PI : DI Ratio Discussion

[KoolBRZ]

Originally Posted by thambu19

If yourintent is best engine response and can sacrifice fuel economy there are other cool things we can do with cams at lower loads. Maybe better for a sports mode calibration if we can switch between the two at some stage. I'd calibrate the cams completely differently if I want performance instead of mileage

Sent from my HTC One_M8 using Tapatalk

How on earth did you figure what numbers to use where? It transitions smoothly the way it is. Should I try using more than .7 at .2 and .3 load? How about using 1.0 PI only at .2 or .3 load in the low rpms?

[KoolBRZ]

Here's my latest attempts to understand and push the limits of the @thambu19 principle. I'm going to try out these four shown.

[thambu19]

Originally Posted by KoolBRZ

How on earth did you figure what numbers to use where? It transitions smoothly the way it is. Should I try using more than .7 at .2 and .3 load? How about using 1.0 PI only at .2 or .3 load in the low rpms?

I am pleasantly surprised it worked. To be honest I plucked those numbers from my a@#. Few simple guidelines were followed
1. Going DI at low speed knock limited regions will increase volumetric efficiency and improve knock resistance
2. But going DI will make combustion slower so need to add more spark
3. The lower the speed the sooner the transition from PFI to DI because of same reason above.
4. Use PFI at loads that are not knock limited to improve fuel mixing and take advantage of faster combustion of PFI. This should reduce combustion variance and make engine feel more stable (less stumble)
5. Make sure to use 100% DI at the torque dip region
6. If it was upto me I would have tried 100% PFI at Idle. I think Toyota went for DI there for emission gains and to avoid knock during HOT RESTART. With PFI they need to model puddle dynamics when the engine is cold and other stuff. With DI they get better fuel control. I dont know if they are even going slightly lean at Idle for fuel economy reasons.
7. Going 100% PFI should also reduce the chirping??? I dont know for sure but worth a try. To really reduce chirping the fuel pressure must be dropped for DI at idle and then going 100% PFI. This is because even if DI isnt being used the pump is still working to maintain fuel pressure. Currently 24bar or 2.4mPa fuel pressure is commanded by the DI system at idle raising to 200bar pressure at WOT.

[thambu19]

OEMs usually play the emissions game rather seriously. Like VERY SERIOUSLY. For example look at this stock map. They went from 50% PFI to 100% DI at 2400rpm because that is where Scavenging peaks in this engine. So that is where going PFI will make the fuel short circuit engine and go directly to exhaust causing HC emissions. So by going DI it allows them to scavenge harder. Upto 1-2% O2 in the exhaust is manageable by the CAT. Anything over that the CAT cannot keep up.

[thambu19]

If you look at this posting by @Shiv@Openflash
http://www.ft86club.com/forums/showp...73&postcount=1
you will see the two distinct peaks on either side of the torque dip. I have a strong suspicion that is it because of the unequal length header working. Equal length headers both tune at the same engine speed but unequal length headers should tune at two different engine speed. The side with shorter tube will tune at higher engine speed (4500rpm?) and the long one tunes at lower speeds (2400rpm?)
What he had done is create more overlap by moving the exhaust closer to intake or retarding the exhaust timing and maybe increasing intake advance and this maximises scavenging at the cost of HC emissions and possibly intake noise. The loud HONK you hear when the engine hits 4K rpm is due to this effect where intake valve is opened so early that combustion gases gush into it at the same time they are escaping through the exhaust. The more the overlap, the louder the noise and OEMs limit the NVH beyond a point. We love the HONK so we can do what we want with the timings

[steve99]

Originally Posted by thambu19

Hey @ztan. I bought a Tactrix but the unit failed to connect/read so sent it back and they shipped me another. Should arrive in couple of days. Since mine is 2013 MY im sure its pre B01C. I want the dealer to update it to B01C since they introduced the Pre-Ignition and idle fixes with it? Im worried that if I update it myself they will find out that the calid is different when I take it in for some warranty work.

I am itching to try out some PFI/DI cal changes and E85. The Takeda Intake has arrived and I want to install that as well with the right MAF scaling.

I love the intake honk from 4K onwards but pretty sure I can make it happen even earlier by changing the AVCS. But that is pretty much what OFT OTS cal does too. They moved the Exh timing creating more overlap and this should make more intake noise although their reason for doing it was for more torque and less torque dip

@KoolBRZ Thanks a ton mate. Got a chance to try out 100% DI? I don't know why the OEM did not calibrate for full DI at low speed WOT. Defeats all my engine knowledge. I can only think of emissions as a likely reason for staying with 50% PFI DI split.

I have found the pins intactrix adapters are quite thin, and the obd sockets on these cars or at least 2013 cars are a bit loose, resulting in most tactrix adapters getting intermittent or no connection. Use a piece of plastic a full size SD memory card works and bend both rows of pins on the tactrix adapter inwards slightly only need less than a milimetre on each row and you will be good.

[steve99]

@thambu19 how much PI is needed and where load rpm wise to maintain intake valve cleaning properties ?

another question, if you have a look at the intake cam tables at low load rpm they do a weird dance go from zero then negitive then back to zero then increase above zero. I have assumed the negitive dance was emmissions related ?

[thambu19]

Originally Posted by steve99

I have found the pins intactrix adapters are quite thin, and the obd sockets on these cars or at least 2013 cars are a bit loose, resulting in most tactrix adapters getting intermittent or no connection. Use a piece of plastic a full size SD memory card works and bend both rows of pins on the tactrix adapter inwards slightly only need less than a milimetre on each row and you will be good.

