IAT ignition comp not working

FlashPro questions & answers specific to the 2006-2011 Americas Civic Si
civicRSA
Posts: 132
Joined: Thu Nov 12, 2009 11:29 am
Location: South Africa
Contact:

Post by civicRSA » Sat Sep 04, 2010 1:20 pm

Hondata wrote:Tested your calibration and the IAT compensation is working. If you still think the compensation is not working, give me the specific test you are using, the expected results and the results that you are getting.
From the above datalogs submitted. I don't see any IAT compensation working. Ignition values at higher IAT's match the MBT values. In 5th and 6th gear there is a 0.5 degree gear based retard.

soxfan143
Posts: 216
Joined: Mon Nov 12, 2007 5:09 pm
Location: lynnfield MA

Post by soxfan143 » Sat Sep 04, 2010 2:55 pm

Here's another log from today. I'm getting absolutely no ignition retard at all from IAT comp table. If you can show me exactly where the IAT is working that would be great. At 18:48 I'm at 104 degrees IAT and there is absolutely no ignition retard happening at all. This is not good for people with turbos. I have like 10 customers waiting for an answer or a solution here and I have no answers for them right now.
Attachments
soxfan FR GT30R V75-6.fpcal
(18.97 KiB) Downloaded 138 times
Still no IAT retard.fpdl
(3.91 MiB) Downloaded 142 times

User avatar
Hondata
Site Admin
Posts: 8783
Joined: Thu Aug 15, 2002 12:13 pm
Location: Torrance, CA
Contact:

Post by Hondata » Sun Sep 05, 2010 6:03 am

At 18:48 I see a table ignition value of 18 degrees, actual ignition of 17 degrees. With the IAT of 98 degrees I would expect to see a retard of about 1.2 degrees. Nothing looks wrong at that point in the datalog.

My point was that you are not using a scientific testing method, and the change you're looking for is much less than the expected accuracy which can be attained from holding the other factors constant. You cannot compare the table advance and final advance and then assume that all the difference is from IAT compensation - there are far too many other compensations / knock adjustments in play. It is not realistic to try to spot 1 degree change in the ignition advance.

What I have done to test the IAT compensation is to vary the IAT with all other conditions being kept as constant as possible, and the IAT compensation does adjust the ignition timing as expected.
Hondata

civicRSA
Posts: 132
Joined: Thu Nov 12, 2009 11:29 am
Location: South Africa
Contact:

Post by civicRSA » Sun Sep 05, 2010 8:19 am

So what we could do to confirm would be to set 4 degrees retard at 86 degrees temp and 4 degrees retard at 125 degrees temp. This way 4 degrees of retard should be seen inbetween these temps? Correct?

The way I see it in Mikes datalog at 18:48 the tables values are interpolating between 18 and 16 degrees and that is why actual values are 17 degrees.

This does not explain why the timing is not actively been retarded at higher temps and knock starts to happen at the higher temps.

civicRSA
Posts: 132
Joined: Thu Nov 12, 2009 11:29 am
Location: South Africa
Contact:

Post by civicRSA » Sun Sep 05, 2010 11:22 am

I did another test by doing the following.
I removed gear based retard. Then I set the IAT comp. as follows. 4 degrees retard at 86F and 4 degrees retard at 131F.

This now had an effect on timing. It looks like the retard was 2 degrees at boost. Why only 2 degrees?

There are a few odd things that I notice with knock tables.
If I open the FG2/FA2 MAF Injen CAI calibration. The knock limit and knock retard tables only have values till 1 bar.
Open the FN2 Group N reflash and there are knock limit and knock retard tables even in boost.
My Calibration has knock limit tables for boost but no knock retard tables for boost.
What is going on here?
Attachments
CivicRSA FR GT3076 FN2 V18-1.fpcal
(17.43 KiB) Downloaded 149 times
Datalog CivicRSA FR v18-1.fpdl
(1.98 MiB) Downloaded 154 times

soxfan143
Posts: 216
Joined: Mon Nov 12, 2007 5:09 pm
Location: lynnfield MA

Post by soxfan143 » Sun Sep 05, 2010 1:18 pm

I figured it out. It's because the knock ignition limit is 2 degrees. You need to get the ignition limit value at 0 degrees and it retards perfectly. In the knock ignition limit tables, high and low, set all the values in boost the proper setting This will make your ignition limit value in boost 0 degrees. Thus the IAT retard will be dead on. I just tested it.

ajpturbopittsburgh
Posts: 145
Joined: Wed May 26, 2010 10:58 am

Post by ajpturbopittsburgh » Tue Sep 07, 2010 9:51 am

Just a quick question then. When I read about the knock tables in the contents section for the 1.1.6.2 it says for knock ignition limit columns 9 and 10 should be 5 degrees or less and in boost should be 0.

I looked at my cals and the one that sox posted here and they are not at all close to that.

Where did all these values come from for this beta version? I just opened my calibration but did not flash my ECU yet. If I change nothing and use my current calibration but with the beta software will everthing be the same? I'm afraid of all the new ignition tables because I am boosted and have shown no knock with 1.1..5.6
Attachments
Track tune with race fuel downloaded and modifided at track.fpcal
(19.86 KiB) Downloaded 136 times

civicRSA
Posts: 132
Joined: Thu Nov 12, 2009 11:29 am
Location: South Africa
Contact:

Post by civicRSA » Tue Sep 07, 2010 10:16 am

Your knock limit tables will need to be adjusted. You cannot open a calibration from 1.1.5.6 in 1.1.6.2 and hope all the settings will be the same. For some reason the way it is explained in the help file is not the way it works.

ajpturbopittsburgh
Posts: 145
Joined: Wed May 26, 2010 10:58 am

Post by ajpturbopittsburgh » Tue Sep 07, 2010 10:25 am

Why would they make all the knock ignition limit tables in boost columns extremely high(i.e. over 20 and 30) when they tell you they should be 0?

