Knock Count and the PV or Are my Pistons in danger

scooterbum46 · May 6, 2015

Since I may have set Fuzzy off on this with my comments about what gauges I watch on the PowerVision, I'll add a comment here (keeping in mind I'm a hack tuner, more used to 8 horse Cushman Huskys, although I'm on my third Twin Cam bike)..

A knock count may be an indicator of precursor events, that you wouldn't even hear or know about without the PV or similar tools on your bike. In the case of 15 - 20 in 150 mile ride, it would tell me that most likely the bike is tuned close to the ragged edge for the conditions your riding it in, including barometric air pressure, ambient air temp, fuel quality engine temp, etc. or you've had a change in something like fuel. 15-20 is not anything to worry about at all... In my situation, I was chasing a bunch of knock events on just one cylinder (turned out to be a plug issue that I got rid of by swapping front to back).

The point I'm making is the PV gauge monitoring is a tool to be used with some underlying understanding of what you're seeing. In this case, some knock event indication given about equal on both cylinders wouldn't be anything to sweat, lots of them all of a sudden, or a lot on one cylinder would raise alarms for me. One area that I know I'm still seeing is a higher count on is WOT as the engine passes through a certain higher RPM - I haven't finished tuning with the PV and I'm pretty certain I've still got some area that haven't had the air flow remapped on yet..

Just my two cents, YMMV.. etc. etc. etc.

msocko3 · May 6, 2015

scooterbum46 said:
Since I may have set Fuzzy off on this with my comments about what gauges I watch on the PowerVision, I'll add a comment here (keeping in mind I'm a hack tuner, more used to 8 horse Cushman Huskys, although I'm on my third Twin Cam bike)..

A knock count may be an indicator of precursor events, that you wouldn't even hear or know about without the PV or similar tools on your bike. In the case of 15 - 20 in 150 mile ride, it would tell me that most likely the bike is tuned close to the ragged edge for the conditions your riding it in, including barometric air pressure, ambient air temp, fuel quality engine temp, etc. or you've had a change in something like fuel. 15-20 is not anything to worry about at all... In my situation, I was chasing a bunch of knock events on just one cylinder (turned out to be a plug issue that I got rid of by swapping front to back).

The point I'm making is the PV gauge monitoring is a tool to be used with some underlying understanding of what you're seeing. In this case, some knock event indication given about equal on both cylinders wouldn't be anything to sweat, lots of them all of a sudden, or a lot on one cylinder would raise alarms for me. One area that I know I'm still seeing is a higher count on is WOT as the engine passes through a certain higher RPM - I haven't finished tuning with the PV and I'm pretty certain I've still got some area that haven't had the air flow remapped on yet..

Just my two cents, YMMV.. etc. etc. etc.

If it constantly has knock in the same areas of the map, there is an issue which needs correcting. It might be something as simple as a little lean spot, or maybe a little to much timing. In Fuzzy's case, his knock is hitting pretty consistently in the same areas of the map, front a rear cylinder taking turns, I've looked at his data logs. I'm thinking it just might be a little lean spot which he can tune out with some more tuning runs.

FuzzyWuzHe · May 7, 2015

Sloufoot said:
Fuzzy, the settings are different from front to rear cylinder so changing the plugs back to front might not make a difference. Sometimes it does, but not always.
You are still ok with just that count anyway. Sometimes I can get some count going from decelerate to accelerate real fast. Not getting them otherwise. Just a fast change in fuel in the cylinders.

I think you are on to something, at least partially .. I did some decal to accel quickly and got knock counts on both cylinders, once the engine was warm they quit happening even when doing the decal/accel. After an ignition cycle off then on and restart the trike .... nadda ... couldn't make it show a knock count ... guessing more tuning needed when cold? Or maybe just let it warm up a little more?

msocko3 · May 8, 2015

FuzzyWuzHe said:
I think you are on to something, at least partially .. I did some decal to accel quickly and got knock counts on both cylinders, once the engine was warm they quit happening even when doing the decal/accel. After an ignition cycle off then on and restart the trike .... nadda ... couldn't make it show a knock count ... guessing more tuning needed when cold? Or maybe just let it warm up a little more?

