Pfc Tps/inj D-jetro Rb26

[edizio]

Hi, I've been playing around with tuning my car for a while on top of the base map that I have and it's been quite successful. Full throttle is perfect, but of course, the problem is part throttle. From what I have been told, the PFC d-jetro is incapable of ever getting good results for part throttle on a GTR due to the multiple throttle bodies plennum.

I've been told that the problem can be solved, using TPS/INJ corrections and although this will never get perfect results, it will, at least, give drivable ones. Currently, at about 5000rpm+ the AFRs drop into the low 10s and then into 9, at which point, I just have to give up or risk flooding.

My question is, what should the curve of the TPS graph look like? I'm guessing it's more complex than just a linear line but I wonder if there is any calculations I can do that will allow me to get an accurate 'estimation'

Furthermore, how do ECUs like LINK/SYVECS allow a better tune? (like I appreciate they have more tables for fuelling, but what are these?)

Edited March 15, 2015 by edizio

[GTRNUR]

Your issues at 5000+ RPM really just comes down to more tuning time is required. Higher engine speeds stabilise engine vacuum and MAP stability in general, regardless of what MAP sensor or how many throttles you have.

How big your cams are, and how many throttles you have will really only affect your low speed tuning. Even then you should be able to get a good lambda cruise tune, and even a decent idle provided your injectors are reasonable.

Use the Map watcher, and watch your wideband input channel and just start pulling fuel out of the map. It really will be that simple.

When I used a djetro in my car, I found the base maps to be less that useless. Ok for a startup (sort of)... but that's about all. Generally you need to completely re-work the map. Its not at all like an ljetro where a factory map will drive for the most part Ok in a stock car. The biggest piss off for the djetro for me was lack of atmospheric correction capability, which meant if I went up the nearby range the car would run like crap.

All the more modern ECUs have boat loads more capability. In comparison the old FC is really in a class of race only, and not much good for anything else.

[edizio]

Using ID1000 injectors; idle is generally ok except after boost it tends to be fairly rich for about 5 seconds and then stabilises to 13.9-14.2 which is where I have it set on my 264 cams. I've got cylinder 6 at 3%, 5 at 2% and 4 at 1% additional fuelling as I believe the plenum on the GTR to be rather poor but I can't be 100% certain that this is perfect as I don't have EGT sensors/widebands on each of the 6 pipes.

I'm getting the part throttle issue sorted a little more now but it's taking a good amount of time and I have to be quite careful with slight adjustments. That's definitely a shame about atmospheric correction, I run on high 10s on high boost and have a custom temp sensor which updates really fast so I can make fairly decent tweeks.

A bit clueless when it comes to the ignition timing though, I feel like that is an air I don't want to touch, even though I have a set of det cans that I made.

[GTRNUR]

I would suggest setting your injector trims back to 0 until you can set them up when the car is on the dyno. The correct method would be to run the car under low load in top gear, and monitor the egt's with an exhaust probe. Then trim up your values till you read the same on all your cylinders.

You shouldn't be making any changes to the tune or even driving it at full load at all unless you are monitoring the ignition map and checking for knock at the same time.

Turn on the Advanced sensor feature in FC Edit so you can use the Map Watcher to monitor for knock. Logged results will vary from car to car, but you will generally want to keep your knock values low 30's at the most. That is until you can have the car on a dyno, and properly monitor for knock with a set of headphones and a knock box. Use the min/max/avg view settings in your map watch so you can see trends beginning to form based on how your driving it.

[edizio]

I never touch the car without monitoring for knock and I listen with my det cans but I'm still, by no means, good enough to fully understand it all. My highest knock count is 11 but I don't trust the sensors as they can be largely inaccurate as well as measuring engine/transmission noise. I tuned WOT fairly well by increasing the timing steadily until I could detect the pitch of the engine changing and then pulled back 3 degrees of timing and then used the temperature sensor to retard timing if the value went over the temperature at which I was tuning at. It's quite fun actually but not having a dyno handy, I have to do this on private roads with friends willing to drive whilst I tune.

Advanced logging is always on but admittedly, I've never probed EGTs; is this common amongst tuners then? how exactly do they probe the individual outlets?

