX, but not yet. There has to be a logical explenation to this, right?

It drives really well on the road. Engine has lots of power and a great sound. Very responsive to the throttle. Lots of pictures here -> http://jep.mine.nu

Thank you for all help!

[GRIMACE]

Another silly thing to check is the hose connecting the AFM to the Plenum, it may have a small crack in it.

EDIT: Sorry just noticed you have already changed the hose.

Also check & clean all earth straps.

Good luck, I would be on the edge of total melt down if I had completed all the above as you have with no sucess.

[Philip A]

One thing you do not mention is the "T" piece in the vent hose from the oil seperator to the throttle plate. this is not a normal T piece but has a very small metered hole in the stub of the pipe leading to the plenum.

Sometimes people break the OE T piece and just replace it with a normal T piece.
This would give an air leak big enough to cause idle at 1500RPM.

Have you set the base idle ? Ie disconnected both sides of the IAC and blocked them , and then seen what the idle is? The base idle is set by turning the screw on top of the throttle plate out to speed up and in to slow. it should idle at 550-600 with the IAC blocked off. I knoew you said it comes down in revs but have you set it .

Have you got the wire telling the ECU the engine revs connected. It is in the MAF sub loom and is connected to coil to amplifier side. ( I don't know whether it would run at all without it connected)

Is the MAF fitted the correct way,as I guess you may not have factory air cleaner.

I know I am nagging about this but if the O2 sensors are fitted to the wrong bank, then one bank will be full rich and one full lean, which could conceivably cause your problem.

Your problems are bizarre, and it is hard to see how they can occur.

If the ECU knows revs then it always tries to stabilise at 750idle in Neutral so maybe it doesn't know the revs.

You say that if you disconnect the VSS the idle is higher. This is the opposite of what should happen. the idle should slow whith no VSS signal.

What voltage on your TPS at idle . it should be under 0.49 volts and ideally 0.33 volts. If the TPS is over 0.49 volts it may not recognise Idle.

On cold start , depending on the resistance of the sensor, it usually goes to say 1100RPM for the first 10 seconds, and then to say 900 for a couple of minutes then down to 750. Mine also hunts while cold a bit since I had a unichip fitted but certainly not while driving, only on idle.

Funny, since I increased my TPS voltage to 0.33 from about 0.2 mine is hanging sometimes at 1500 for a short time,so it may be the TPS.
After all that I wonder if the ECU is damaged . Have you tried another ECU?
Regards Philip A

[jepper]

Thank you for your reply again.
I can't understand where the problem is either. Seems really weird!

The T piece is also brand new, and there is only a tiny hole there. I think the size of the hoses makes it impossible to put the wrong way also.

TPS voltage was a little bit low, around 0,2 V I did turn it very little and got it spot on 0,33V.

Earlier I did set the base idle at 520rpm, didn't effect the problem. I have tried to adjust both lower and higher of base idle but going to look at it again so it's not way out now.

I think the ECU doesn't start the fuel pump (after initial 2-3 seconds pressurizing) if it doesn't get any signal back through a 6.8kohm resistor from coil. So I asume it's correct. Now when I'm thinking of it, I'm not 100% sure it's on the right side of the coil. One more thing to check!

MAF is correct I think, the connector for the wires is pointing out from the engine. I use the round type of aircleaner (as the 3.5 EFI has). Not the more square-like box.

More ToDo:s
* Measure the four wires to the Idle Stepper Valve.
* Measure Oxygen sensor wireing so they arn't on wrong side.
* Double check ECU terminal 4 to earth.
* Set base idle
* Measure with a oscilloscope on the signals to Idle Stepper Vavle to se if ECU tells the Valve to lower engine revs or not.
* Add another earth strap

Regards
John-Erik

[chimpboy]

I think the ECU doesn't start the fuel pump (after initial 2-3 seconds pressurizing) if it doesn't get any signal back through a 6.8kohm resistor from coil.

I can confirm this; I had a dead resistor and it acted as you describe.

[Philip A]

Jepper,
I have the manual which sets out a complete step by step testing procedure for 14CUX.
If you give me an email address I can send you the injection part Pdf and maybe you should do it step by step.
Regards Philip A

[jepper]

That would be nice of you.
e-mail: remskiva A gmail.com
Replace A with @

Going to look at it again this weekend.

Regards
John-Erik

[jepper]

Thank you Philip A and others that have helped me here.
Finally I think it goes rather OK.

One thing I found that did a slightly difference to the hunting at idle was adjusting MAF down to 170ohm's AFTER switching to no-cat spec. (As you said earlier, but it made the throttle unresponsive when in cat spec.) I think I only tried adjusting it when on cat spec earlier. Sorry for that.

Everything I measured that was in the manual looked good. All leads has god continuity. Despite that, I never got the idle ok when warm and stepper motor connected. So for now I use the simple valve that I can adjust by my self. This is a strange fault/ behavior that must point to the difference spec of the engine. Could be bigger head ports, higher compression ratio, camshaft profile/timing or maybe the rather big free flowing exhaust

I'm going to finish the last few bits before the MOT test now. When it has passed MOT, it's easier to fine tune everything since I don't need to hide from the police

Again, thank you all!
Will update later on!

[Philip A]

One thing I found that did a slightly difference to the hunting at idle was adjusting MAF down to 170ohm's AFTER switching to no-cat spec. (As you said earlier, but it made the throttle unresponsive when in cat spec.) I think I only tried adjusting it when on cat spec earlier. Sorry for that. Embarassed

Jepper, I did not look back at the entire thread but I did not mention the MAF and 170Ohms.

The correct way to adjust the MAF is to set the voltage with ignition on to around 1volt for non cat or 1.8 volts for cat tune. The terminals to back probe are the ones on each end.

Mine ran well at 1 volt, but I presume that it is easier for the o2 sensors to pull out fuel rather than add it. I have always had a bit of a hot stall into D problem so I recently adjusted it and it is a little better.

I have seen this Ohm method mentioned on forums but the manual says volts.

Regards Philip A

[jepper]

I think 1Volt equals 178ohms but not sure. Going to measure again tomorrow and also doing CO gas analysis to see where we land :-)

Thank you!

[jepper]

I just found that the vacuum timing advance membrane was faulty. It appears that there is a big hole in it as I can suck air through it. Hopefully then I can go back to lambda control and reconnect the idle stepper valve!

So, now I'm going to buy a new distributor (to get a spare amplifier module also)
If I look at lrseries there is a high compression and low compression type. I guess I should get the high compression one. Is the springs in the distributor different to the low compression one maybe?