Nistune Forums

Hi guys.

I've been using the autotune feature to dial in the correct AFRs for my currently stock spec vehicle.

The vehicle is a 44f ecu manual swap without hot start solenoid or crank case breather(intake pipes blocked) and no neutral switch set up.

The image shows the delta compare of the autotuned injector values compared to the stock D009D map, which has me a little concerned. I set it to 2%, using 465mV decel cut.

I have an adjustable fpr, which has been shown to be set correctly to 43psi above reference pressure.

I reconditioned my stock injectors a little while ago with new seals, caps and filters so whether this is effecting them I don't know.

I did drill out the maf potentiometer, however I believe I have put it back to as close to original as possible, but I would expect any problem with this to be pretty even all the way along the map.

Every now and then I pull over and glance at the screen and notice the lambda has fixed to 18.60 on the nistune, and LIMIT REACHED" is written in autotune, whether this is causing the bad behaviour or not I don't know.

Using the stock values, the boost lags, and until the boost actually kicks in it stays around 17 afr on my AEM UEGO.

I am very appreciative of all your help Matt.

Regards,

Pete

Don't use the auto tune if it causing issues. It is only needed if fine tuning the maps to your intended AFRs

In this case it is adjusting the map values based on your AFR trace output which seems to be off. Check your stoich mixtures first at cruise are correct with factory fuel maps

Hi Matt,

My primary concern was why there could be such a big error using a stock engine on a stock map. The engine pulls over 7.25 psi vacuum, so I would have thought a major boost leak after the manifold would be unlikely.

Using the autotune is just about the only way I can actually drive the car without running lean on acceleration. The non linear adjustment screams out a possible unreliable readout of a sensor which could be causing problems. What do you think?

I am wondering whether the replacement injector caps may have had a slightly different sized hole, I bought them from Turbonoz I believe.

Cheers,

Pete

Oh, I should point out, I am using my AEM UEGO through RS232 mode so that I do not get the inaccuracy from the arduino DLP input voltage. What kind of 'update' value do you think makes sense for an rs232 connection? I remember a bit of documentation saying it is only 9600 baud so doesn't update too quickly.

Cheers

I put my stock air box back in incase the cone filter I had on was causing turbulence in the intake, but it seems it is not the problem.

When I am pulling away the map seems to get stuck around the cell which is in the attached photo, which happens to be where the 0 transitions to 15/20. This is when the aem uego is saying either --- for lean or 17, then when boost finally kicks in it richens.

Using the autotune is just about the only way I can actually drive the car without running lean on acceleration

There is another problem here

The non linear adjustment screams out a possible unreliable readout of a sensor which could be causing problems. What do you think?

No enough information to comment

I am wondering whether the replacement injector caps may have had a slightly different sized hole, I bought them from Turbonoz I believe

So what is the setup?

44Fxx ECU with factory CC injectors and MAF? However the pot adjustment on MAF is removed (which will affect mixtures) and are injectors flow tested? What CC?

So car runs lean on factory 44Fxx tune file?
Take a log of a run with AFRs included so we can see AFRs on stock tune file

Is there anyway of reading the adc that the mixture tuning potentiometer on the maf is connected to in the hardware/software?

I think I will try to use the estimated pulse width in the fuel mixture correction page to adjust the potentiometer to get the idle pulse width correct, and hopefully that will sort the problem.

The injectors and maf are all stock.

I will take datalogs before and after.

There are two datalogs attached, both are slow roads, before a fast road, then slow roads again. First one is coming up to temperature, and second one is up to temperature, with a few turns out on the MAF potentiometer. It didn't seem to alter the injector duty cycle or AFR at idle, but the fueling on acceleration seems better.

Every now and then the lambda sensor reading in nistune seems to lock, so I hope it has taken enough correct readings to get useful readings off of it.

Thanks.

I have completely removed the mixture control pin from the ECU connector (Pin 37 thick green and yellow wire) on the assumption that the ecu will have a pull down resistor of some sort, and a lack of any input will cause it to revert to just the map. The hope is that without a signal going in, it will not float and cause random mixture adjustments.

Attached is a log file, to me it looks like the fueling is correct, however I am uncertain how much of a 'pinch of salt' I should take with the estimated air fuel ratios, especially on a non O2 vehicle.

Is there anyway of reading the adc that the mixture tuning potentiometer on the maf is connected to in the hardware/software?

Not directly. If we knew the memory location, could use RAM tracer. I dont know this location at the moment

I think I will try to use the estimated pulse width in the fuel mixture correction page to adjust the potentiometer

I would not do that... estimate is a rough guess. Put vehicle in cruise until 14.7:1 and O2 feedback 0% then adjust pot

I have completely removed the mixture control pin from the ECU connector (Pin 37 thick green and yellow wire) on the assumption that the ecu will have a pull down resistor of some sort

Wouldn't do that. About 370ohm was used on RB30 AFM screws. Offset the mixtures but would need to read code gain to see how (additive or multiplier)

Last log shows around 15.19 on cruise (2400rpm). bit lean

Attached is a log file, to me it looks like the fueling is correct, however I am uncertain how much of a 'pinch of salt' I should take with the estimated air fuel ratios, especially on a non O2 vehicle.

ECU has no idea of actual AFRs. Put in filtered view and use the TRACE window viewing your wideband AFR instead

About the MAF potentiometer, I can see 2 ways it may be wired up.

1) both elements of a voltage divider in the maf, connected to what I think is battery voltage, and ground. This would mean the signal voltage would vary with battery voltage.

2) only half of the potentiometer is used, and the other half of the voltage divider is mounted inside the ECU.

I remember during the first log, the resistance from signal to either ground or + was 1500 ohms, and the other was 500ohms.

Last log shows around 15.19 on cruise (2400rpm). bit lean

What point along the Load axis would cruising be (or rather, where I should I aim to get to 14.7)? When looking at the tracer, there doesn't seem to be any regularity with what cell is running at afr 10, and what is afr 16. I presume once the pot is tuned correctly these will come back into order.

It does seem confusing that the tracer puts in overrun/deceleration plots in as well, when the tps isn't fully closed, but less than needed to sustain speed.

From quick glances at the afr gauge while cruising, if I put a tiny bit of throttle on it would shoot down to 13 afr, and releasing it, going up to 18.5 afr. I don't really know how precise a primitive 1980s control system should be, and how responsive, and repeatable to altering changes in afr.

Forgive me if I am going around in circles, it is midnight here

About the MAF potentiometer, I can see 2 ways it may be wired up

RB30 ECUs pass a signal directly to ECU. Yours does this also I believe. Put it back to factory to fix it properly or find correct resistance and use a resistor

Cruise about 2000rpm. Usually column 2-3 and steady

Hi,

Yes, setting the pot to 370 ohms has sorted the problem out, the fueling is pretty good now, no random leaning out.

Cheers,

Pete