[GTRNUR]

Unless you have individual sensors in each runner, the only option is one of these:

http://www.ebay.com.au/itm/Best-IR-Infrared-Laser-Thermometer-Temperature-Measurer-Gun-Wireless-Temp-Tester-/290840273731?pt=LH_DefaultDomain_15&hash=item43b76e6343

[edizio]

That external temperature sensor looks really good! I might purchase one of those for when it goes on the dyno.

Can anyone tell me why my ignition timing drops down when I release the throttle? Is this something to do with overrun?

[edizio]

Infact, sorry, surely that temperature gun is pointless: -

• Temperature range: -32 to 350degree

[GTSBoy]

That particular IR pyrometer is useless, because it maxes out at 350°C and the ex manifold surface temperature would certainly get hotter than that. But you can buy much better ones with ~900-1000°C max for $100 ish.

You would have to hold the engine under load on the dyno for a reasonably long time to saturate the manifold runners with heat to be able to tell if there was any difference caused by gas temperature inside. You could then check all 6 of them (as long as you can see them - ie, no heat shield or turbo in the way) fairly quickly. The real problem though is that runners that are bends will probably read higher than runners that are straight out into the log part of a manifold. This is because you'd get better heat transfer from the gas to the wall on a bend. You'd have to have a fair bit of experience with playing this game to get reliable results out of it.

There's nothing wrong with drilling all the runners and fitting little 3mm thermocouples to all of them. Even if you don't have them hooked up to a readout all at the same time, or all the time, or even installed all the time, it would make the question of tuning to temperature so much easier to answer.

[LWO]

"The biggest piss off for the djetro for me was lack of atmospheric correction capability, which meant if I went up the nearby range the car would run like crap."

This breaks my brain a little. What am I missing? I thought that 15psi of post turbo mountain air would have the same o2 as 15psi of sea level air? Surely the map sensor is doing the atmospheric compensation you require? To create the 15psi of mountain air the turbo goes into a different efficiency zone as it is working harder now, you may visit areas of the map in the ecu you have not visited frequently and the waste gate spring will open a little earlier due to less pressure on the atmospheric side of it.

What am I missing? It doesn't physics well with me.

[GTRNUR]

Then answer to that brain bender is that I am referring to light loads and off boost drivability. At idle and low speed cruise, less atmosphere caused the map sensor to move about 500 on the PIM scale towards 0psi. This meant that the tune that was stable at sea level would end up rich when at altitude.

It didn't matter what map sensor I used. Apexi 3 bar, apexi 3.5 bar, GM 3 bar. They all handled atmospheric change terribly. In short the rate if any that they did work differently did not even remotely work with the FC.

Also with the laser IR thermometer I meant to show you what I was talking about. Don't buy a cheap ebay one as GTSboy said. It has to read at least 500-700 degrees. Also when you are checking header temps, you must be holding the engine at load (WOT/full load is good if your not also at full boost at that time) say 3000, but importantly it must also be a rich tune in the 12's at the time. You then take a reading from each header, and make your trims from there. It will require very small amounts, no more than 1-1.5% as even that on an ID1000 is a significant amount.

Edited March 26, 2015 by GTRNUR

[edizio]

GTRNUR, if I'm running 440cc injectors and at WOT, cylinder 6 requires 3% more fuelling, surely with an engine tuned and requiring 1000cc injectors at WOT would still require 3% more fuelling.

I wouldn't have thought that it would matter what injectors are used; if you're fuelling and the rear cylinders are running lean, then wouldn't the coefficient remain the same, no matter what injectors you use?

I use 3%,2% & 1% because I was under the impression that every GTR engine that uses the standard intake plenum will run lean on 4, 5 and 6. I'm concerned about EGTs but since I run water meth injection, the exhaust temperatures, shouldn't be that high.

[XGTRX]

Depends on what afr you are running at above 50% max boost. If you are on the conservative (richer) side, then wouldn't worry about it as 0.4 leaner or so wont make a huge difference, But if you are running closer to mid 12s afr then it may be worth considering.

[STATUS]

The PFC can be tuned extremely well when the map sensors are fitted correctly (drilled and tapped into runners, if you just run them off the vaccum bar your gonna have a bad time.

There are 2 sensors for a reason.

