PDA

View Full Version : RTACS using STFT



EagleMark
January 24th, 2013, 01:20 PM
The RTACS is working!

Trying to set up the RoadRunner RTACS using the STFT outlined here:
http://forum.efilive.com/showthread.php?6318-A-way-to-cheat-and-fix-your-strims-ltrims-using-Rtac-Very-quickly

As far as engine displacment I am using g*K/kpa in the {B0101} Main VE, but lso have engine displacment set under log file information.

A. Set Displacement: open a Log File, go Edit->Log File Information->Vehicle Options, and next to Engine enter the displacement in cc, click Ok, and save the log file.
This matters only if you're using [%] VE units; if you're using [g*K/kPa] VE units then you don't need displacement.



Calc pids.txt added are, and I think the issue is my calculation?

*CLC-00-001
% -20.0 20.0 .2 "({SAE.SHRTFT1}+{SAE.SHRTFT2})/2"

*CLC-00-002
factor 0.0 2.0 .1 "{CALC.STFT_AVG}/100+1"


SAE-generic.txt untouched.

Cal_links text has this one change to first entry after RRR=

;Main VE Table
B0101.ROW=SAE.RPM
B0101.COL=SAE.MAP
B0101.RRR=CALC.STFT_AVG,CALC.BEN1,CALC.BEN2,


Updates are happening but they don't stop, just keeps adding till it hits the limit? What do I have wrong?

joecar
January 24th, 2013, 02:24 PM
Do you reference CLC-00-002 in the PRN section of the calc_pids.txt file...?

Where is CALC.STFT_AVG defined (in the PRN section, what CLC does it reference)...?

If CALC.STFT_AVG references CLC-00-002, then CLC-00-002 cannot reference CALC.STFT_AVG (recursively circular reference).

joecar
January 24th, 2013, 02:26 PM
You probably want CLC-00-002 to compute either one of these:

"{SAE.SHRTFT1}/100+1"

"{SAE.SHRTFT1}+{SAE.SHRTFT2}/200+1"


edit: correction added.

joecar
January 24th, 2013, 02:27 PM
Post you calc_pids.txt here.

EagleMark
January 24th, 2013, 02:37 PM
I am trying to get average of both STFT as outlined in this post:
http://forum.efilive.com/showthread.php?6318-A-way-to-cheat-and-fix-your-strims-ltrims-using-Rtac-Very-quickly&p=56607&viewfull=1#post56607

EagleMark
January 24th, 2013, 02:48 PM
You probably want CLC-00-002 to compute either one of these:

"{SAE.SHRTFT1}/100+1"

"{SAE.SHRTFT1}+SAE.SHRTFT2}/200+1"If I add a "{" in front of SAE.SHRTFT2 it seems to be working! EDIT: Not...

Do I even need *CLC-00-001 ?

joecar
January 24th, 2013, 03:08 PM
If I add a "{" in front of SAE.SHRTFT2 it seems to be working!

Do I even need *CLC-00-001 ?Ah, thanks, yes I forgot the {.

CLC-00-001 is not needed.

joecar
January 24th, 2013, 03:12 PM
Actually, CLC-00-002 should read like this: " ({SAE.SHRTFT1}+{SAE.SHRTFT2})/200+1"


see attached.

joecar
January 24th, 2013, 03:16 PM
...

Do I even need *CLC-00-001 ?


Ah, thanks, yes I forgot the {.

CLC-00-001 is not needed.CLC-00-001 and F001 can be deleted.

EagleMark
January 24th, 2013, 03:19 PM
Without the { it has an X through PID when connected?

STFT % VE Fixer {CALC.STFT_AVG}

Expression:
factor = ({SAE.SHRTFT1}+SAE.SHRTFT2})/200+1
is NOT valid because:
Unknown parameter {SAE}, at position 19.

PID value cannot be determined because
the following error would occur:
Expression not valid: Unknown parameter {SAE}, at position 19.

joecar
January 24th, 2013, 03:36 PM
Yes, the missing { confuses it.

EagleMark
January 24th, 2013, 04:00 PM
But when you add it it works SWEET! :good:

EagleMark
January 24th, 2013, 05:27 PM
Looks like I hit the 2.4145 limits set from EFI Live in a couple cells. What are limits to VE table in g*K/kpa ? I know what limits are in VE%...

joecar
January 25th, 2013, 05:41 AM
Look at B0101... in the upper RHS of the window it shows the min/max allowed cell values.

EagleMark
January 25th, 2013, 05:59 AM
I saw that for g*K/kpa is 10.

