Ion knock detection sensitivity

[Mr. Wizard]

Thought you would like that...   

[FBRR]

I don't log on much anymore. But the adaptive decay is based on speed and load. So it will decay out faster or slower depending on engine operating speed and load!

One thing to remember when removing ALL adaptive knock retard is the system was calibrated "assuming" ALL spark retard would be available. And what I mean by that is the "Adaptive" portion PLUS the "fast retard", have calibrated limits of retard. When you zero out the adaptive, yes it will not be "learned", but you have also lost some of the RESOLUTION in the amount of retard available. And one should also know that the "onset" of detonation takes MORE retard than the actual "spark amount that caused the event! So a couple of extra degrees of spark may "trigger" detonation, but it may take a quick retard of 5 or 6 degree to stop the event! Then the "adaptive learn" has triggered and slowly decays back to "FULL Spark" when no further "events" are registed. So fast attack and recovery rates are designed to work in conjunction with the "fast retard."

The big reason a production cal doesn't just "run" on fast retard is the adding and subtracting spark during knock can be felt. The adaptive pervents the "fast retard" from being active and impacting the surging that can be felt in gear.

And while the adaptive "VALUE" is in memory, there is a calibration that can be "Set" that will change the decay rate IF the FUEL GAGE has changed. (i.e. a fill up has taken place. THAT IS AVAILABLE IN THE SOFTWARE, BUT I don't KNOW IF HARLEY USED that function!!) If the fuel gage is not changed the "normal" decay rate remains after a key cycle! Just turning the key on and off doesn't trigger a change. The software uses both "key cycles" an ENGINE RUN flags for various functions. So some functions can be "reset" with just a key cycle, off/on/off. But many resets only "flag" After and "ENGINE RUN" flag has been set. The Knock system uses "ENGINE RUN" not key cycle.
Hope that helps understand the system a little!

[wurk_truk]

I enjoy seeing you post!!!!!

[HogMike]

FBRR:
Thanks for helping me understand just HOW and WHEN my spark settings will default to "setpoint" after an event ( or series of events) that pulls timing.

[whittlebeast]

So how little timing will you guys go to at WOT before you guys start thinking the problem is somewhere else.  Compression, air flow/cooling issues, squish issues, bad head machining...

AW

[HogMike]

At WOT lots of things change.
I'm not really hearing any ping or other issues there, and on the dyno the bike is tuned properly at that load.
I ride 90% of the time at 2200-3000 RPM at medium MAP and THAT'S  where I am most concentrated in my tuning.
My riding style, and the load I have will NOT be optimum for others, but, it is where I am.
So far, I would guess that I'm probably 99% satisfied with my tune.
I DO need to change my warmup table just a little though!
JMO

[HogMike]

So how little timing will you guys go to at WOT before you guys start thinking the problem is somewhere else.  Compression, air flow/cooling issues, squish issues, bad head machining...

AW

In my engine (basically stock with mild cams) in the mid rpm area of 2200-3000 I use about 9 to 18 spark advance at 100 MAP. This is by dyno with knock control monitored.
Other issues with bad heads, bad gas, cooling, etc, etc was NOT used in tuning.

[cts1950]

When you change the engine size or combustion chamber or “camshaft timing” those values might in some areas be close enough to function, but it doesn’t surprise me that you would either get “false” retard, or failure to detect real detonation under some speed and loads!

This is the closest answer to my earlier question concerning Ion sense and modified motors. I have been contemplating the sensitivity issue and will throw this out for discussion. By increasing spark plug gap that would effectively increase the resistance for the ion sense in the combustion chamber and help reduce false fast retard events on modified motors? Has any one experimented with that and what was the results.

[hrdtail78]

I usually close the gap on built engines.

I have notice best power in the mid range isn't always driven by where it starts to ping.  I Have found were a bike will run without ping and went back and removed 5 more for best power.  I do believe for the most part these are ignition walled, but not at every MAP and RPM.

