Tuning Issue That Has Got Many Of The Best Minds Stumped.

[D_Stirls]

Hello All,

One of my mates has encountered an issue with his GTR which his tuner (as well as many of the top workshops in the eastern states) have not been able to figure out.

Lewis Built RB26

260 x 9.15 tomei cams

SARD 700cc injectors

CNC ported head

Trust T517Z's (8cm)

Haltech Pro Plugin

No O2 sensors

Walbro intank pump with 044 from external from surge tank.

The car had a few electrical gremlins that were sorted out before tuning, it then went on the dyno and was tuned and it made 360kw @ 20psi (T517Z's).

The issue came the next day when they started it it was mega rich and was pumping out thick black smoke. They adjusted the tune and pulled out 25% fuel to bring it back into line. Then as the engine warmed and it got to around 50 degrees it started to lean itself out and needed the fuel added again. They shut it down and let it cool, then started it again and it did the same thing.

They though the pump might be dying (even though it made power) so they put a fuel pressure/flow gauge on it and watch it again as the engine warmed and once again it went rich then lean but the pressure and flow were unchanged.

They then scoped the injectors to see what was happening at the injector and there was no change in the firing time with the change in AFR's. They then fitted another set of the same injectors the same thing happened, they removed the resistor and changed the impedance at the ECU there was no change, the voltage is only fluctuating by 0.1V-0.2V so it's not latency related.

They though that it might be the ECU itself even though the injector scope said that the firing time wasn't changing, so another ECU was used and it changed nothing.

The engine was previously running a Power FC L-Jetro and there was no issue with the AFR's changing themselves.

The Tuner is Simon at Morpowa and he has rang CRD, Godzilla Motorsport, and Haltech themselves and no one can think of what would cause the symptoms. Haltech have now been sent the tune to run on their simulator so see if that can shed any light but when they were trying to diagnose the fault Simon fixed all the cells used for idle to the same value to make sure that the ECU wouldn't be changing injector time, but that had no effect on the issue.

I spoke to Simon yesterday and he gave me a rundown but any specific question that he didn't give me the info for will take a day to get the answers for the thread.

Has anyone got any ideas or come across anything similar?

Edited September 17, 2011 by D_Stirls

[Adriano]

Is the ecu temp reading closely, what water temp compensation is it running?

[Guilt-Toy]

make sure they have the correct relay on the main power for the ECU.

[D_Stirls]

They had played with both the ATC and WTC initially and it had no effect on the issue. But both those corrections it would have changed the injector firing time and that would have been seen on the scope.

[Rolls]

Graph all injector values vs time and compare when it is fine and when it isn't, see if stuff like pulsewidth/latency is changing etc. This way you don't need to look for what COULD change these things, you can just look at what IS changing and then work backwards. This is assuming the ECU is powerful enough to graph random variables like pulse width etc which would never normally change.

Also get someone to look at the fuel pressure when it is fine and when it isn't making sure it doesn't change.

Ultimately it sounds like an ECU issue where one of these variables is getting reset or changing for whatever reason, either that or some weird trim/compensation map is kicking in.

Something else worth graphing is the rpm and load (whether it is afm or map) when this happens, see if the RPM signal changes dramatically or the load suddenly doubles even though it hasn't actually. Could be a faulty sensor.

Edited September 17, 2011 by Rolls

[T04GTR]

what "electrical gremlins" were "sorted out" before the tune?. using onboard map sensor in the ecu, or an external one? doyou have the full injector dead time figures (8-16v)? if not SET THEM TO ZERO. remove injector resistor. set injectors in the main set up table to the propper setting.(drop down menu explaines)

was it tuned throttle over map,afm,or just map? ve or injection time? set coolant temp correction to zero and warm up. (all this will have to be done at the dyno)

injector latency is not nessesary, all it does is add a time to each event. if it is not set correctly the thing wont evan start. (had the same issue at work with id1000's. there values were heaps differant to the default haltech ones. changed them and started first kick.

so if you dont have the exact documentation on the values, set it to zero as it will make tuning and cold starts uber hard with the varing crank voltage.