But was not sure if that was accurate or possible because if I use % the max is 500%... pretty sure 100% is max?

EagleMark
January 25th, 2013, 03:58 PM
The STFT worked well. Disabled the LTFT as discribed, also EGR, PE, Lean Cruise, DFCO and it was a pretty clean VE table when done.

Thought I'd try another way using part of the New Twist to Calc VE. Went back and enabled LTFT and used CALC.SELBEN, now that works well!

joecar
January 25th, 2013, 05:23 PM
I saw that for g*K/kpa is 10.

But was not sure if that was accurate or possible because if I use % the max is 500%... pretty sure 100% is max?

You can have more than 100% VE... this is what long tube headers can do with NA (I've seen 109% at one of the resonant rpm's)...

if you add boost then as you correct the VE table you will definitely see above 100%.


Take note:
- the tunetool uses cylinder volume to view the VE table when using [%] units;
- the tunetool ignores cylinder volume when using [g*K/kPa] units;

it's up to you... but I would rather use [g*K/kPa].

joecar
January 25th, 2013, 05:30 PM
The STFT worked well. Disabled the LTFT as discribed, also EGR, PE, Lean Cruise, DFCO and it was a pretty clean VE table when done.

Thought I'd try another way using part of the New Twist to Calc VE. Went back and enabled LTFT and used CALC.SELBEN, now that works well!Good job :cheers: post pics of your resulting MAF and VE tables.

Take note:
- the VE kicks in during transient airflow... but the transient filter you apply to the map excludes frames where VE kicked in, so you're left with MAF only correction data;
- the VE is calculated from the corrected MAF (from the same log);

how does vehicle drive now, how does throttle respond (be objective)...?

the procedure is fairly easy, wouldn't you say...?

and you even got it going for Vortec (which lacks various definitions and pids), good job well done :cheers: :cheers:


Thanks for posting back, I really appreciate the feedback.

EagleMark
January 25th, 2013, 07:56 PM
You can have more than 100% VE... this is what long tube headers can do with NA (I've seen 109% at one of the resonant rpm's)...

if you add boost then as you correct the VE table you will definitely see above 100%.


Take note:
- the tunetool uses cylinder volume to view the VE table when using [%] units;
- the tunetool ignores cylinder volume when using [g*K/kPa] units;

it's up to you... but I would rather use [g*K/kPa].I am using g*K/kpa, you don't think Im ignoring your advice do you? AFR is also gone, Lambda for wideband and EQ for data. It takes a little while to brake OBDI tuning habits. Like VE can never go over 100%, 90 to 95% injectors go static! Differant worlds in MAF...

EagleMark
January 25th, 2013, 08:14 PM
Good job :cheers: post pics of your resulting MAF and VE tables.

Take note:
- the VE kicks in during transient airflow... but the transient filter you apply to the map excludes frames where VE kicked in, so you're left with MAF only correction data;
- the VE is calculated from the corrected MAF (from the same log);

how does vehicle drive now, how does throttle respond (be objective)...?

the procedure is fairly easy, wouldn't you say...?

and you even got it going for Vortec (which lacks various definitions and pids), good job well done :cheers: :cheers:


Thanks for posting back, I really appreciate the feedback.I'll post more as I said I would, it's the least I can do for all the help you have given. Maybe you haven't noticed all I share back elsewhere... Would have been a lot easier to start without the Vortec and then the wideband issue? :doh2:

Yes it's easy once you catch up to procedures!

But it's all coming together now.


how does vehicle drive now, how does throttle respond (be objective)...? Crisp! :good:

joecar
January 26th, 2013, 07:34 PM
I am using g*K/kpa, you don't think Im ignoring your advice do you? AFR is also gone, Lambda for wideband and EQ for data. It takes a little while to brake OBDI tuning habits. Like VE can never go over 100%, 90 to 95% injectors go static! Differant worlds in MAF...Cool :cheers:


I'll post more as I said I would, it's the least I can do for all the help you have given. Maybe you haven't noticed all I share back elsewhere... Would have been a lot easier to start without the Vortec and then the wideband issue? :doh2:

Yes it's easy once you catch up to procedures!

But it's all coming together now.

Crisp! :good:Cool :cheers:

I like crisp.

EagleMark
January 28th, 2013, 10:40 AM
I got something wrong? Closed loop seems to be working fine, but when I go PE... well the numbers aren't good?

Looking at calc pids, tutorials, pdf etc... it seems it stated with a calc to change from LTFT to WO2BEN
"iff({CALC.PE},{CALC.WO2BEN1}, {CALC.LTFTBEN})"
I understand CALC.PE was changed to CALC.CL but what about 1 space before {CALC.LTFTBEN} ?