If your seeing knock counts on dead cold it wouldn't surprise me, it isn't in closed loop yet. You won't be able to auto tune it until it does go in to closed loop. When I was looking at the logs you sent me, it was a completely warmed up engine. If you got it to stop knocking when warm, your in tall clover. I wouldn't be sweating anything when its stone cold.

Decel and a quick accel is just asking it to knock, especially if the rpms drop down to where your on the verge of lugging the engine.

FuzzyWuzHe · May 8, 2015

msocko3 said:
If your seeing knock counts on dead cold it wouldn't surprise me, it isn't in closed loop yet. You won't be able to auto tune it until it does go in to closed loop. When I was looking at the logs you sent me, it was a completely warmed up engine. If you got it to stop knocking when warm, your in tall clover. I wouldn't be sweating anything when its stone cold.

Decel and a quick accel is just asking it to knock, especially if the rpms drop down to where your on the verge of lugging the engine.

I think I'm good ... went for a ride today ,,, let it warm up ... no knock count.

msocko3 · May 9, 2015

FuzzyWuzHe said:
I think I'm good ... went for a ride today ,,, let it warm up ... no knock count.

Now all you have to do is ride the snot out of it.

FuzzyWuzHe · May 9, 2015

Trust me ... That's not a problem ... I average about 15 k a year ... Not bad for a guy who does most of his work in the summer ... I think about retiring all the time ... Just think I'd go nuts if all I did was ride ... Maybe in a couple years or so

FuzzyWuzHe · May 20, 2015

Update .... did some more auto tuning in the 2-3k range (thanks msocko3) and now I only get the knock counts if I do some hard decal then accel.

msocko3 · May 21, 2015

FuzzyWuzHe said:
Update .... did some more auto tuning in the 2-3k range (thanks msocko3) and now I only get the knock counts if I do some hard decal then accel.

Love it when a plan comes together

epper:.

scooterbum46 · May 21, 2015

Has anyone seen Delphi (or from any of the popular tuners) descriptions of each of the data channels coming from the ECM. The reason I would like to see it, pertaining to this thread, is that on the 14/15 Twin Cooleds, I've seen an statement that the meaning of some of the channels has changed, including what the Knock Count means. I really like the fact that you tuned this out with AT, rather than messing with advance settings, etc. I'd note that using MegaLogViewer, some of what were called "knock events" on my PV were actually precursor events that the Ion Knock process sensed and started pulling timing, before there was any real "rattle"....

FuzzyWuzHe · May 21, 2015

scooterbum46 said:
\ I'd note that using MegaLogViewer, some of what were called "knock events" on my PV were actually precursor events that the Ion Knock process sensed and started pulling timing, before there was any real "rattle"....

I have a feeling most of mine where just as you described ... never did I hear or feel anything.

msocko3 · May 21, 2015

scooterbum46 said:
Has anyone seen Delphi (or from any of the popular tuners) descriptions of each of the data channels coming from the ECM. The reason I would like to see it, pertaining to this thread, is that on the 14/15 Twin Cooleds, I've seen an statement that the meaning of some of the channels has changed, including what the Knock Count means. I really like the fact that you tuned this out with AT, rather than messing with advance settings, etc. I'd note that using MegaLogViewer, some of what were called "knock events" on my PV were actually precursor events that the Ion Knock process sensed and started pulling timing, before there was any real "rattle"....

I've never seen any descriptions of each data channel. I do know the ECM part number changed with the introduction of the 14 MY, I would have expected something being changed with the introduction of liquid cooling, left and right fan and introduction of Canbus or Hdlan. Which 1st came to market with the 2011 MY softails, and then being introduced into the touring line starting with the 2014 MY.

Usually the measured even is after the fact, the cause of the knock happens before the detection. It is my understanding, this is true for all the Delphi ECM's regardless of the model year. The reason he was able to tune this out with the AT is I always like to make sure it isn't a leaner mixture causing the knock event 1st, before timing is pulled. If you can't make the knock event go away with a little more fuel, then its not lean mixture induced, it has to be timing induced at that point. Also the assumption is the engine isn't being lugged.

