Pinking on WOT

[Danblez]

Hi all,

I was out having some fun at the weekend with the roof down and noticed when on WOT from about 4k revs it was pinking quite heavily. It only happens on WOT on haevy load and high revs. Pulling from lower down the rev range and all is fine. I haven't noticed this before.

The only thing I have changed recently are the plugs and went for some BPR6ES which the car seems to be happier with. I also haven't changed the fuel filter but have this planned.

When I got the car one of the lambdas had been accidentally unplugged and I recently had the other one fail. I wonder if someone has set the mixture with the lambda unplugged?

I have the ecumate lite and there are no errors stored.

It is a standard 4l and I always run premium fuel. I am a little worried about this as we have a trip to Italy planned in a month which will only make matters worse I suppose so really need to get on top of it. Where should I start trying to diagnose?

Thanks
Dan

[nigelj77]

I used to have a 500 which pinked terribly under the same conditions as yours, it was eventually cured by turning the timing down, not sure if that's a possible on yours with an aftermarket ecu.

If it's something that has just started could you have an air leak leaning out the mixture?

What did the plugs that came out look like in terms of colour?

I hope this gives you a start on things to look at.

[macdeb]

DON'T induce the pinking, it will cause damage. Get the map looked at and maybe retarded a bit if you have aftermarket ECU or wind back the distributor slightly if you still have one. ASAP
Just done an Italy trip and it was epic, you'll love it. Be prepared for the attention the car will get.

[Danblez]

Thanks both it is running standard ECU. I hope I havent done any damage. It may have been doing it since I bought it and not noticed and it has been given the beans fairly frequently.

I haven't looked at anything yet as I only noticed last weekend.

Anyways it is going in to XWorks (great garage) for some bits I wont have time to do before I go so have asked Heath to have a look. although I would like to understand the potential issues for myself. Hopefully it is as simple as timing.

I am a little apprehensive about the trip but cant bloody wait! Bring on the tunnels!

[Danblez]

My the car is still pinking, only under heavy load and high revs. I cant get it to pink at all under heavy load and low revs and is otherwise running perfectly. I have knocked the timing back to 28 degrees and it hasn't made any difference at all. When I fist got the car one of the lambdas was unplugged and it didn't pink (or wasn't noticeable at all), this is leading me to believe it may be fueling related.

I have been reading about a few possible causes but could do with some real world experience of where to start my investigations. Thanks!

It is a completely standard 400 apart from an ACT Y piece.

[SMB]

What fuel are you guys using, mines been fine until after i filled up with vpower at a shell station down in Littlehampton. Ever since then I have exactly the same pinking as you have described. Nothing else has changed so I suspect the fuel. Fill up was on the 1st august

[Danblez]

I generally user 98 RON but occasionally 95. This doesnt seem to make any difference though.

[Smokey Boyer]

From what I have read, running a car with the lambda sensor physically in the manifold, but not plugged in to its electrical socket can damage the sensor. I am not sure if that would cause the problem you have, but might be of use.

[blitzracing]

This is very likely down to an over lean mixture under load as its almost impossible to make the RV8 pink with wrong ignition timing if the mixture is correct. The effect of a lean mixture is to increase the temp in the combustion chamber, so pre detonation is much more likely once the flame front is ignited so you would be looking at a drop in fuel pressure / clogged fuel filter being the most likely cause. It wont have anything to do with Lambda sensors, as these are not used above 3400 rpm anyway.

You can check for a lean mixture with a test meter across the lambda outputs- the black (positive) and white (ground) wires if you run a long cable into the cockpit you can read with a test meter under load. Below 3400 rpm it should cycle between 0 and 1.2 volts approx, but then on WOT it should be at least 1.2 volts all the time- up to 1.4 or so, depending on the probe age. A lean mixture will show as the voltage drops away to zero as the load increases showing the AFR is running leaner than 14.7:1. This is bad news and needs to be rectified.

Edited by blitzracing on Tuesday 11th August 12:25

Edited by blitzracing on Tuesday 11th August 12:26

[Pupp]

I'd be looking at going back to a 7 on the plug and/or moving from the projected nose type if it were me...

[QBee]

Mind did this after an engine change, to a modified 5 litre engine. AFR went from about 13:1 to over 20:1 at 4600 rpm, as if someone had thrown a switch. Changed the injectors for larger ones, upped the fuel pressure, all to no avail.

In the end I took the car to Mark Adams, who remapped the 14CUX chip. Been perfect ever since.
Mark is based south of Shrewsbury. He uses a rolling road in Shrewsbury amongst others.
Jools is based in Chesterfield and can also do the required re-map. He could be cheaper, as he has his own rolling road.

One thing you could try first is a different 4litre ECU. I have loaned my spare to someone else, and will need to get it back, but Xworks should be able to lay their hands on one and try it for you, before you go to the expense of a re-map. My car is on iridium 6 plugs with no issues. And I only use higher octane unleaded on track days.

