PI : DI Ratio Discussion

[Kodename47]

And then you lose the reason the PIs are used.... to help keep the valves clean and to help get a better mixture. It's a nice feature that ECUtek have got which means you can log the end of DI to spark timing, but only really is useful for boosted applications when you might have enough fuel to get close to the spark timing.

[KoolBRZ]

Originally Posted by Kodename47

And then you lose the reason the PIs are used.... to help keep the valves clean and to help get a better mixture. It's a nice feature that ECUtek have got which means you can log the end of DI to spark timing, but only really is useful for boosted applications when you might have enough fuel to get close to the spark timing.

I just logged the results of the tunes with those tables and PI is most definitely not staying at 20%. Per the threshold settings it was supposed to be at 0% up to 2000 rpms and out to 1.0 load, and on the other tune it was with oem threshold settings, so it should have been 0% up to 5000 rpms and out to 1.5 load. At certain loads and speeds under 1.0 and under 5000 rpms the PI would actually exceed the DI, even though the whole table was 20%. Then I tried just zeroing out the area affected by the threshold and it made absolutely no difference. I zeroed out the whole table and finally PI shut off. There has to some other coding at work here, because it didn't happen at all times or all rpms. I'll load up the table I partially zeroed out and upload a log for you guys to look at. then maybe you can tell me what the *&^%$ is going on with this crazy PI system.

[KoolBRZ]

This should be 20% PI. It not only isn't, it's all over the place. There is obviously something else affecting PI. I don't think it's just being used for tip-in. I think there are more thresholds or something.
http://datazap.me/u/koolbrz/log-1458...g=0&data=12-13
http://datazap.me/u/koolbrz/log-1458...0&data=1-12-13

[Kodename47]

Originally Posted by KoolBRZ

I just logged the results of the tunes with those tables and PI is most definitely not staying at 20%. Per the threshold settings it was supposed to be at 0% up to 2000 rpms and out to 1.0 load, and on the other tune it was with oem threshold settings, so it should have been 0% up to 5000 rpms and out to 1.5 load. At certain loads and speeds under 1.0 and under 5000 rpms the PI would actually exceed the DI, even though the whole table was 20%. Then I tried just zeroing out the area affected by the threshold and it made absolutely no difference. I zeroed out the whole table and finally PI shut off. There has to some other coding at work here, because it didn't happen at all times or all rpms. I'll load up the table I partially zeroed out and upload a log for you guys to look at. then maybe you can tell me what the *&^%$ is going on with this crazy PI system.

Did you change all 3 ratio tables?

You're logging injector open time for PI and DI, just because the PI IPW is longer than the DI does not mean that it's providing more fuel.

There are min/max DI and PI IPW tables, are these still untouched?

[KoolBRZ]

Originally Posted by Kodename47

Did you change all 3 ratio tables?

You're logging injector open time for PI and DI, just because the PI IPW is longer than the DI does not mean that it's providing more fuel.

There are min/max DI and PI IPW tables, are these still untouched?

I left the min/max tables untouched. I considered the PW size difference, but regardless, PI should have followed DI, not jumped all over the place. Try making all three tables all zeroes except for a line of 20% down the right side and across the bottom. Drive and log that. Tell me if it does it to your car as well. When I zeroed out the whole tables it turned off PI completely, except for idle. At idle it came on some of the time.
(edit) Could the PI pulse width have been magnified by the manifold pressure? Since I have the Phantom SC with the Procede controller it was on and boosting when the PI was jumping around. Could the PI have been using more PW to compensate for pressure in the manifold?

[Kodename47]

I have run with 0% PI and it works as it should. I would suggest something not being set right. If the PI ratio table is set to 0 then no PI will be used unless under certain idle or you exceed the max DI fuel volume parameters.

[KoolBRZ]

Originally Posted by Kodename47

I have run with 0% PI and it works as it should. I would suggest something not being set right. If the PI ratio table is set to 0 then no PI will be used unless under certain idle or you exceed the max DI fuel volume parameters.

0% PI tables work fine showing Pi only at idle, sometimes, 100% DI cures every single problem I have ever had with fueling. Problems I have tried to fix with AVCS tuning. I hate PI now. It's louder, less economical, and accelerates worse than DI. I'm trying to work with AVCS timing now to get the Intake Late/Exhaust Late internal EGR working correctly. I'm not sure if I can do it with DI alone, or if I need PI to get it to work correctly. I'm just concentrating on the area from .3-.5 load, and from 1200-2000 rpms. That should boost my highway mileage.