Then the new version
"iff({CALC.CL}, {CALC.LTFTBEN}, {CALC.WO2BEN})"
Now we have 2 spaces? Do the spaces matter? Nothing else has spaces? And iff {CALC.CL),{CALC.LTFTBEN} wouldn't it need an else or? {CALC.WO2BEN}?

Or am I missing something else?

joecar
January 28th, 2013, 02:25 PM
Check that there is no mention of CALC.PE.
Check that CACL.CL is defined.

The space(s) should make no difference in that context (spaces can enhance human readability).

The iff(x,y,z) function performs this:

if (x) then y else z

where x, y, z are evaluated at the time that the iff() is executed;

( the else/or is not required because the parser can determine the clause by it's position wrt to the commas).

joecar
January 28th, 2013, 02:26 PM
What do the PE numbers look like...?

Post pics showing what is going wrong.

EagleMark
January 28th, 2013, 04:43 PM
Check that there is no mention of CALC.PE. Not on my end, just old tutorials, one specifaclly has CALC.PE mentioned in calc_pid.txt file yet calls for CALC.CL PID to be selected.


Check that CACL.CL is defined. Check, always has been. "{GM.EQIVRATIO}=1" and GM.EQIVRATIO PID selsected.

The problem was when in PE the adjustment made by files and RoadRunner went through the roof! Was earlier today, logs are gone, no pictures.

Problem solved.

{B4206} "Use O2 with Open Loop Commanded fuel table" in this calibration is Enabled. It is also Enabled in all truck cals I checked. It is Disabled in all Camaro and Corvette cals I checked. I disabled it and all is well.

Log with {B4206} Disabled shows no STFT in PE areas. Works good!

Log with {B4206} Enabled shows STFT of up to 10% in PE areas. Works bad!

I only get a couple hours a day to dial this in and stuff like this interferes in my sleep!

Tonight I can sleep...



Starting VE table and the reason engines fall on there face before a shift point.
http://forum.efilive.com/attachment.php?attachmentid=14437&d=1359433837

After first round and some smoothing. 5.7L Vortec engine so RPMs are not going to match LS engines. But it does pull hard to shift points
http://forum.efilive.com/attachment.php?attachmentid=14438&d=1359433884

Almost done. Not much fuel LTFT left, PE AFR/Lamda steady.
http://forum.efilive.com/attachment.php?attachmentid=14439&d=1359433885

joecar
January 28th, 2013, 04:56 PM
In OL B4206 will trim any cell in B3605 that is set to stoich... this means you cannot use STFT to trim away from stoich (PE is commanding away from stoich).


Do you have CL/LTFT disabled...?

EagleMark
January 29th, 2013, 02:03 AM
Do you have CL/LTFT disabled...?EDIT: No, that would not work with
"iff({CALC.CL}, {CALC.LTFTBEN}, {CALC.WO2BEN})"

But since the thread started as STFT I can see why you ask. LTFT is only turned off for
"({SAE.SHRTFT1}+{SAE.SHRTFT2})/200+1"
Which is not implemented in what I am doing now. That works well for Non PE and a vehicle without a wide band. I started there because it was the only idea I got from reading the RoadRunner forum.



In OL B4206 will trim any cell in B3605 that is set to stoich... this means you cannot use STFT to trim away from stoich (PE is commanding away from stoich).
Correct, but since all truck cals are set {B4206} to Enable you get STFT with PE. If not you spend hours over days wondering why CL LTFT is working well and OL PE sucks...

Since there are way more trucks and conversions from trucks with 0411 PCM then there are Camaro or Corvette (which don't have this issue, it's disabled)? This should really be in summary/logging notes with turn off EGR, Lean Cruise and DFCO etc...for tuning only. {B4206} Disable during tuning and left off permanatly.

EagleMark
January 29th, 2013, 10:56 AM
Question? What does the cal_setup.txt file do?

I had made a mistake in cal_link.txt, copied CALC.CALC.SELBEN and of course it would not work! RTACS showed CALC.WO2BEN which was next in line. So I fixed it but still did not work. I remember in another thread, long ago someone posted they put cal_link.txt in
C:\Program Files (x86)\EFILive\V7.5\Configuration which is correct.
But had also put it in
C:\Users\Mark\Documents\EFILive\V7.5\User Configuration which is not correct.

But I found cal_setup.txt in
C:\Users\Mark\Documents\EFILive\V7.5\User Configuration
and looked... it had this

[LS1]
B0101=.,4,.,.,.,.
I removed it and left the file blank and Roadrunner RTACS started working again properly.

