Traction Control + Knock Control Strategy

Post Reply
jun1or
Posts: 137
Joined: Tue Apr 14, 2009 10:09 am

Traction Control + Knock Control Strategy

Post by jun1or » Tue Jul 30, 2019 9:52 pm

Hello Doug and Hondata Team,

Long time FlashPro user and self-tuner. Curious about recommendations on TC + OEM Knock Control strategies.
Just finished up a CD009 swap and am finally able to test the limits of Hondata Traction control. I have 3 comments and 3 related questions:

Cliff notes:
  • 70% HPDE-Road Course Car / 30% Street Car / 0% Drag Strip Car
  • 640whp/480wtq e85/ford 8.8/CD009

Comments:
  • 1.) I am noticing high amounts of knock counts when enabling TC (15-20 with TC on, 0-3 off). Given how TC works via TC.R (Traction Control ignition retard), I assume this is normal behavior.
  • 2.) I currently have misfire detection enabled. I have played with this on and off, and usually kept it off for HPDE/Time Attack but enabled it during recent street driving & testing. I am getting CEL for random misfires. (likely due to k.retard)
  • 3.) Timing - I noticed medium/high amounts of timing retard and high "Knock Control" values when enabling Traction Control.

Questions related to above comments: (yes by now you know the questions)
  • 1.) Is there a better knock control strategy (or lack thereof) to be used if reliant upon Traction Control. In other words, shall I severely limit/disable the OEM knock strategy when using Traction Control? I run e85 and see <3 k.count when TC is disabled and just saw 16 on a 20m street drive with TC enabled. I also saw a K.retard value of 6 degrees tonight after traction control enabled, which is new / (caveat of TC?)
  • 2.) Can I safely disable misfire detection when using Traction Control?
  • 3.) I would love to limit the OEM knock control strategy when using TC, but still enjoy using the OEM knock detection failsafes: I.E. (K.Count[1-4], K.Level, K.Control). Is there a happy medium? Is the best solution to reduce the Knock Retard High table and leave the Knock Sensitivity High table alone? Any other recommendations are welcomed. I am open to testing and sharing data to better the TC help file too.

Files:
  • Datalog #1 attached - Street drive before enabling and testing Traction Control [K.Count 0, Ignition 11-18 - perfect pull]
  • (no logs - tested TC for 30-45 minutes)
  • Datalog #2 attached - Street drive after testing Traction Control for 30+ minutes [K.Count 16, Ignition 7-12 - breakup/CEL/High K.Retard/K.Control]
Both the datalogs are with TC disabled, but notice the 2nd datalog induced a CEL due to high K.Count, K.Control, and related K.Retard)

Calibration also attached, for reference.

Thanks,

Mike (JR) Riley
Attachments
S2000 (e85) v117 23lbs pte5865 ID1700's see notes (CD009).fpcal
(11.54 KiB) Downloaded 32 times
2019-07-30 - Hondata TC disabled testing1.fpdl
Pre-TC testing (TC disabled)
(7.07 KiB) Downloaded 29 times
2019-07-30 - Hondata TC enabled testing2.fpdl
Post-TC testing (TC disabled)
(34.82 KiB) Downloaded 29 times

User avatar
Hondata
Site Admin
Posts: 8750
Joined: Thu Aug 15, 2002 12:13 pm
Location: Torrance, CA
Contact:

Re: Traction Control + Knock Control Strategy

Post by Hondata » Wed Jul 31, 2019 9:12 am

1. Nope, not normal. In the two datalogs I see the knock control is vastly different (55% vs 100%), and that will give different running and knock counts.
2. Disable the misfire detection. There is no downside.
3. The ECU will look at engine load and knock noise, and then adjust the knock control accordingly. So if the load is high and the noise low knock control will come down.

For knock control I would do a test where you reflash with the TC enabled and disabled. See what knock control is immediately after a reflash. It should be 55%. If that is ok then drive and see where knock control increases. With E85 knock control should stay at 55%.

I can't tell because the datalog is short but I think that you're getting incomplete combustion from too much retard (unrelated to TC) and that is increasing the knock control, which makes it worse. A longer datalog would confirm this. Also on E85 you might need to run leaner.
Hondata

jun1or
Posts: 137
Joined: Tue Apr 14, 2009 10:09 am

Re: Traction Control + Knock Control Strategy

Post by jun1or » Wed Jul 31, 2019 12:22 pm

Hondata wrote:
Wed Jul 31, 2019 9:12 am
1. Nope, not normal. In the two datalogs I see the knock control is vastly different (55% vs 100%), and that will give different running and knock counts.
2. Disable the misfire detection. There is no downside.
3. The ECU will look at engine load and knock noise, and then adjust the knock control accordingly. So if the load is high and the noise low knock control will come down.

For knock control I would do a test where you reflash with the TC enabled and disabled. See what knock control is immediately after a reflash. It should be 55%. If that is ok then drive and see where knock control increases. With E85 knock control should stay at 55%.

