PI : DI Ratio Discussion

[KoolBRZ]

Originally Posted by Kodename47

The ECU will show 14.7 at stoich for any fuel, so when it reads 14.7 you are at stoich for your fuel. The AFR reading is derived from the O2 sensor scale but the sensor works in lambda, lambda 1 is stoich and the sensor scale means that it shows 14.7. All you're doing is making it a percentage richer than that. I'm also not sure why you're doing what you're doing, all you're doing is reducing economy....

The real "workaround" for this would be to change your O2 scale, however the ECU will try and target 14.7 so you'll actually make it run leaner

Although I have no doubt you are correct, I sometimes doubt my understanding, so I thought a comparison would prove your point. I brought up the same tables in OFT OTS 91 Oct and E85 tunes. They are identical up to .5 load, then they branch off from there. Stoich is radically different for both fuels, yet they are identical. This proves your point, yet shows a different principle at work too. Above .5 and .6 load, there is a compensation applied that is unique to the fuel being used. The question now is, how to compensate for E10 above these loads to get the best economy and power.

[steve99]

Stoic ratio for petrol mixture is 14.7 parts air to 1 part fuel.
Stoic ratio for e85 mixture is 9.8 parts air to 1 part fuel.

The O2 sensor does not know how much air or fuel you started with. It only knows oxygen levels left in the spent charge. The O2 sensor reports in terms of lambda. If the lambda value is 1.0, the fuel is burned optimally, and started with the right mix. Whether you start with gas at 14.7:1 or e85 at 9.8:1, the lambda will be the same at the O2 sensor.

The ECU recalibrates the lamda value to read AFR for petrol that what you see in logs.

The fact that your running straight petrol or E85 does not matter to sensor and the ECU will try to still target stoic if you set the "afr" to 14.7.

The fuel trims will easily adjust for E10

All they really do in E85 tunes fueling wise is adjust the injector scaling so you don't run huge trims like 25%

[solidONE]

Here's a short article about afr's for flex fuel.
http://ethanolpro.tripod.com/id213.html

AFR to Lambda calculator:
http://www.wallaceracing.com/air-fuel-lambda.php

Should come in handy while dialing your fuel tables.

[thambu19]

@KoolBRZ what changes did you have to make to PFI BRZ and DI fuel pressure to get the LTFT back to where it was before the new PFI DI surface?

[steve99]

Originally Posted by thambu19

@KoolBRZ what changes did you have to make to PFI BRZ and DI fuel pressure to get the LTFT back to where it was before the new PFI DI surface?

People seem to be finding the port injection fueling is a bit low compared to the di.

Seems arround the 5% mark , i moved by PI injector scalar (injector scaling brz in romraider) down 3% which increased port fueling 3%, Did not touch th di pressures ect figured it was easier to match the port to the di than the other way round

[KoolBRZ]

Originally Posted by thambu19

@KoolBRZ what changes did you have to make to PFI BRZ and DI fuel pressure to get the LTFT back to where it was before the new PFI DI surface?

I increased the Injector Scaling BRZ by 3%, than by an additional 1.6%. I think 5% should be a good place to start. Load, drive, use the Shiv quick-learn method of accelerating gradually from 2000 to 5000, log, then make changes based on the log.
I'm using custom AVCS mapping, so my trims are going to be off until I re-scale my MAF to suit. I'm making the major changes first, then worrying about fuel trims last, so long as they are trimming richer from lean. That way if it applies the trims to Open Loop it will only make it richer, which is much safer than applying lean trims to Open Loop. I'm also experimenting with large cell groups of -.001 compensations to the closed group compensation tables. It had been running too rich at low loads with less power and worse economy. It's also more unsafe to run rich in Closed loop, because it applies those trims to Open loop. I'm finding out it has more power at Stoich in the low load areas of .3 through .6, and from 1600 RPM's through 2800. I'm using more increased Tip-in instead, to compensate for low-load throttle changes. More power, and more economical.

[thambu19]

Thanks! I am about to try the PFI DI map. Wanted to make sure my baseline is good before I did that. I have a Takeda intake so have to work on some MAF scaling and get trims and OL AFR correct before messing with PFI ratios. I am starting with the scaling @steve99 had posted

[KoolBRZ]

Remember, it is far far better to make it leaned out in Closed loop, so it applies richer trims in Open loop. You can make up the fuel needed to accelerate with added Tip-in, and if you really need to accelerate, you'll be in Open loop anyway. I'm beginning to think that the entire closed loop should be Stoich. I'm going to try this out today. I left the temp comp tables alone, I still think they are necessary.

[ztan]

Originally Posted by KoolBRZ

Remember, it is far far better to make it leaned out in Closed loop, so it applies richer trims in Open loop. You can make up the fuel needed to accelerate with added Tip-in, and if you really need to accelerate, you'll be in Open loop anyway. I'm beginning to think that the entire closed loop should be Stoich. I'm going to try this out today. I left the temp comp tables alone, I still think they are necessary.

Are you logging your CL/OL transitions? The ECU holds CL waay too long. With slightly modified transition settings, I've had the ECU hold closed loop into load 1.4-1.5 with boost of 3-4 psi early in the rev range.

[KoolBRZ]

100% Stoich doesn't work. It bogs down after every shift. Went back to tune with Stoich between 1600 and 2800, from.2 to.6 load, and that runs well. @ztan, what minor changes to CL transition did you make?

[Kodename47]

Reducing or setting the open loop delay counters to 0 really seems to help. I've seen plenty of people suggest to set them to zero.

[thambu19]

Originally Posted by KoolBRZ

Remember, it is far far better to make it leaned out in Closed loop, so it applies richer trims in Open loop. You can make up the fuel needed to accelerate with added Tip-in, and if you really need to accelerate, you'll be in Open loop anyway. I'm beginning to think that the entire closed loop should be Stoich. I'm going to try this out today. I left the temp comp tables alone, I still think they are necessary.

Depends on what you are trading for. If it is FE then yes stoich or slightly lean will give you that but doing so you need to add some spark to bring combustion back to where it was set by the OEM.
Going slightly rich (very slightly) in the order of 0.95 lambda will give excellent drive feel but ofcourse at cost of FE.
OEMs tune the FAR to get best catalytic conv efficiency. So the FAR can vary from lean to rich depending on speed/load and this is engine specific so I cant tell you how it works for FA20.
Going CL too long can dip into Scavenging region and it can confuse the O2 sensors causing engine to run rich to compensate for it.

[thambu19]

My car feels flat all the way to 2.2K rpm. I may go back to OEM cam timings below 2K and follow it up with going 100%DI with some additional spark timing. Right not the IAM =1 all the time so I am sure there is some potential for extra spark timing.
I have seen people get scared by seeing some -0.5 deg knock retard. In the world I work half a degree of spark retard is nothing at below WOT conditions. At WOT ofcourse you dont want a lot of spark pull out but upto a degree is okay as it doesnt change the numbers so much.

[KoolBRZ]

I've been playing around with straight DI, and for an AT it shifts better and accelerates better. Knowing it is limited in the higher rpms and loads I made a map of solid 20% PI, then using the thresholds to limit it, I limited PI to 0% up to 2000 rpm, and out to 1.0 load. I made a pic of the table, above, and the effective percentages, below.