Road tuning your R33 with PFC. How do you work around a PFC?

[JasonO]

R31 Power,

I was involved in a lengthy debate about this over on SDU a while back, we never really finally figured it out but we did discover a few critical items.

You need to have the idle set up properly, before expecting the ECU to be able to control it. To do this you need to unplug the Idle Control Valve, and adjust the Idle Set Screw on the side of the AAC assembly to get your idle into a ball park of where the PFC expects it to be.

You also need to adjust the idle stop on the throttle body, but only if the butterfly is not closing cleanly when you lift off.

You also need to set the Throttle Position Sensor (TPS) voltage accurately to show less than 0.5V at idle, 0.49 would provide the best response, but may not always appear closed, so aim for 0.45V. You can see this voltage on the PFC Commander while you adjust it. To adjust the voltage, you need to loosen the two screws on the side of the throttle body, leaving them just nipped up enough that you can lightly tap the TPS around on the spindle to adjust the voltage. Tap the TPS around while observing the Commander which you've pulled through to the engine bay. Make sure you tighten both the screws when your finished adjusting it.

First thing I would check is the TPS voltage, it only takes a second. Then make adjustments from there.

I'll see if I can find the thread on SDU, and link to it for you.

[JasonO]

R33 Gts25t idle instability issues... thread on SDU.

It's a long read, but just about everything is covered init.

[R31 POWER]

Thanks guys will give me something to tomorrow.

Steve it has mesh on both sides but the one closest to the turbo is frayed abit also it is a little lean at idle to so I will richen it up.Thanks

JasonO the tps is at 0.43 at idle is that ok or should I put it to 0.45 :headspin:

[Revhead]

My TPS is at 0.65V at 0% throttle. Setting it here may have fixed the hunting problem I had, but I did a few other things at the same time so it could have been any of them.

[JasonO]

TPS at 0.43V will be OK, but may cause the engine to kind of stutter/surge a bit on very light throttle, ie. when you are in very slow moving traffic. If it doesn't stutter/surge, then I'd leave it as is.

TPS at 0.6V will be enough to lift the ECU out of the Idle Control Loop, so effectively it's running on a single cell of the INJ/IGN map and may not be the most economic idle available to you. You will most likely find that at this setting, it struggles to maintain a level idle when you switch on the Air Con, or put an electrical load on the engine.

[Revhead]

Aah ok, nah mine seems fine with load on or off the engine. I've just had a look in the engine manual, they recommend between 0.4 -> 0.5V at 0% throttle.

[R31 POWER]

When I back off the throttle it does'nt de-accelerate straight away,You know when you back off the zorst goes booooooor well mine goes boooor vrom booooooooor like someones pressed the go pedal.Could the TPS be doing that?

[Steve]

R31, sounds bizarre. When you lift off, the power fc cuts fuel, or it should. Have you checked what your fuel cut recovery rpm is set at? also, have you checked the throttle isnt sticking at WOT?

I used to have my TPS voltage at 0.61V, as anything under 0.6 and it wouldnt idle properly.

The AAC valve still did its thing, and the car idled ok, except if the voltage fluctuated enough to bring it to 0.59 where the throttle (AFAIK) indicates shut (between 0.4 and 0.6V)

After finally finding the smallest vacuum leak - It now idles with the TPS voltage set to 0.48. I also had issues with my mesh, the rear was removed - I was told by a jap performance mechanic to replace the mesh, as air can revert off the compressor wheel. That worked well.

I adjusted the AAC vavle a little different to JasonO. I just wound it open, until the revs started to raise, then screwed in until it started to struggle to breathe. From there I went to the mid way point.

The final thing I did was replace my O2 sensor, now the idle is damn near perfect, decel is perfect - its all good. The only problem is (well this is hypothetical) the cold air duct is 4" and it has a funnel at the front behind a hole in the front bar. The slightest breeze into it registers on the AFM - when its really windy, the voltage seems to jump enough that it appears to throw the idle a bit. I will be relocating the CAI in the near future to test the theory.

Oh, and letting the car sit at idle for 10 minutes is how the power fc learns its idle, that seemed to work nicely too:)

[R31 POWER]

Thanks Steve and guys for taking time to anwser

Steve when you say "Have you checked what your fuel cut recovery rpm is set at?" do you mean that injector table.The throttle does'nt stick coz I was checking it out while we did the piping.

Can the O2 sensor affect idle,I thought it was for closed loop it came into play?Mines done 90,000km's so it's maybe time to buy.

I could'nt put my air-con on during the learning process coz it's not hooked upto the R32 loom yet but at times it hunts but is getting better.Can the learning process still be done if the PFC has a map in it or do I have to start agian?

[Steve]

fuel cut recovery rpm is set in the idle option. When you back off, the ecu cuts fuel, the fuel cut recovery rpm is where the idle will pick back up again. Mine is set for 1150 from idle and 1200 from idle with AC on. Idle is set to 900 and 950.

I did find it recovers better after changing the O2 sensor. My O2 sensor was cactus when I changed it, and I had only 68,000ks on the clock which I believe are genuine:)

Have you got someone to try part throttle and WOT with the engine off while you are looking at the throttle linkages? The only reason I ask is it sounds to me like it may be sticking - sometimes even a vacuum line in the way or a dirty/frayed accel cable can cause problems.

