Rogue_Ant

I still fail to see how your making a connection with an aftermarket DME chip altering the function of the KLR...


Rogue

Oddjob

Are you sure Im the one that is mistaking one for the other? My whole point was that knock detection/reaction and diagnostic fault detection/safe mode are two separate functions, unrelated. It sure seemed like you were not understanding that distinction.

Regardless, since both knock and limp mode are functions of the KLR, if you leave the stock KLR chip in the stock KLR, there is no elimination of the factory knock control or the diagnostic fault/limp mode/blink code functions - regardless of who's DME chip you are using.

nize

if that is true, why does russell, the actual chip maker, say different ?

xsboost90

awsome thread! i just installed my counter- 701 model- i was wondering if there is ANY way to reset it. I thought by reading that it didnt have a spot to put a button to reset it but that turning the ignition or power off would reset it. It seems that after testing mine- turning by boost up some to get alittle more chance of detonation- and getting on it on the backroads by my house i got 36 counts of knock. I unplugged the thing and did a final install, then when i fired it back up still had the same # showing! Most of these were while i was just pulling away from a stop sign or something w/ a low load- no biggie. Anyway, see how you like my slick install- i still have all the OE parts in a box should i decide to swap back be asured.

- sorry for the crappy pics its dark in my garage currently and my camera doesnt like night pics much.

hally

very clean install, will do the same

333pg333

Yep, nice install there. However there seems to be some speculation of the various abilities of the different brands and where to locate the sensors if you're not running off the stock ones. Also with the knock counts, what do you read as good, bad, or ugly? What's the scale out of? I know that most here believe that we shouldn't tune to the knock count, so what do we do with the numbers then? Does this give us the ability to determine the safe boost levels of various fuels and octanes? Has anyone used the Link knock box?

Rogue_Ant

I think that the parameters haven't been identified - thus the lack of people tuning via knock count.
I tune for less then 3 counts per gear... I typically see 0-1 count. (running E85 is much less likely to knock though)


Rogue

Trucho-951

+1 agree the parameters have not yet been identified, not enough people logging (or sharing) their knock data. Plus, there are so many variables, knock intensity being the largest variable, I think. Also, we cannot compare my knock data to say Rouge_Ant's data, because we do not know which knocks occurred at low rpm and which at high rpm, nor can we differentiate the high amplitude ones from the low amplitude knocks.

I wish I could log my knock data as a curve, some type of histogram that not only showed the number (or frequency of knocks) but also their amplitude graphed vs. rpm vs. boost. That way I could see where in the powerband my engine is most sensitive to max boost.

I suspect the answer to be correlated with the engine's volumetric efficiency (VE), i.e., I would expect the most knock activity to occur in the rpm range where the VE is highest, perhaps in the 4,000 rpm to 5,500 rpm range (that is where the combustion chamber would see the highest peak cylinder pressure (PCP).

However, things to keep in mind are that even if an engine tune (cam timing, ignition timing, AFRs, boost curve) is optimized for a certain octane level to keep the knock counts below some predetermined range, what may be considered "safe" for one engine may actually not be safe for another engine. Some parameters that should be considered are: 1) age of head gasket, 2) type of gasket material ( i.e., MLS, copper, Stock?), 3) type of fastening hardware (Raceware studs vs. stock), clamping torque, 4) condition of cooling system, especially the radiator, 5) additional oil cooler?, and 6) type of use, track vs. street.

I think we have a long way to go to begin really understanding the use of these limited knock data, but I like the fact that people are thinking about it and are posting their results. It’s a good start.

Oddjob

Dan,

I installed mine in the same spot (former clock location). No way to reset it, not that I have found. Mine is around 15k-16k counts now. Maybe it will "rollover" after 99,999.....

Pauerman

I log the knock signal from the KLR's pin #15 through my piggyback. I can see the number of knocks, the amplitude on a 5V scale and reference each event against RPM and boost. I can tell you that with 91 pump I get between 3-5 knock events at 14 psi. This is with WOT timing values at a very conservative 20-22 degrees of advance. If I increase boost to just 16 psi, knock activity as well as the amplitude of each event increases.

At 14 psi, the typical amplitude of my knock events range between 1.75 volts to 2.8 volts. Most events occur at or near when the turbo is approaching 14 psi. I've had to make adjustments to my part throttle ignition map because the lower RPM range in the map uses a lot of advance. I usually see more knock activity accelerating at part throttle to redline than I do at WOT from low RPM to redline.

Hopefully more guys can log this signal cause I'm not sure if the activity I'm seeing is "normal"

Trucho-951

Vic, that is great info.!

333pg333

Yes thanks Vic although I still don't know what I'm reading there in terms of the numbers. What is good or even acceptable? How close do we run the gauntlet? Sounds like a case of risk/reward with possible big consequences.

Rogue_Ant

Vic,
Low-load knock is fairly common - and is something I wouldn't worry about. Which is why I have parameters for logging knock. So I don't worry about harmless events.


Rogue

blown 944

To resurect an old thread

I think I have figured out a way to distinguish low load knock vs knock at high boost pressures.

I have installed another Hobbs switch that when about 12psi (availible in adjustable pressure values) sends the 12 volts to activate the counter, otherwise it is not on and not detecting knock. Then I also have a toggle that will power the switch to see what the counts were during the run.

I thought I would share this info for those who are not using a piggyback and cannot log the counts accurately.

I also use E85 and see very few counts at high boost.

nize

nice !