Tuning the transient parameters

[5.7ute]

I have attached a spreadsheet for dual data log addition for the transient tables. Hopefully it can be of some help since there is not a lot of data to use after filtering.

[MICK]

okay so the cos3 and altered transient tables are working well so far, still rich decel, and a few points of slightly lean tip in, but i can also tell a good difference in throttle response.

quick question for anyone though. i asked in the other transient fueling thread but never got an answer. i'm trying to attempt to understand the transient tables better and was wondering about the degress/c in the transient tables. is that referring to IAT or ect? i would think both have a factor on the transient fueling.

Think of it as Engine Temp for now. if it were related to valve temp it would not surprise me.

Cheers

Mick

[Aloicious]

I have attached a spreadsheet for dual data log addition for the transient tables. Hopefully it can be of some help since there is not a lot of data to use after filtering.

hmm, how are you using this? like what PID's are you logging to fill the tables in? I tried logging ECT vs MAP with the BEN's as the data, but it didn't yeild too much usefull info, or maybe I'm going about it wrong. I've gotten pretty good results making educated guesses on where in increase the table, but if there's a better way, I'm all for it.

on another note, while doing some adjustments to the tune last night I did adjust the boiling time constant table to attempt to fix the rich decel, it seemed to help SLIGHTLY but not a whole lot, I may have to look elsewhere to fix it.

on some totally unrelated topics, I solved my hard starting problem using redhardsupra's cranking VE extrapolating spreadsheet (thanks RHS!!!!), and I found out that semi-open loop CAN be disabled using COS 3 without resorting to 1.01 EQ ratios in the commanded fuel tables and altering the stoich value, etc, etc, basically I zero'd out ALL the trimming tables (LTFT and STFT) as well as the switching points table, and maxed out the CL enable temp table. viola, 3 logs, NO trimming (STFT or LTFT, although the LTFT never were a problem), so its a TRUE OLMAF tune I'm running with COS3, and I can use regular EQ values in any table with the correct stoich value and not worry about trimming. (this is good because one of my O2's are bad and adding 50%!!! at times and throwing a code, I really didn't want to buy a $60 sensor that I don't wanna use anyways.)

[hquick]

Can you send me some info as to what tables you did what to, to acieve this?
I'd like to do the same.
As for rich decel.....you should adjust the impact factor table in the same way as for lean tip in. Bump it up in those area's where rich decel occurs.

[Aloicious]

Can you send me some info as to what tables you did what to, to acieve this?
I'd like to do the same.
As for rich decel.....you should adjust the impact factor table in the same way as for lean tip in. Bump it up in those area's where rich decel occurs.

yeah I did bump it up quite a bid with not much luck, I'm wondering if I didn't filter out the transients well enough when I was doing AutoVE and the problem actually lies in the VE table, it may be time for another VE adjustment...on another note, the lean tip in issue is practically eliminated, there is still a tiny bit now and then, but its just leaning out to say ~15.7 around 15% of the time or so, instead of ~17+ at 50% of the time. throttle response has improved like you mentioned as well.

yeah I can send you the tables to disable for a true OL in COS3, I'm at school now though so I don't have my laptop with me, I'll let you know exactally which ones when I get home tonight. I also noticed on one of your recent tunes that you have the stock cranking VE table, does it give you a hard time starting (especially when warm/hot)? I was having to give it gas every time I was trying to start it with the stock cranking VE. if you are having issues you should go to RHS's website and get his cranking VE tool, it extrapolates what the cranking VE should be (instead of ~1.8 across the board like stock) from the regular VE, and 5-10 starts so far after doing that I've never had to give it any gas to help it start.

[hquick]

Yes...it does have trouble starting on a warm re-start. I have to feather the throttle until it settles at idle.
Thanks for that....I'll look into it.

[joecar]

...

on some totally unrelated topics, I solved my hard starting problem using redhardsupra's cranking VE extrapolating spreadsheet (thanks RHS!!!!), and I found out that semi-open loop CAN be disabled using COS 3 without resorting to 1.01 EQ ratios in the commanded fuel tables and altering the stoich value, etc, etc, basically I zero'd out ALL the trimming tables (LTFT and STFT) as well as the switching points table, and maxed out the CL enable temp table. viola, 3 logs, NO trimming (STFT or LTFT, although the LTFT never were a problem), so its a TRUE OLMAF tune I'm running with COS3, and I can use regular EQ values in any table with the correct stoich value and not worry about trimming. (this is good because one of my O2's are bad and adding 50%!!! at times and throwing a code, I really didn't want to buy a $60 sensor that I don't wanna use anyways.)

I'm also interested in which tables...

[5.7ute]

hmm, how are you using this? like what PID's are you logging to fill the tables in? I tried logging ECT vs MAP with the BEN's as the data, but it didn't yeild too much usefull info, or maybe I'm going about it wrong. I've gotten pretty good results making educated guesses on where in increase the table, but if there's a better way, I'm all for it.

At the moment just use AFR,s as the data to give a visual on which cells the lean tip in or tip out is occuring. The main problem is we are talking about fractions of the pulse width being added & BEN factor is not the correct adjustment.
The main reason I had made the sheet is that even on a long log with lots of throttle action there was only a tiny amount of data left for the lean tip in. With some good data we should be able to work out a custom pid that will take out the guesswork & bring it all into line fast enough for roadrunner use.

[Aloicious]

At the moment just use AFR,s as the data to give a visual on which cells the lean tip in or tip out is occuring. The main problem is we are talking about fractions of the pulse width being added & BEN factor is not the correct adjustment.
The main reason I had made the sheet is that even on a long log with lots of throttle action there was only a tiny amount of data left for the lean tip in. With some good data we should be able to work out a custom pid that will take out the guesswork & bring it all into line fast enough for roadrunner use.

That makes sense with the AFR. your spreadsheet should help out alot. I'll give it a try, how are you filtering it? when I tried filtering it I couldn't get a good enough 'just transient' filter.

[Aloicious]

Okay here's the tables I altered to get a true OL tune in COS3. I probably went a bit far by disabling EVERYTHING I could find on trimming, but hey, it works.... basically every table is completely zero'd out, except for B3802 which is maxed to 140*C. I'm going to be doing some more logging the next few days I'll try to post one up with a tune if anyone wants to see 'em

{B3802} "Minimum Coolant Temp to enable LTFT" - Maxed out to 140
{B3803} "Maximum Coolant Temp to Enable LTFT" - Zero'd
{B3805} "Minimum LTFT Adjustment" -Zero'd
{B3806} "Maximum LTFT Adjustment" -Zero'd
{B3807} "Maximum TPS for Idle LTFT" -Zero'd
{B3808} "Maximum speed for Idle LTFT" -Zero'd
{B4115} "STFT Maximum Correction" - Zero'd
{B4116} "STFT Minimum Correction" - Zero'd
{B4117} "STFT Multiplier" - Zero'd
{B4118} "STFT Adjustment at Idle" - Zero'd
{B4119} "STFT Adjustment at Non-Idle" - Zero'd
{B4109} "STFT Base Correction at Idle" - Zero'd
{B4110} "STFT Multiplier at Idle" - Zero'd
{B4111} "STFT Base Correction at Non-Idle" -Zero'd
{B4112} "STFT Multiplier at Non-Idle" - Zero'd
{B4113} "STFT Base Delay" - Zero'd
{B4114} "STFT Delay Multiplier" -Zero'd
{B4105} "O2 Switch point" - Zero'd
{B4107} "Closed Loop Mode" - Zero'd