What causes KR
01-31-2007, 08:33 PM
#3
Moderator
iTrader: (11)
Join Date: Mar 2002
Location: East Central Florida
Posts: 12,605
Likes: 0
Received 6 Likes
on
6 Posts
There is an audio transducer and electronic filtering
to detect the characteristic ring of the cylinder at
ping. Anything that rings the block can be heard.
Some sensors are more sensitive than others and
some motors make more mechanical noise than others.
There is signal processing done in the attempt to
exclude the irrelevant. Blanking (for intervals when
spark is not active), thresholding (for below-caring
levels of noise) and bandpass for the characteristic
ping note(s). In whatever order.
Of these the bandpass and the blanking used to be
done in the interface electronics. But the threshold
appears to be what we have access to. It's possible
that it's all DSP now and the other filter params are
buried somewhere.
But if you have top end mechanical noises they may
still register even if there is no ping. And that's your
prime question, is knock retard received, ping driven?
I wonder what a fundamentally different head casting,
than original (and embedded in the PCM knock detect
expectations) might do to knock sensing. Put the ping
ring at a different frequency, or put rocker train noise
onto the presumed ping-note? Who can say without a
transducer, a spectrum analyzer and way too much
spare time? Sure does look like you've worked a lot of
changes on the top end of the motor, though. Clakkity-
clack, might talk back....
Then there is the real ping. Which would come from
excess spark timing, too hot a head temp, too lean a
mixture, etc. But I'd expect we're past that? What
have you excluded fuel / air / spark / temp -wise so
far?
to detect the characteristic ring of the cylinder at
ping. Anything that rings the block can be heard.
Some sensors are more sensitive than others and
some motors make more mechanical noise than others.
There is signal processing done in the attempt to
exclude the irrelevant. Blanking (for intervals when
spark is not active), thresholding (for below-caring
levels of noise) and bandpass for the characteristic
ping note(s). In whatever order.
Of these the bandpass and the blanking used to be
done in the interface electronics. But the threshold
appears to be what we have access to. It's possible
that it's all DSP now and the other filter params are
buried somewhere.
But if you have top end mechanical noises they may
still register even if there is no ping. And that's your
prime question, is knock retard received, ping driven?
I wonder what a fundamentally different head casting,
than original (and embedded in the PCM knock detect
expectations) might do to knock sensing. Put the ping
ring at a different frequency, or put rocker train noise
onto the presumed ping-note? Who can say without a
transducer, a spectrum analyzer and way too much
spare time? Sure does look like you've worked a lot of
changes on the top end of the motor, though. Clakkity-
clack, might talk back....
Then there is the real ping. Which would come from
excess spark timing, too hot a head temp, too lean a
mixture, etc. But I'd expect we're past that? What
have you excluded fuel / air / spark / temp -wise so
far?
01-31-2007, 10:38 PM
#4
Originally Posted by jimmyblue
There is an audio transducer and electronic filtering
to detect the characteristic ring of the cylinder at
ping. Anything that rings the block can be heard.
Some sensors are more sensitive than others and
some motors make more mechanical noise than others.
There is signal processing done in the attempt to
exclude the irrelevant. Blanking (for intervals when
spark is not active), thresholding (for below-caring
levels of noise) and bandpass for the characteristic
ping note(s). In whatever order.
Of these the bandpass and the blanking used to be
done in the interface electronics. But the threshold
appears to be what we have access to. It's possible
that it's all DSP now and the other filter params are
buried somewhere.
But if you have top end mechanical noises they may
still register even if there is no ping. And that's your
prime question, is knock retard received, ping driven?
I wonder what a fundamentally different head casting,
than original (and embedded in the PCM knock detect
expectations) might do to knock sensing. Put the ping
ring at a different frequency, or put rocker train noise
onto the presumed ping-note? Who can say without a
transducer, a spectrum analyzer and way too much
spare time? Sure does look like you've worked a lot of
changes on the top end of the motor, though. Clakkity-
clack, might talk back....
Then there is the real ping. Which would come from
excess spark timing, too hot a head temp, too lean a
mixture, etc. But I'd expect we're past that? What
have you excluded fuel / air / spark / temp -wise so
far?
to detect the characteristic ring of the cylinder at
ping. Anything that rings the block can be heard.
Some sensors are more sensitive than others and
some motors make more mechanical noise than others.
There is signal processing done in the attempt to
exclude the irrelevant. Blanking (for intervals when
spark is not active), thresholding (for below-caring
levels of noise) and bandpass for the characteristic
ping note(s). In whatever order.
Of these the bandpass and the blanking used to be
done in the interface electronics. But the threshold
appears to be what we have access to. It's possible
that it's all DSP now and the other filter params are
buried somewhere.
But if you have top end mechanical noises they may
still register even if there is no ping. And that's your
prime question, is knock retard received, ping driven?
I wonder what a fundamentally different head casting,
than original (and embedded in the PCM knock detect
expectations) might do to knock sensing. Put the ping
ring at a different frequency, or put rocker train noise
onto the presumed ping-note? Who can say without a
transducer, a spectrum analyzer and way too much
spare time? Sure does look like you've worked a lot of
changes on the top end of the motor, though. Clakkity-
clack, might talk back....