[GTSBoy]

There are 2 sensors for a reason.

Yeah, but most of that reason is that there are 2 AFM inputs to the original ECU and that just sort of carried over!

[LWO]

"Then answer to that brain bender is that I am referring to light loads and off boost drivability. At idle and low speed cruise, less atmosphere caused the map sensor to move about 500 on the PIM scale towards 0psi. This meant that the tune that was stable at sea level would end up rich when at altitude.

It didn't matter what map sensor I used. Apexi 3 bar, apexi 3.5 bar, GM 3 bar. They all handled atmospheric change terribly. In short the rate if any that they did work differently did not even remotely work with the FC."

Yeah nah brain still broken. Does not compute with my thoughts about map sensors. I understand aftermarket ecu's needing barometric compensation on NA cars when running a tps vs rpm map. But barometric compensation over a djetro map which already runs manifold pressure vs rpm map has me stumped.

Edited March 27, 2015 by LWO

[GTSBoy]

Yeah nah brain still broken. Does not compute with my thoughts about map sensors. I understand aftermarket ecu's needing barometric compensation on NA cars when running a tps vs rpm map. But barometric compensation over a djetro map which already runs manifold pressure vs rpm map has me stumped.

It's actually pretty simple. The MAP sensor measures the pressure in the inlet manifold. Let's say that it is currently measuring that pressure is 50 kPa absolute. That is 50 kPa above an absolute vacuum. If you are at sea level, then the atmospheric pressure outside the engine is 101.325 kPa absolute. That means there is -

101 - 50 = 51 kPa (give or take the 0.325 kPa I dropped off).

of pressure difference between the outside and the inside pushing air in against the various restriction in the way (which is mostly the throttle plate at that sort of MAP).

If the only pressure sensor that the ECU has is the MAP sensor, then the only number that the ECU knows about is the manifold pressure. If the VE table in the ECU is set up on the assumption that you are at sea level, then the ECU will calculate the correct amount of air flow into the engine for that MAP reading and the fuelling and the timing will be right.

If you are at, let's say, 2000m elevation, halfway up a nice mountain, then the atmospheric pressure is only 81 kPa absolute. If the MAP sensor is reading 50 kPa absolute, then there is only 81-50 = 31 kPa pushing air across the throttle plate. The MAP sensor says that the load is exactly the same, but the real air flow is actually VERY different to what it would have been at sea level. The throttle position to get that MAP would be quite different too, so even using TPS vs MAP for helping tuning is only going to be useful to a certain extent.

What you really need is for the ECU to have a sensor to tell it what the absolute pressure is outside the engine, so that it can account for the reduced flow obtained at any given MAP reading as you go up in elevation. And of course it also needs decent maths programmed in to work it out properly too.

[edizio]

In response to xgtrx's response, I run my boosted afrs at 11.5. Regardless though, surely all gtrs with the std plenum will go lean on 6 and surely that percentage is the same no matter what injectors you use. So id1000 should be no different.

[GTSBoy]

In response to xgtrx's response, I run my boosted afrs at 11.5. Regardless though, surely all gtrs with the std plenum will go lean on 6 and surely that percentage is the same no matter what injectors you use. So id1000 should be no different.

I suspect that there is a little misunderstanding between your side of the conversation and XGTRX's. You're saying that a 1% fiddle on injector is the same regardless of whether they are big (1000cc) injectors or small (500cc say) injectors. What adds 0.5ms to a small injector will add 0.25ms to a big one. And you're pretty much correct, assuming that the ECU has the resolution to drive the injectors to such fine tolerances. I suspect that XGTRX's point was along the lines that if you add 1% of duty cycle, rather than just 1% of "trim" then you will get a lot more fuel with a 1000 than a 500. Because 1% of duty cycle is similar to 1% of trim when up near full duty cycle, but it's a hell of a lot more when down at only 10% duty cycle. ie, 1% duty cycle added on to 10% duty cycle is an extra 10% fuel. 1% trim on 10% duty cycle is 10.1% duty cycle.

It would be a difference of interpretation rather than the basis of a proper e-fight.

[XGTRX]

Umm yeah basically what I was saying.... You just said it a whole lot better... Out of interest how lean does cylinder 5 and 6 get with a standard plenum.