Wouldn't that cause you to possibly have knock? Becuase it could potentially cause you to go above the main ignition table values.

If I set all the tables in the knock ignition limit tables to 0 would that assure me that the ignition would never exceed the main ignition tables?

civicRSA
Posts: 132
Joined: Thu Nov 12, 2009 11:29 am
Location: South Africa
Contact:

Post by civicRSA » Tue Sep 07, 2010 11:07 am

In reality you are correct it seems to be working that way. We have actually found that the knock limit in boost needs to be -2 in the tables for the compensation tables to work.

But this is what the help file says and I wish that Hondata would clear this up "The normal ignition tables contain MBT ignition advance values (black line above). An additional table contains the knock ignition limit values (blue line above). It is possible for the knock ignition limit to be lower than the MBT timing. In this case, the ECU will use the lower of the MBT timing and knock ignition limit timing so that the engine will make the most power possible, but will not knock."

That would mean that the ignition timing in boost would be 0 degrees.

ajpturbopittsburgh
Posts: 145
Joined: Wed May 26, 2010 10:58 am

Post by ajpturbopittsburgh » Tue Sep 07, 2010 2:38 pm

civicRSA wrote:Your knock limit tables will need to be adjusted. You cannot open a calibration from 1.1.5.6 in 1.1.6.2 and hope all the settings will be the same. For some reason the way it is explained in the help file is not the way it works.
What should I adjust them to? What is safe or conservative? 0 for all of the knock ignition limit tables? Will I then only run on the main ignition tables?

Wouldn't the default settings that come in the beta version be safe? I don't understand why the knock igntion limit is so high in the boost columns if they tell you in the help section that they should be 0 in boost. I don't like messing with ignition tables since I am not a professional tuner.

soxfan143
Posts: 216
Joined: Mon Nov 12, 2007 5:09 pm
Location: lynnfield MA

Post by soxfan143 » Tue Sep 07, 2010 7:13 pm

Guys, the knock ignition limit tables aren't directly related to the MBT igniton table like you think. The knock ignition limit tables isn't an actual ignition value that effects the ignition value like you think. And to be honest I really don't know how it works exactly. I do know that you have to set the knock limit value in boost so that the actual knock ignition limit value in the sensors window is 0 degrees when in boost for the IAT and ECT comp tables to effect the MBT value properly. When you look at your log open the sensors window and look at the sensor named "knock ignition limit". If that number is more or less than zero it will not let the comp tables effect actual ignition properly. I hope that clears things up for you.

mucter
Posts: 87
Joined: Tue May 16, 2006 11:03 am
Contact:

Post by mucter » Wed Sep 15, 2010 2:53 am

So I'm just going to put this bluntly.

THIS LOGIC SUCKS ASS

Derek, please change the logic here and make all ignition compensations under the "calibration" section independent of the knock logic.

This is nearly impossible to tune consistently the way it works now. k.control is the stupidest thing ever and it messes with the ignition limit and makes it whack out all over the place. We need to be able rely on the IAT ignition compensations.

We can set all of the values to -2 in the k.i.l. tables under boost, the IAT comp will work.... UNTIL you get knock. Once you get knock, k.control shoots up like a rocket because the ignition limit is so low. When it starts to come back down, the ignition limit mysteriously goes from 0 up to whatever the hell the ecu feels like making it. And guess what ?!? The IAT ignition compensation gets offset by whatever the ignition limit has drifted to, thus making it useless.

I don't have room for error with 19psi of boost on the stock engine. There are tons of other boosted people that need to rely on this IAT compensation. This is a MAJOR problem for all of us, a problem that we should have never had.

soxfan143
Posts: 216
Joined: Mon Nov 12, 2007 5:09 pm
Location: lynnfield MA

Post by soxfan143 » Wed Sep 15, 2010 8:24 am

mucter wrote:So I'm just going to put this bluntly.

THIS LOGIC SUCKS ASS

Derek, please change the logic here and make all ignition compensations under the "calibration" section independent of the knock logic.

This is nearly impossible to tune consistently the way it works now. k.control is the stupidest thing ever and it messes with the ignition limit and makes it whack out all over the place. We need to be able rely on the IAT ignition compensations.

We can set all of the values to -2 in the k.i.l. tables under boost, the IAT comp will work.... UNTIL you get knock. Once you get knock, k.control shoots up like a rocket because the ignition limit is so low. When it starts to come back down, the ignition limit mysteriously goes from 0 up to whatever the hell the ecu feels like making it. And guess what ?!? The IAT ignition compensation gets offset by whatever the ignition limit has drifted to, thus making it useless.

I don't have room for error with 19psi of boost on the stock engine. There are tons of other boosted people that need to rely on this IAT compensation. This is a MAJOR problem for all of us, a problem that we should have never had.
AMEN!!!!! I was just gonna post something similar. The k.control gos bananas with the littlest knock count and skews everything to the point where someones gonna lose an engine. Please do something to separate the IAT ignition comp from the knock control logic.

User avatar
Hondata
Site Admin
Posts: 8783
Joined: Thu Aug 15, 2002 12:13 pm
Location: Torrance, CA
Contact:

Post by Hondata » Wed Sep 15, 2010 9:03 am

We do the best we can to improve the behavior of the ECU but you must remember that we did not design the ECU - complain to Honda.

I'll see what I can do.
Hondata

Post Reply