Then there is the real ping. Which would come from
excess spark timing, too hot a head temp, too lean a
mixture, etc. But I'd expect we're past that? What
have you excluded fuel / air / spark / temp -wise so
far?
I havent excluded anything yet. I use 93 octaine, and my compressions is about 10.85 :1 The exhaust is loud, so I dont hear any pinging. I do have alot of valve train. I did notice that I had the car dyno'd over the weekend and my AFR drops under load to under 10.5 and levels out at 10.75.
I just noticed it on a recent log, and I havent gone back to check other logs.
Foff, Ill have to read what you linked
02-01-2007, 12:10 AM
#6
Its actually pretty easy to determine the frequencies involved in spark knock. It depends boils down to the bore size and speed of sound which can be found with the ratio of specific heats in your fuel/air charge, temperature, and gas constant/molar mass of the fuel air charge.
Depending on your uncertainty the band needed should be fairly tight.
Depending on your uncertainty the band needed should be fairly tight.
Trending Topics
02-01-2007, 09:40 AM
#8
TECH Resident
Join Date: Oct 2005
Posts: 787
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by jimmyblue
"In theory, there is no difference between theory and
practice. But, in practice, there is."
~ Jan L. A. van de Snepscheut/Yogi Berra
practice. But, in practice, there is."
~ Jan L. A. van de Snepscheut/Yogi Berra
I like that and it is so very true. I wonder if it came out that way when Yogi Berra quoted that. He was pretty funny.
02-01-2007, 09:44 AM
#9
TECH Resident
iTrader: (17)
Join Date: Sep 2004
Location: NC - Charlotte area
Posts: 3,747
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by Phil'sC5vette
That's a lot of food.
I havent excluded anything yet. I use 93 octaine, and my compressions is about 10.85 :1 The exhaust is loud, so I dont hear any pinging. I do have alot of valve train. I did notice that I had the car dyno'd over the weekend and my AFR drops under load to under 10.5 and levels out at 10.75.
I just noticed it on a recent log, and I havent gone back to check other logs.
Foff, Ill have to read what you linked
I havent excluded anything yet. I use 93 octaine, and my compressions is about 10.85 :1 The exhaust is loud, so I dont hear any pinging. I do have alot of valve train. I did notice that I had the car dyno'd over the weekend and my AFR drops under load to under 10.5 and levels out at 10.75.
I just noticed it on a recent log, and I havent gone back to check other logs.
Foff, Ill have to read what you linked
so why don't you lean that out to like a 12.4ish down low to a 13.4 up high
02-01-2007, 12:53 PM
#11
Originally Posted by andereck
A few gallons of 100 octane unleaded can help you determine if you're a victim of false knock or its a real tuning issue.
02-01-2007, 02:44 PM
#12
Originally Posted by Xtnct00WS6
A less expensive way if you have htp or efi live is to just take a bunch of timing out and see if u still get KR.
02-03-2007, 07:49 PM
#13
I checked out my log, and it seem to do it around the 2400 mark. I get what looks to be 1-2*. Seems like my advance is around 36-40 when it does it, and it doesnt do it under load. Looks like just cruising. Tomorrow Ill take the car out for a run, and do a couple cruising logs and WOT runs to see what the car is doing. It doesnt seem as bad as I orginially thought
02-03-2007, 09:36 PM
#14
Moderator
iTrader: (11)
Join Date: Mar 2002
Location: East Central Florida
Posts: 12,605
Likes: 0
Received 6 Likes
on
6 Posts
Seems like marginally overadvanced to me, if that's at
any kind of load. Or you are holding off PE. You want to
see the commanded EQ, the cylinder air and the MAP
and get an idea of what this load-point means to you,
whether you want PE or not, whether you want more fuel
or less spark. If this is cruise or with a light roll-on, probably
less spark. If you really must lug it loaded then enrichment.
any kind of load. Or you are holding off PE. You want to
see the commanded EQ, the cylinder air and the MAP
and get an idea of what this load-point means to you,
whether you want PE or not, whether you want more fuel
or less spark. If this is cruise or with a light roll-on, probably
less spark. If you really must lug it loaded then enrichment.
02-04-2007, 01:15 PM
#15
Launching!
Join Date: Dec 2005
Posts: 207
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by jimmyblue
Put the ping
ring at a different frequency, or put rocker train noise
onto the presumed ping-note? Who can say without a
transducer, a spectrum analyzer and way too much
spare time? Sure does look like you've worked a lot of
changes on the top end of the motor, though. Clakkity-
clack, might talk back....
ring at a different frequency, or put rocker train noise
onto the presumed ping-note? Who can say without a
transducer, a spectrum analyzer and way too much
spare time? Sure does look like you've worked a lot of
changes on the top end of the motor, though. Clakkity-
clack, might talk back....
I had 1 loose rocker arm and I would get HELLA KR!! In
ALOT of cells (15 is the highest a cell can go)