[Lithium]

Under what conditions would it "lean itself out"? Was it only idle, or under load, or everything? Just trying to work out whether you should be trying to resolve an issue with too much fuel, or conversely - not enough air going through the engine as the conditions change for some reason.

[D_Stirls]

From my understanding it was lean everywhere, I'll have to get more answers on Monday as I'm not the one working on the car.

[blind_elk]

How is the surge tank plumbed? Is it possible that hot (less dense) fuel is being fed to the engine after warmup?

[Guilt-Toy]

check relay power to the ecu. make sure its wired correctly with correct parts.

[Crans]

Can you provide more details on the ecu relay? I'm having similar issues on my 33 gtst random leans outs and miss firing at idle. during the lean out i am not seeing a change in injector duty cycle

[zebra]

Are they saying its leaning out by just looking at the Wideband read out, or is it lean popping and farting etc?

[Crans]

Lean on the wideband as well having a rough idle and miss fires(Then again it miss fires idling at stoich)

[zebra]

Myself and quite a few other people have found that RB's need to be fairly rich to idle properly.

But that is another matter.

If you have swapped ECU's and it is still doing the same thing than that is not the cause.

I'm also dudious that there is no change in injector pulses, have you checked the voltage at the injectors when both lean and rich?

[Crans]

I have basically swapped out everything apart from ECU and FPR.

I'll borrow my mates occeloscope to check injector voltage.

I suspect my issue is electrical, also suspec OPer's issue is electrical aswell.

I took my car to an auto electrician to be told nothing is wrong. I see an alternator putting out 13v and the ecu seeing 10.5v as a problem

New alternator has mostly resolved my voltage issue tho, but still seeing all these weird problems.

I will be putting in a PFC and Sard FPR as testers ASAP

[blk94r33]

Under what conditions would it "lean itself out"? Was it only idle, or under load, or everything? Just trying to work out whether you should be trying to resolve an issue with too much fuel, or conversely - not enough air going through the engine as the conditions change for some reason.

+1

Fuel presure, flow, and injector duty cycle are the same, meaning it *shouldnt* be fuel, so not enough air?

[Crans]

Amount of fuel to flow out of the injector is dependant on the opening time(For this purpose we will ignore injector lag times) and fuel pressure.

The more fuel pressure the more fuel will flow at the same opening time(This is also discounting changes in manifold pressure)

If you have a faulty FPR it could make the fuelling go all over the place.

That said if injectors are not getting a constant good voltage source, injector opening times will change, this will change the amount of fuel being injected but wont nessesarly change the pulsewidth's especially if the ECU is not seeing the same voltage drop

[Rolls]

As I said you need to graph all injector values to see why it is getting more fuel. Opening time, pulse width, fuel pressure, voltage.

Fuel pressure I imagine you'll have to have someone sitting there looking at a gauge but it is better than nothing.

[Crans]

As I said you need to graph all injector values to see why it is getting more fuel. Opening time, pulse width, fuel pressure, voltage.

Fuel pressure I imagine you'll have to have someone sitting there looking at a gauge but it is better than nothing.

Agreed totally with this, I will be attempting this on the weekend, just need to acquire the tools or some one with the tools to do this.

[Lithium]

As I said you need to graph all injector values to see why it is getting more fuel. Opening time, pulse width, fuel pressure, voltage.

Fuel pressure I imagine you'll have to have someone sitting there looking at a gauge but it is better than nothing.

Given the people involved, the diagnostics they have been resorting to and generally giving them credit I'd think (or hope) that its almost insulting to make these suggestions on the assumption they weren't the first points of call. It'd be very interesting to see a datalog file as the car warms up, however - if that were a possibility?