I looked at his map and didn't see any unreasonable timing, so I was pretty sure a little more fuel would take care of the problem.

FuzzyWuzHe · May 21, 2015

And all I did was what I was told ...

scooterbum46 · May 22, 2015

msocko3 said:
the cause of the knock happens before the detection.

Actually, with the ION sensor system, it's monitoring a voltage derived from conduction of the ionized area in the plug gap after every spark. I believe it can sense the approach of preignition based on the value of that voltage and start pulling timing.. Again, I've derieved some of this from looking at logs with MLV. Before I start sounding argumentative, I'll just say, between the Delphi ECM and the PV, I've got a great running trike, without audible preignition, which is a whole lot better than previous ant-knock systems I've seen.. I'll do some more digging regrading the channel definitions and share if I can find anything - it's quite possible it's not openly shared, with DynoJet and the other tuner mfrs having to reverse engineer to figure them out .. Back to Google and search on the forums..

msocko3 · May 22, 2015

scooterbum46 said:
Actually, with the ION sensor system, it's monitoring a voltage derived from conduction of the ionized area in the plug gap after every spark. I believe it can sense the approach of preignition based on the value of that voltage and start pulling timing.. Again, I've derieved some of this from looking at logs with MLV. Before I start sounding argumentative, I'll just say, between the Delphi ECM and the PV, I've got a great running trike, without audible preignition, which is a whole lot better than previous ant-knock systems I've seen.. I'll do some more digging regrading the channel definitions and share if I can find anything - it's quite possible it's not openly shared, with DynoJet and the other tuner mfrs having to reverse engineer to figure them out .. Back to Google and search on the forums..

It's sensing resistance at the plug tip, if detonation or preignition starts, the resistance changes. It is not voltage dependent.

I think flash tuner choices are pretty much like an oil choice, at the end of the day they all pretty much do the same thing. Some have some more bells and whistles, some are geared more towards the professional tuner and some are geared more towards the home hobbyist.

scooterbum46 · May 23, 2015

msocko3 said:
It's sensing resistance at the plug tip, if detonation or preignition starts, the resistance changes. It is not voltage dependent.

I think flash tuner choices are pretty much like an oil choice, at the end of the day they all pretty much do the same thing. Some have some more bells and whistles, some are geared more towards the professional tuner and some are geared more towards the home hobbyist.

Whatever..... I understand how the value is derived and it is there all the time, not just when there is preignition. My point was the value is constantly changing and expressed as a voltage in the 0-5 volt range that the ECM can evaluate and act on. (you can watch it on the PV - the channels are called IONQ1 and IONQ2 and it's expressed in volts).

As far as the choices in tuners, I bought the PV based on what it could do for me without much experience, but also what it offered as I gained experience beyond the basic Autotune. That has played out well for me. Through three bikes. The other tuners all looked to require much more starting knowledge and looked to more oriented to pros with dynos.....

msocko3 · May 23, 2015

scooterbum46 said:
Whatever..... I understand how the value is derived and it is there all the time, not just when there is preignition. My point was the value is constantly changing and expressed as a voltage in the 0-5 volt range that the ECM can evaluate and act on. (you can watch it on the PV - the channels are called IONQ1 and IONQ2 and it's expressed in volts).

As far as the choices in tuners, I bought the PV based on what it could do for me without much experience, but also what it offered as I gained experience beyond the basic Autotune. That has played out well for me. Through three bikes. The other tuners all looked to require much more starting knowledge and looked to more oriented to pros with dynos.....

And computers understand ones and zeros. Of course you have to take a signal and change it into a voltage and that voltage is then read and acted upon by the programming of the ECM, that is understood. But to keep things simple for most folks, all they care to know is what type of knock control they have and how it works.

The PV is a good tuner for the novice, a pro can also get good use of it also. I've fiddled with the PV a little bit, I've also used the PCIII, PCV, SEPST, TTS and the old SERT (Screamin Eagle Race Tuner) which was also made for Harley by the folks at TTS. They do all have their own learning curves.