[whittlebeast]

FBRR

With production motors, what is the least timing you have ever seen as a normal answer at WOT?  Sure I understand that water cooled is a little different and 4 valve motors are different, but if you got down to 10 degrees base timing and you are still getting detonation on the fuel you were designing to....  What did you do as a factory calibrator?  Was it time to send things back to engineering? 

AW

[HogMike]

FBRR

With production motors, what is the least timing you have ever seen as a normal answer at WOT?  Sure I understand that water cooled is a little different and 4 valve motors are different, but if you got down to 10 degrees base timing and you are still getting detonation on the fuel you were designing to....  What did you do as a factory calibrator?  Was it time to send things back to engineering? 

AW

Not sure I really understand the question.
Are you looking for (x) timing at some given rpm/load at wot?
If you look at some calibrations, you will see anywhere from 0 advance to 26 (or more) depending on RPM and load.

[Sporty 48]

Hardtail78,
What do you mean by "these are ignition walled?"
Are there limits imposed by timing?
Or by the EFI setup?

I usually close the gap on built engines.

I have notice best power in the mid range isn't always driven by where it starts to ping.  I Have found were a bike will run without ping and went back and removed 5 more for best power.  I do believe for the most part these are ignition walled, but not at every MAP and RPM.

[strokerjlk]

By increasing spark plug gap that would effectively increase the resistance for the ion sense in the combustion chamber and help reduce false fast retard events on modified motors? Has any one experimented with that and what was the results.

you are correct.
I have seen,increasing the gap to 40-42 on motors that had false retard so bad that I would normally turn the knock retard off , calm retard down enough to go ahead and enable spark retard.

[Rider57]

When you change the engine size or combustion chamber or “camshaft timing” those values might in some areas be close enough to function, but it doesn’t surprise me that you would either get “false” retard, or failure to detect real detonation under some speed and loads!

This is the closest answer to my earlier question concerning Ion sense and modified motors. I have been contemplating the sensitivity issue and will throw this out for discussion. By increasing spark plug gap that would effectively increase the resistance for the ion sense in the combustion chamber and help reduce false fast retard events on modified motors? Has any one experimented with that and what was the results.

Does it work? Yes, but not all the time. CR was the major stumbling block. At about 10.0:1 the "trick" didn't work as well.
Above 10.5:1 it didn't work at all.
I have used this on EVOs and Twinkies many times with good results.
Experimentation is about the only way to find out on any particular build.

[cts1950]

Rider 57 and strokerjlk
Thanks for the reply I have been chasing this problem for several years and now I have a direction to experiment with. I am sure that the data logs will show the changes. I typically ride the same route for each data log so I can go through them and campare them side by side for acceleration and loads on hills and the flats.

[hrdtail78]

Hardtail78,
What do you mean by "these are ignition walled?"
Are there limits imposed by timing?
Or by the EFI setup?

I usually close the gap on built engines.

I have notice best power in the mid range isn't always driven by where it starts to ping.  I Have found were a bike will run without ping and went back and removed 5 more for best power.  I do believe for the most part these are ignition walled, but not at every MAP and RPM.

Ignition limited.

[HD_04Ultra]

Is "Adaptive Knock Retard" a new feature in the TTS software?  I have reviewed all the SERT materials and can find nothing on the subject.  Admitedly, my SERT is "old" software.  Infact, so old it does not pass the SERT Upgrade Qualifier.  It is P/N 32107-01E dated 04/25/06.

So if I am right about the lack of AKR in my SERT software, how does that change tuning the timing tables.  As noted in a different post. I tried running 89 octane fuel and went from 0 knock retard to well over 8 degrees of knock retard.  I filtered the CSV file of the data log for greater than or equal to 4 degrees of knock retard for each cylinder and noted the kPa and RPM range each cylinder covered.