[Danblez]

blitzracing said:

This is very likely down to an over lean mixture under load as its almost impossible to make the RV8 pink with wrong ignition timing if the mixture is correct. The effect of a lean mixture is to increase the temp in the combustion chamber, so pre detonation is much more likely once the flame front is ignited so you would be looking at a drop in fuel pressure / clogged fuel filter being the most likely cause. It wont have anything to do with Lambda sensors, as these are not used above 3400 rpm anyway.

You can check for a lean mixture with a test meter across the lambda outputs- the black (positive) and white (ground) wires if you run a long cable into the cockpit you can read with a test meter under load. Below 3400 rpm it should cycle between 0 and 1.2 volts approx, but then on WOT it should be at least 1.2 volts all the time- up to 1.4 or so, depending on the probe age. A lean mixture will show as the voltage drops away to zero as the load increases showing the AFR is running leaner than 14.7:1. This is bad news and needs to be rectified.

Great thanks for the reply, this would seem like a sensible first step. Does the lambda have to be plugged in to the loom to test?

I am also going to connect up to Roverguage, would this shed any light?

Just for info I have already changed the fuel filter to eliminate that (there is only one isnt there?) and have switched back to the 7 plugs also.

I will bear in mind your comments QBee when I have done some eliminating

Edited by Danblez on Tuesday 11th August 14:46

[blitzracing]

You have to leave the lambda plugs connected, and push some test probes down the back of the connectors, or at a push I have a specific connector, loom and test meter for this- I just want to be sure I get it back if I loan it out. You need all the connections in place as the heater supply (the red wire) also supplies the resistive Lambda probe, to allow it to generate a voltage output. Unfortunately RoverGauge cant read this voltage, as the trim values it shows you is the ECU response to the voltage from the probe- but this is ignored at WOT as its open loop, although the probe is still reading and producing a signal all the time you can make use of.

[blitzracing]

Just a thought about the mapping- you can check the map fuel delivery by looking at the live fuel map data- The horizontally rows is basically load / airflow, low airflow top row, that increases towards the bottom row. The vertical rows are RPM bands. What you should see under WOT is the lowest Row highlighted at whatever RPM, hopefully with a fuel cell value of FF (its in Hex) showing peak injector time and fuel flow. If you are not reaching the bottom of the map at WOT then a sensor value is wrong or the map scaling is wrong- although this is unlikely unless you are running the wrong map.

[Danblez]

Thanks a lot Blitz, I may take you up on your generous offer but will give it a bash first. I presume a normal multimeter will read this output? I have got an oscilloscope too but wont be able to use this on the move!

[blitzracing]

The best type is an analogue meter with a 2.5 volt (best) or 10 volt DC scale. A Digital will do, but they are not so good as the voltage constantly shifts, so the DMV display is not that stable, where as an analogue one damps out the voltage shifts in its mechanical lag. For WOT the voltage wont shift that much as the mixture does not cycle so the readings should be better.

If you are running lean you will see the lambda voltage drop away as the load increases- its quite clear.

[Danblez]

I haven't managed to do the lambda thing yet but have had it hooked up to RoverGuage at the weekend.

I also checked it was running the white cat map.

A couple of odd things to note

Long term Lambda trim is sitting at +89% and +61% which I believe means the ECU is trying to increase fueling. The short term trim also hits 100% occasionally on. So this seems to tie in with the suspected running lean.

I have also done some reading of other potential causes. I believe engine load and therefore fueling is calculated from the AFM reading and engine RPM.

So I have just bought a fuel pressure tester to test system pressure and am going to test the AFM voltages also. I am also going to try and confirm the lambda readings on the move.

Am I on the right track?

Edited by Danblez on Tuesday 1st September 10:12

[gruffalo]

Danblez said:

I haven't managed to do the lambda thing yet but have had it hooked up to RoverGuage at the weekend.

I also checked it was running the white cat map.

A couple of odd things to note

Long term Lambda trim is sitting at +89% and +61% which I believe means the ECU is trying to increase fueling. The short term trim also hits 100% occasionally on. So this seems to tie in with the suspected running lean.

I have also done some reading of other potential causes. I believe engine load and therefore fueling is calculated from the AFM reading and engine RPM.

So I have just bought a fuel pressure tester to test system pressure and am going to test the AFM voltages also. I am also going to try and confirm the lambda readings on the move.

Am I on the right track?

Edited by Danblez on Tuesday 1st September 10:12

Just a quick thought have you checked the throttle POT is doing what it should, that is also used to decide fueling and they can wear.

[blitzracing]

Not far off- but i would not take too much heed to the closed loop values- as long term trim is is set at idle, and for those conditions only- so in you case of WOT and open loop the fuel requirements will be way in excess of those at idle, so a whole set of new variables comes into play. Certainly fuel pressure is worth checking- but it needs to be at WOT, which can be somewhat difficult to read a gauge strapped on somewhere at 100 plus leptons, unless you have access to a rolling road. The nice thing about the Lambda output is it perfectly possible to get a passenger to read a test meter when you floor it.

[Danblez]

gruffalo said:

Just a quick thought have you checked the throttle POT is doing what it should, that is also used to decide fueling and they can wear.

Good call, this is on my list of things to test although because there seems to be an issue with the car underfuelling at idle I put this lower down the list!