I can't tell because the datalog is short but I think that you're getting incomplete combustion from too much retard (unrelated to TC) and that is increasing the knock control, which makes it worse. A longer datalog would confirm this. Also on E85 you might need to run leaner.
Thank you for the prompt reply and suggestions.

You are correct, the two datalogs show vastly different k.control values, and I wanted to seemingly blame TC and TC.Retard for it, which may not be accurate. I did notice k.count2 spiking to 15 counts in low rpm low speed parking lot driving. See the end of this datalog where it starts to (phantom?) knock. (See 16:22 to end of file, this k.count reaches 15 eventually when the initial datalog2 in post 1 was taken!) I do have some NVH (Noise, Vibration, Harshness) from the CD009 and twin disc right now, so perhaps this is all phantom knock? But why would it only register on cylinder 2?

I think you're on to something. I think this has nothing to do with traction control has everything to do with OEM knock control strategy. I checked several other datalogs and see many k.count on cylinder 2, but mostly low speed, low load, which is odd.

As far as e85 and your comment about going leaner. I'm at 12.0:1 currently! I had better luck at 11.9 target to be honest.

Thanks again,

Mike (JR) Riley
Attachments
2019-07-30 drive to golf BAP disabled.fpdl
(3.74 MiB) Downloaded 30 times

jun1or
Posts: 137
Joined: Tue Apr 14, 2009 10:09 am

Re: Traction Control + Knock Control Strategy

Post by jun1or » Wed Jul 31, 2019 12:33 pm

I also found this thread which references common phantom/ghost knock on cylinder #2, which leads me to believe it's mechanical noise/NVH induced.

viewtopic.php?t=13016

I'll check valve clearances as well.

User avatar
Hondata
Site Admin
Posts: 8750
Joined: Thu Aug 15, 2002 12:13 pm
Location: Torrance, CA
Contact:

Re: Traction Control + Knock Control Strategy

Post by Hondata » Wed Jul 31, 2019 12:57 pm

Last datalog at 19:15. Knock control goers up at 70kPa / 4000 rpm. It's not real knock so probably is from mechanical noise.

You could zero the knock retard tables and that would at least give you consistent ignition timing, but be careful as it may result in more ignition than you planned.

On one of the earlier datalogs I see 11:1. You need to clean up those rich spots. E85 will misfire if it gets too rich. But look at the knock first.
Hondata

jun1or
Posts: 137
Joined: Tue Apr 14, 2009 10:09 am

Re: Traction Control + Knock Control Strategy

Post by jun1or » Wed Jul 31, 2019 1:16 pm

More info proving this is ghost knock.

Here is a sensor overlay of a 20 minute hpde track session showing 37 knocks, all on cylinder #2, all below 4000rpm, all on low cam. There is zero k.counts in boost or on anything on high cam for that matter, yet the high cam pays the price (so to say).

Image
image hosting

Datalog attached shows 25 k.counts during low load paddock driving and the start of a warmup lap! ouch.

Million dollar question:
  • Can I safely reduce the likelihood of knock count detection in the above problematic rpm and load points by increasing values in the "knock sensitivity low" table?
Thanks again!
Attachments
2019-07-13 WMHM 2019 session3 - phantom knock.fpdl
(147.39 KiB) Downloaded 30 times

User avatar
Hondata
Site Admin
Posts: 8750
Joined: Thu Aug 15, 2002 12:13 pm
Location: Torrance, CA
Contact:

Re: Traction Control + Knock Control Strategy

Post by Hondata » Thu Aug 01, 2019 9:08 am

You can try changing the sensitivity but it usually does not work that well. It is worth looking at the engine / trans because sometimes it is something simple like the exhaust hitting the frame which gives the noise.
Hondata

jun1or
Posts: 137
Joined: Tue Apr 14, 2009 10:09 am

Re: Traction Control + Knock Control Strategy

Post by jun1or » Mon Aug 05, 2019 8:18 am

Hondata wrote:
Thu Aug 01, 2019 9:08 am
You can try changing the sensitivity but it usually does not work that well. It is worth looking at the engine / trans because sometimes it is something simple like the exhaust hitting the frame which gives the noise.

Thank you!

I will report back results of both possible remediation paths!

jun1or
Posts: 137
Joined: Tue Apr 14, 2009 10:09 am

Re: Traction Control + Knock Control Strategy

Post by jun1or » Sun Aug 18, 2019 10:47 pm

Hondata wrote:
Thu Aug 01, 2019 9:08 am
You can try changing the sensitivity but it usually does not work that well. It is worth looking at the engine / trans because sometimes it is something simple like the exhaust hitting the frame which gives the noise.
I tried desensitizing the knock sensitivity low table and it did seemingly nothing. I wish I could simply disable knock detection on low cam and/or in vacuum.

Image

jun1or
Posts: 137
Joined: Tue Apr 14, 2009 10:09 am

Re: Traction Control + Knock Control Strategy

Post by jun1or » Tue Aug 20, 2019 11:23 pm

Not alone, reference:

viewtopic.php?t=21352

Post Reply