Front Cyl Knock Retard was 4 degrees or more from 77.5-103.3 kPa and from 2523-4754 RPM.  I took 2,50 degrees of advance out of the front cylinder from 70-100 kPa and from 2500-5000 RPM.

Rear Cyl Knock Retard was 4 degrees or more from 77.5-94.9 kPa and from 2523-2936 RPM.  I took 2.50 degrees of advance out of the rear cylinder from 70-100 kPa and from 2500-3000 RPM.

I believe these would be the recommended edits based on the SERT user's manual.  Then reprogram the ECM and head out for another data run to collect more data so the affects of the change can be analyzed.

As I understand it, even if AKR was part of the SERT software (), reprogramming the ECM resets the values to zero anyway; so the AKR code really would not affect tuning in any way even if it were there.  After all, the ECM must be repogrammed every time the .MT6 file is modified.

Interesting thread, I am enjoying the reading and think I am learning a lot, even if my software is "old" and perhaps some of the things being discussed do not apply to that particular version of the SERT/TTS software.

HD_04ultra

[ultraswede]

Just as a FYI, all the functionality discussed here has nothing to do with which tuning software is used.
The functions exist in the stock ECU, the difference is if you can tune them or not.
That is the difference in the various tuners out there.

[whittlebeast]

Has anyone verified any of this on a timing light with a crank that has been properly indexed/degreed?

[FLTRI]

Has anyone verified any of this on a timing light with a crank that has been properly indexed/degreed?

Got any idea how to accomplish that?
HD ceased offering a timing hole over 10 years ago with the introduction of the TC.
Bob

[whittlebeast]

A duel trace recording scope with one lead on the CPS and the other on the coil lead would do it.

AW

[hrdtail78]

Better be pretty quike with that dial back knob.

[mike 120]

When you change the engine size or combustion chamber or “camshaft timing” those values might in some areas be close enough to function, but it doesn’t surprise me that you would either get “false” retard, or failure to detect real detonation under some speed and loads!

This is the closest answer to my earlier question concerning Ion sense and modified motors. I have been contemplating the sensitivity issue and will throw this out for discussion. By increasing spark plug gap that would effectively increase the resistance for the ion sense in the combustion chamber and help reduce false fast retard events on modified motors? Has any one experimented with that and what was the results.

Does it work? Yes, but not all the time. CR was the major stumbling block. At about 10.0:1 the "trick" didn't work as well.
Above 10.5:1 it didn't work at all.
I have used this on EVOs and Twinkies many times with good results.
Experimentation is about the only way to find out on any particular build.

Is it related to compression ratios or cylinder pressure? Ccp can vary a fair amount depending on cam at same compression, yes?

[FBRR]

Whittlebeast,
On normal aspirated engines the "min. Spark" at high speed and loads that I have done were about 5 degrees. On a supercharged engine that number was down around 2 degrees, vs. high boost.

On production engines the knock system allows a little "wiggle" room vs. Octane and deposits. Before the knock systems were as robust as they are now WOT spark tables might be a couple of degree less on noramly aspirated engines.

Also when I mention those "low limits" that is in the "LOW Octane Spark advance".(High Octane spark limits would be 4 to 5 degree higher)
The adaptive spark runs between two different spark tables by a multilpier. (i.e. is the multiplier applied by the adaptive knock is .5 the value for delivered spark would be hafl way between the "High oct. table and the "low "ocatane table. Ther eis a complete spark MAP for both high and low octane. Then there are modifier tables based on Power enrichment vs. time in P.E, as well as spark reduction BEFORE fuel cut off! The spark reduction before the "hard limit" fuel cut off for Max. RPM is a cushion as you approach "over speed fuel cut off."

I'm not sure that helps with a Harley engine. Those values are dependant on "deposit over miles" and octane used for each spark map.

[BVHOG]

One other interesting thing I have noticed is that you can have whole numbers in your timing tables yet have .25 increments during data logs.