Self learn will work even if you have it mapped - just leave it sitting for 10 minutes. But, if you havent changed your O2 sensor recently, I would definately recommend you do this first - only cost me $80 for an EL falcon one which is easy enough to adapt with a soldering iron. (interesting, the nissan stock sensor is bosch, the EL ford one I got was NGK????)

Good luck, thats a nice clean looking engine bay there.

[R31 POWER]

Thanks Steve I'll look into a Oxy sensor and the throttle thing and put those idle adjustments in I've done the idle from there before but did'nt know what it was called LOL.

[Robo's]

Im just bringing up an old thread, but how you measuring A/F ratios without a dyno?

[Guilt-Toy]

Robos: You really need a Wide band air fuel ratio meter which you can buy for about $2500 from Autronic...

Otherwise there is no real way of truley knowing the correct air fuel ratio

[Robo's]

Thanks mate, looks like i will stick it on a dyno then

[Emre]

Im just bringing up an old thread, but how you measuring A/F ratios without a dyno?

The dyno itself doesn't measure your AFRs, the AFR meter is used which is linked with the dyno software.

The Autronic ones cost a fair amount as mentioned above, but there are some cheaper alternatives out there for the interested individual.

I've been meaning to get one of these for quite some time now, but haven't got around to it.

[bombastic]

Suppose AF ratios were set at optimum levels on a dyno, and you have the datalogit equipment. Suppose I have the brains to interface with the software.

Rather than having a passenger adjust the ignition, have a laptop do it for you. Start with the base map and let it advance until a certain level of knock is detected, then back off and lock the value.

Plug the laptop in every time you drive. Drive it like a hat driver. Drive it like you just got your L's. Drive it like a soccer mum. Drive it like you stole it. A few insurance claims and gaol sentences later all the points on the map will be perfect.

Now why hasn't anyone done this yet? I don't see why it can't be done, since if I tune it myself I'm just going to be applying the same algorithm. Yes, I am a robot.

[Sydneykid]

Suppose AF ratios were set at optimum levels on a dyno, and you have the datalogit equipment. Suppose I have the brains to interface with the software.

Rather than having a passenger adjust the ignition, have a laptop do it for you. Start with the base map and let it advance until a certain level of knock is detected, then back off and lock the value.

Plug the laptop in every time you drive. Drive it like a hat driver. Drive it like you just got your L's. Drive it like a soccer mum. Drive it like you stole it. A few insurance claims and gaol sentences later all the points on the map will be perfect.

Now why hasn't anyone done this yet? I don't see why it can't be done, since if I tune it myself I'm just going to be applying the same algorithm. Yes, I am a robot.

What you are describing is closed loop tuning, or Self Tune as it is called in Motec and Autronic speak. You set an A/F ratio target and the ECU self adjusts until it achieves that target. For example you pick 12 to 1 A/F ratio and the ECU merrily goes off and tunes to that A/F ratio as you drive around. It is surprisingly quick, I saw Ken Douglas (ex Motec GM) tune a production Mazda in 2 laps of Bathurst using self tune.

The problem is the engine will not produce its best power at 12 to 1 A/F ratio at all rpm ranges, sometimes it needs 11.8, 11.9 or 12.1 to produce the most power. What you are aiming for is a smooth power curve, not a smooth A/F ratio. To achieve that you need a dyno.

What I have been doing for a few years now is tuning on the road (and/or track) then going to the dyno and tidying up. Whether I use self tune or not, I always find (actually Dyno Dave always finds) a few more horsepower on the dyno.

Check out the Tech Edge thread for more details

cheers

[rob82]

Suppose AF ratios were set at optimum levels on a dyno, and you have the datalogit equipment. Suppose I have the brains to interface with the software.

Rather than having a passenger adjust the ignition, have a laptop do it for you. Start with the base map and let it advance until a certain level of knock is detected, then back off and lock the value.

Plug the laptop in every time you drive. Drive it like a hat driver. Drive it like you just got your L's. Drive it like a soccer mum. Drive it like you stole it. A few insurance claims and gaol sentences later all the points on the map will be perfect.

Now why hasn't anyone done this yet? I don't see why it can't be done, since if I tune it myself I'm just going to be applying the same algorithm. Yes, I am a robot.

The problem with that is your algortihm for leaning ignition in has to be very good otherwise your going to run into big problems. The factory subaru system post 02 is capable of learning igntion timing and can be quite a fiddle to tune with. The other issue is that the knock sensor wont pick up all detonation. This is very evident when tuning at low loads and rpm as if you've ever tuned a PFC you can hear detonation through the chassis ear but it doesn't register on the hand controller.

[paulr33]

my experiences with tuning are here;

http://www.skylinesaustralia.com/forums/in...ic=79377&st=260

[SATO GTS]

Hey Guilt-toy

Your first comment about your mate going out in a 31 with a 25pfc, that was me and anthony with his old boys autronic many years ago and yes i only had the hand controller back then. My god how times have changed using a laptop and datalogit software.

I cant believe i have not viewed this thread earlier.