I will try this out when I get my new one. If it does not connect again it could just my be OBD port and will ask the dealer to look at it while I am there

[thambu19]

Originally Posted by steve99

@thambu19 how much PI is needed and where load rpm wise to maintain intake valve cleaning properties ?

another question, if you have a look at the intake cam tables at low load rpm they do a weird dance go from zero then negitive then back to zero then increase above zero. I have assumed the negitive dance was emmissions related ?

@steve99
PFI at extremely low PWs (pulsewidths) wont really help clean the valve (like idle) because the injection duration is so short the gas fully atomizes rather quickly. What is really needed for clean up is substantial PW. I cannot tell how much PW in ms exactly because I do not know to be honest. These engine will start seeing knock at 40-50% load at speeds between 1000 and 3000rpm and higher loads at higher engine speeds so increasing the PFI% below that is always going to help valve cleaning.

Regarding the Intake cam timing I will need to make some wild assumptions shall I?
1. Generally 1200rpm to 2500rpm is the region this car would run in an FTP cylcle (one of the fuel economy test cycles)
2. Toyota wanted to go to LATE Exhaust Opening as soon as they could. So you will see Exh timing go to 40 as soon as they can. They cannot go to 40 too early in load because if they do they deteriorate the blow down of exh gases causing combustion quality issues. The reason for going late Exh opening/closing or retarded timing is for Fuel economy to get maximum expansion of the gases in cylinder.
3. With Exhaust timing so retarded the intake cannot be advanced beyond a point. If advanced it will cause the exhaust to be pushed into intake manifold during the exhaust stroke and rebreathed during intake stroke and this will increase incylinder residuals causing combustion quality issue. So by going Late Intake opening/Closing (-10) they are using a late/late strategy. Meaning open Exh late and intake Late.
4. By moving the intake late they are reducing the Dynamic Compression Ratio of the engine and this helps in reducing pumping losses (something like drag) and also helps in reducing knock.

So overall this would have been a move to maximise fuel economy. Once engine load is at 0.7 the intake is being moved to a more advanced position because
1. The throttle has maxed out in authority or in other words the throttle is almost fully open and the MAP is already 90kPa or above and the only way to get more air to flow is to change cam timing (throttle is maxed out)
2. At this time scavenging kicks in and moving intake to more advanced position helps with scavenging
3. They can simply tolerate more internal residuals at these higher loads

[KoolBRZ]

Here are 3 more attempts to improve on @thambu19's table. His Total Injection PI to DI ratio table is so smooth. There is only a slight delay transitioning to full DI, and I think timing advance could fix that. The question I want to answer with these attempts is really "How much is too much?".

[thambu19]

@KoolBRZ you definitely need to adjust the timing where we have increased DI % from stock cal

Everytime you flash are you making sure to relearn the background noise of the knock sensor by doing a partial load sweep from low to high engine speed? Normally you start at 2K rpm low load and slowly let rpm build until 6K to 7K (not WOT). This helps ECU learn background noise caused by valvetrain and other components. Helps ECU call knock more accurately and overall a consistent knock detection instead of too early or too late.

My only concern with flashing is ECU loses the chain stretch information. This is usually accounted for because when the company flashes a cal at the plant the engine will learn the relationship between crank and cam and then on wards it will use this relationship to estimate how much the chain has stretched. For example initially when the chain stretches a lot when new the cams can get more and more delayed in actuation or cams can get retarded. Some OEMs can use the chain stretch info to position the cam after discounting the stretch. What I mean is physically the Intake cam could have moved to 42deg instead of 40deg retard. What Ecu can do is move the Cam to 2deg more advance thereby engine is actually getting the needed 40deg instead of 42deg.

[jvincent]

I'm subbed to this thread, and it's a very interesting discussion, but I have a question for the modders.

What is your ultimate goal with the changes you are making?

- Increased power for the track?
- Better driveability (i.e. smoother acceleration, reduced knock, etc) ?
- Improved mileage?

[thambu19]

Originally Posted by jvincent

I'm subbed to this thread, and it's a very interesting discussion, but I have a question for the modders.

What is your ultimate goal with the changes you are making?

- Increased power for the track?
- Better driveability (i.e. smoother acceleration, reduced knock, etc) ?
- Improved mileage?

If I were to guess Id say increased top end torque, better drivability.

Drivability and fuel economy is always a trade off because:
1. For FE you move the cams in such a way as to increase MAP as soon as possible and achieve de-throttling this kills driveability
2. If one wants pure driveability, id move the cam timing more towards WOT cam timing earlier in the load. I mean blend the cams towards WOT cam timing not just pop into WOT cams timing. Then Id increase spark at the lower loads to slightly oversparked by a degree or so (non knock limited region). This will improve combustion stabilty at the cost of some FE.

So yea to answer the question we are trying to understand the engine better that way each person can tune it to his/her liking. Some guys are all about FE and we can do that and some are all about performance (like me) and we can do that too.

[KoolBRZ]

I want drive-ability and power, and I want it smooth with no flat spots. I want smooth economy up to .7 load, and acceleration from .7 to 1.0 load, then sheer power from 1.0 to ~3.0 load. Yes, I am slightly super-charged with the Phantom ESC, and that's how I can get to 3.0 load. That change thambu19 made to the PI to DI ratio table has made more of a difference to open-loop acceleration than any other improvement I have made so far. Closed loop is pretty good too. Since I am super-charged I can't boost my timing by too much, since I am already using Shiv's Phantom ESC tune's timing tables. If I'm having a hard time improving on his table, that's my way of showing respect for his work. I'll always keep trying though.
P.S. @thambu19 can you please get an OFT? There are always a few for sale in the member classifieds. OFT users need your input. I can also send you a tune with my revised AVCS tables, and you can show me where I went wrong.
PP.S. Do I make timing advances to Base timing A, B, or some other table?