[Kodename47]

1st you thought PI was beneficial, now you want to run 100% DI.....

There's a reason the OEM engineers chose what they did. You could run more DI if you like at lower load, but I wouldn't eliminate the PI completely for cruising or you'll end up with all the horrible deposits that the PI is there to help with. And as stated, the PI will help mixture formulation in the cylinder under some circumstances.

EGR is not necessarily beneficial to efficiency, it is more to do with emissions, but that's for a different thread. However, you don't really want the DIs to be injecting when there's hot exhaust gasses in the cylinder IMO.

[shr133]

Originally Posted by KoolBRZ

0% PI tables work fine showing Pi only at idle, sometimes, 100% DI cures every single problem I have ever had with fueling. Problems I have tried to fix with AVCS tuning. I hate PI now. It's louder, less economical, and accelerates worse than DI. I'm trying to work with AVCS timing now to get the Intake Late/Exhaust Late internal EGR working correctly. I'm not sure if I can do it with DI alone, or if I need PI to get it to work correctly. I'm just concentrating on the area from .3-.5 load, and from 1200-2000 rpms. That should boost my highway mileage.

I wouldn't run any egr timing...... but to do so just bump up the exhaust timing...

For MPG at lower rpm go to about 10-12 on the intake and 8-10 on the exhaust... That should still produce good cruising power and good MPG....

Changing the DI / PI percentage will also change your F/A ratio, I think you van run more fuel with port injection...

[freerunner]

Originally Posted by shr133

Changing the DI / PI percentage will also change your F/A ratio, I think you van run more fuel with port injection...

I have made some adjustments to 'Port Injector Scaling BRZ' until changing DI/PI ratios did not affect AFR anymore.

Had to decrease the value slightly for my car.

[KoolBRZ]

I've been comparing DI to PI with DI and found that DI alone doesn't have as much power or acceleration. That is over-simplification, since there are firing times and angles to consider with DI as well, but I happened on a good thing because of it. From working with it in the past, I know that straight DI can handle more spark advance than PI, so I wanted to use the majority of the PI in low load areas. After several phenomenally bland and boring failures, I tried out this table, changing the PI Ratio Threshold to 15% to enable 20% PI. I was surprised to find I liked it. Knowing most people don't have access to the PI Ratio Thresholds, I tried the same design table, with the 20% ranges boosted to 35%, so the normally hidden thresholds wouldn't apply. I didn't like it as much, but it wasn't too bad. Take a look, try it out, and tell me what you think.

[pro_turbo]

Originally Posted by KoolBRZ

I tried the table shown, 23, and a new one, 24, which turned out to be better. From the stock thresholds and my experiences, I conclude that anything less then 35% is worse than nothing, and anything more than 75% might as well be everything. I'll keep on trying out different tables and posting them here, all by myself. (sound of crickets chirping)

This was a very look like a base map scale from motec

[callisto]

For those who are still studying 86 VVT, here's an update from the distant future of gen1.5 red intake ROMs: the intake tables are 3 rows taller, having been tuned specifically to reduce intake advance at load 0.0-1.0 x rpm 4700-5200. This is right about where we all hear that tinny VVT resonance on K-series tunes, and it doesn't silence that resonance - but it does limit how loud it gets compared to K tunes, and I figure if they went through the effort of resizing the table to put this in, there must be something worth considering in that. So, duly noted.


EDIT: Also! Some injection-specific changes in S10C/U01A that y'all may be interested in (the S10C XML and BIN elsewhere in the forum can be used to see these in RomRaider):



One value for Direct Injection Quantity Maximum has changed: RPM 2000, from 90 => 120 (increased by 30mL).


The GDI Pressure A/B tables have had nearly every value tweaked.


The Intake/Exhaust Duty Correction A/B/C/D tables have had most values tweaked.


The GDI Flow Rate (correction) table has had most values tweaked, with many cells being set to 1.000 (no correction).

[callisto]

Alongside the VVT in/ex table above, here's the ZA1JU01A port injection ratio table. It's been expanded since K-series as well, to make room for the notched series on the right, but you can compress this back down to K-series to fit without losing any fidelity by renumbering the RPM axis to (600, 1600, 2000, 2400, 2800, 3200, 4800, 5200, 6400, 6800) and copying only those rows from U01A below (or S10C, which has a romraider XML elsewhere on the forum).

This is also accompanied by a change to the PI ratio values for RPM and Load from K-series, so make sure to consider those: PI Ratio Thresholds RPM = 3200; PI Ratio Thresholds Load = 0.6. (The table PI Ratio Thresholds is unchanged.)