Everytime I change cal_links.txt or calc_pid.txt I restart scantool and tunetool which I beleive is correct procedure?

joecar
January 29th, 2013, 11:51 AM
Yes, scantool reads those files when it is started.

joecar
January 29th, 2013, 11:53 AM
EDIT: No, that would not work with
"iff({CALC.CL}, {CALC.LTFTBEN}, {CALC.WO2BEN})"

But since the thread started as STFT I can see why you ask. LTFT is only turned off for
"({SAE.SHRTFT1}+{SAE.SHRTFT2})/200+1"
Which is not implemented in what I am doing now. That works well for Non PE and a vehicle without a wide band. I started there because it was the only idea I got from reading the RoadRunner forum.


Correct, but since all truck cals are set {B4206} to Enable you get STFT with PE. If not you spend hours over days wondering why CL LTFT is working well and OL PE sucks...

Since there are way more trucks and conversions from trucks with 0411 PCM then there are Camaro or Corvette (which don't have this issue, it's disabled)? This should really be in summary/logging notes with turn off EGR, Lean Cruise and DFCO etc...for tuning only. {B4206} Disable during tuning and left off permanatly.I'm getting confused (happens easily)...

when STFT's are enabled (in OL) are you saying that:
- RTACS diverges...?
- Calc.VET diverges...?

I didn't consume sufficient caffeine nor did I get enough sleep... :)

EagleMark
January 29th, 2013, 12:52 PM
I'm getting confused (happens easily)...

when STFT's are enabled (in OL) are you saying that:
- RTACS diverges...?
- Calc.VET diverges...?

I didn't consume sufficient caffeine nor did I get enough sleep... :)I have the same lack of sleep issues... but to much caffine! :anitoof:

I diverge! Hell I don't know Joe, it's not like there's instructions. This is what I'm trying to accomplish. Instructions.

When I do this on old ECM/PCM PE goes OL has no STFT/INT and no LTFT/BLM. Just Desired AFR and adds PE (simplified version)

My understanding of PE is it is Open Loop. Commanded AFR or in this case EQR and EQR added from PE table.

When I Disabled {B4206} my STFT at WOT/PE are now 0

Now I look today at log and I have LTFT at WOT/PE? Why?

Maybe a better way would be if you get some sleep and then have lots of coffee! :cool:
Then give me some ideas for implementing this for VE table only? Give me some basics, make me rethink! Start from scratch.

:wallbash:

Which first? MAF or VE? How? Then would running RTACS on both at same time be wise?

joecar
January 29th, 2013, 03:39 PM
lol, how much coffee can you drink before it puts you to sleep :cucumber::banana:


PE can be OL or CL...

in OL: CFOL is already enabled, PCM enables PE (so the active tables are PE B3618 and CFOL B3605), selects the richest of the active tables;

in CL: CFOL is not enabled, PCM enables PE, (PE B3618 is the only active table), stops trimming, rounds any negative trim to zero, applies any positive trim on top of PE;


was your PE LTFT positive...?


Sometimes you have to go in reverse... try doing the MAF first (but make sure you exclude any VE contribution)...

I don't think RTACS (or BEN corrections in general) on both at the same time would work because there is no indicator showing which (VE or MAF) the airmass is contributed from.

joecar
January 29th, 2013, 03:41 PM
So while your trims (LTFT) are positive you have to be careful...

EagleMark
January 29th, 2013, 03:53 PM
lol, how much coffee can you drink before it puts you to sleep :cucumber::banana:Does that work? I haven't sleep more then 4 hours a day in almost three years now...


So while your trims (LTFT) are positive you have to be careful...This is it! +6 to +8 LTFT...

I'll get my 3-4 hours sleep and re-read to retune.

:good:

joecar
January 29th, 2013, 04:20 PM
Does that work? I haven't sleep more then 4 hours a day in almost three years now...

...Sometimes... especially when you're determined to stay awake... and then suddenly Zzzzz...

SW1911
August 30th, 2014, 10:31 AM
where are these lines inserted ?

joecar
August 31st, 2014, 06:39 AM
where are these lines inserted ?Add then right after the .ROW and .COL for the tables...

e.g.


A0007.ROW=SAE.RPM
A0007.COL=SAE.TP
A0007.RRR=CALC.WO2BEN,CALC.STFTBEN


A0009.ROW=SAE.RPM
A0009.COL=SAE.MAP
A0009.RRR=CALC.WO2BEN,CALC.STFTBEN