Recently, another vendor here has been making various claims regarding my product and methods.

One of his claims was that the equipment I use to monitor the voltage of the factory knock sensor for occurence of knock activity was "useless" and "inaccurate"

This weekend, while having an opportunity to tune many cars on the famous 91 octane **** gas, I have been seeing a lot of knock. In fact much of 91 octane tuning is tuning the car to make it run smooth on the road and not have any knock activity. When I see knock in the data log or I hear it or when I see the ecu pulling timing it is a situation which needs a tuning adjustment.

EVERY time I hear or feel knock - EVERY TIME - I see it in the data log which logs at a rate of approx 12 times per second

Here are some examples


The green is the knock

Keep in mind these are all minor small knocks. Big knocks show up to 4 plus volts but I do not ever see those high a knockk when I am tuning (at least not be deign)

Anything over 1 volt will cause the ecu to pull timing and go to a richer a/f

 

Is there a way to post the data log images at about half the size? I feel I'm viewing through a microscope.

 

After the ecu sences >1vdc knock and runs a rich program, does it revert back to the normal tuned state after the WOT pull (with knock) is completed or after the ecu is reset from battery disconnection?

My car kept getting 1.75vdc at 7500rpm just from stock catback.

 

I think Shiv thinks your method is useless" and "inaccurate" is because of the sample rate.
The XEDE samples at 100 times per second when it logs knock...

Just a guess though.

 

Not exactly on topic but...

Is it possible for you to save these graphs on say a usb drive or cd? A lot of times I lose papers.. Just asking because I'm set for another custom tune Feb 4th and I'd really like to have all the data I can possibly get

I thought I saw a thread about this somewhere a couple months ago.. but either I'm blind or its buried now.

 

All i know is that i got my ECU back from Al, logged a few HARD pulls on 92, 93, and 110 oct. No knock was recorded, and my car is running stronger than ever! Thanks Al. Some time after my military days are over, when i can move to a real state (North Dakota actualy belongs to Canada), i'll have Al set me up with some good road tuning.

 

Just out of curiosity..
On your graphs when the voltage goes over 1 volt for knock there is no change in a/f?

 

12 samples a second is really slow when an engine is at 7500RPM. 7500RPM is 125 revs per second so at 12 samples per second you are only recording 8% of what is happening.

 

Not only that, but those 12 instantaneous samples in that one second aren't going to be timed to listen during the small portion of the combustion cycle that knock occurs. So it's a crap shoot anyway you look at it. That's why they make det cans since most human brains can sample a lot faster

shiv

 

The way I see it, you are just full of hot air, negativity and sales talk. Your followers may hang on your every utterance - but over here at Dyno Flash I am about action and results - not speculation and negative vibes.

The data logs I use are IN ADDITION to and supplimenting other more traditional methods of knock detection. I have found that this method os very helpful and to date it has never missed a knock event. Of course, you have to know what to look for.

BTW - not that you care, but remember the low ball BR stage II car you dynoed a few weeks ago ? A customer with nearly identical mods and 93 octane went 12.0 in the 1/4 mile with PUMP gas in Florida with the same identical tune that you claimed did not make any power on your dyno.

 

But at 3500 to 5500 the sample ratio of samples per combustion events is much more favorable

Not much escapes the method I use at this time

 

Innovate sample rate: Unsynchronized 12 times per second
Combustion event rate @ 4500rpm: 75 samples per second
Combustion event rate @ 7800rpm: 130 samples per second

No negativity or speculation there. Just facts. Even XEDE data logging (which samples at 100/sec ) isn't good enough for truly accurate knock detection. This is because the sample events are not synchronized with the spark events. In other words, the sample event could easily occur too early or too late in that particular combustion cycle, completely missing a knock event altogether. What you are hoping in looking at datalogs is that you conduct enough logs to cast your "monitoring net" wide. Take this information and put it to good use. Don't insult me for providing facts... sheesh.

-shiv

 

No. Sample rate is the same. You're just having fewer engine events per second at lower engine speeds.

 

I guess the only real way to go then is with the exede piggy back. ( )

After all you have logged so many impressive victories in so many forms of motorsports with that lovely piggy back.

I hear that MOTEC is going top thorw its data logger dash system out in the trash and also get the exede very soon

Shiv - you are really a funny guy.

 

Shiv is not saying to buy a piggyback, but to find a better way to log knock more effectively. At 3800rpm you are only getting 19% of the actual data. How would you know if the other 80% didn't have knock or just an outlier? There is high chance of getting a false positive w/ such low rate.