Knock boxes

this year i will be trying out the phormula knock detection unit.
http://www.phormula.co.uk/KnockMonitor-KS-4.aspx

here is a video how it works.
http://www.phormula.co.uk/Video.aspx

i may be testing out a more hitech unit also.
http://www.jandssafeguard.com/VampirePage/J&S_Vampire.html

now i run a stand alone engine management system, so i can make it reduce some timing when it is sensing detonation.

what do you guys think?

Please give some feed-back either way on what you find with either system. I am very interested in a knock sensor that is user friendly. I do not ever plan on exposing my engine to detonation, but we all know sometimes unplanned things happen.

Hopefully this is basic enough to understand.
The Phormula "knock box" at least is a move in the right direction. It's still missing some crucial information though. The Phormula guys at least put in some frequency banding and ask your your bore size to set the center frequency.
Engines, depending on bore size, have a unique harmonic frequency. They resonate at specific frequencies when they knock. That's why in the old days you have specific knock sensors for different engines. They were resonant sensors or "tuning fork" designs with a very narrow band response. A sensor from an engine with a 4" bore wouldn't work on something with a 3.8" bore. Hit your engine block with a hammer and measure it's frequency response. This is what the engine will do when it knocks.
As ECU's became faster OEM's started to use wide band sensors and use custom chips and software to do all the filtering needed. The wideband sensors work faster and the sensors have much less ringing on the output signal.

Basically most all of the "knock boxes" are just fancy microphones. Very basic and very simple to design. The fluff thrown on top like LCD screens and such is kinda cool though. If you watch the Phormula video where they recorded engine noises, the box was saying the engine was knocking when it wasn't. This is because it's not really detecting knock. It's just looking for sounds. If a sound is "louder" than another sound, might be knocking, might not. The run where the engine WAS knocking, whatever they had hooked to the ECU obviously didn't work as it kept knocking. I'd hate to be a head gasket, rod bearings or piston in that motor.

Knock can occur throughout the combustion event. Knock that occurs early on you will not hear, nor will your "knock box" and this early knock is the silent killer of bearings and connecting rods. When it occurs around TDC to 20 Deg ATDC, then you hear it. This is the knock that takes out head gaskets and pistons. And if you heard it, some damage has been done.

To really detect knock, you need to know a few things.
Since knock really only occurs within a specific window of crank rotation for a given cylinder event, you need to what the crank angle is. And you need to know RPM. You have to measure the background noise of the engine. All sorts of things make noises in engines. Valve train, chain case, your new turbo manifold touching the frame, your exhaust hitting the tunnel, whatever. These things MAY also resonate at the same frequency of your engine so the "knock box" will falsely detect them as knock and yank your timing back.
Usually you open two "windows" and look at the knock sensor output. 1st thing is to measure the background noise starting at a CA (Crank Angle) where you know knocking will not occur. At what CA you open your windows depends on RPM. Once you've measured the background noise, you start looking at the sensor in the CA region where knock can occur. The sensor output needs to be run through a bandpass filter to get rid of frequencies outside of the engines resonant frequency, then be gated by your windows. If the sensor output during the knock detection window is higher than some ratio of the sensor output during the noise detection window, your knocking. This ratio changes engine to engine and also with RPM. You NEED to retard the timing at this point. If your running 20 degrees of timing and you detect knock, you may need to retard the timing back to 12 degrees to get it to stop knocking. Ideally you keep track of all this stuff so as you advance the timing back up (in steps), you can remember at what point it knocked previously and just not go there. If you jam all the spark back into the engine, it will probably knock again.

Now the question is do these "knock boxes" work? Well, they do provide some useful information, albeit probably too late alot of the time. If whatever your using to control spark on your engine has some sort of input to retard the timing based on the "knock box" output, that will help a bunch. I'm not a big fan of Water/Meth to control knock. Stuff runs out and it's usually a sign of a poor charge cooler design. Seen way too many smoked motors/lost races when the tank went dry. But if it's all ya have, it's definitely not going to hurt anything.

Problem with the boxes is, IMO, they don't have enough calibrateable control to be truly effective. Most don't look at CA, RPM or any of that stuff. They use fixed gains to figure the ratio of non-knock to knock. They don't keep track of the engines operation. Can they save your #*$&@? Possibly. I know when I am riding hard, I don't have time to look down to see if some light is blinking at me. Too many trees that seem to be magnets for snowmobiles I need to be paying attention to.

Key thing to remember is common sense will go a long ways in protecting your engine. Octane is your friend.

People have been bugging me to make a replacement EFI box that is plug and play for the Yamahas. I have some other projects to finish so I'll look at doing that sometime early next year. I still have to find some time to crack open a Apex box and see what all I can find in there. If it's anything like the R6 and watercraft boxes I have reverse engineered.. might not need to make a new box.

Thanks RunninRX1 for the writeup, very educational for a rookie like me.

I found it worthy to go into a sticky - Turbo- and Supercharger information - http://www.ty4stroke.com/viewtopic.php?p=616664#616664

How a knock box works - http://www.ty4stroke.com/viewtopic.php?p=670923#671017

rxrider
TY Management

insane - your post is now in the sticky

Turbo- and Supercharger information - http://www.ty4stroke.com/viewtopic.php?p=616664#616664

Knock boxes - http://www.ty4stroke.com/viewtopic.php?t=76757

A lot of good info on the subject of knock boxes and knock detection.

rxrider
TY Management

it sure does

There is actually more to it, but I didn't want to get too complicated and go off the deep end. Sometimes I start to explain stuff to people and see that glazed look staring back at me and in their mind thinking "This fool might as well be talking Chinese, don't understand a damn thing he is saying."

RunninRX1 said:

There is actually more to it, but I didn't want to get too complicated and go off the deep end. Sometimes I start to explain stuff to people and see that glazed look staring back at me and in their mind thinking "This fool might as well be talking Chinese, don't understand a damn thing he is saying."

Click to expand...

Thats the great ting about this forum, you don't have to see that look

I say, keep it coming

I've been running the Parker box for three years now. Generally mine does not go off so when it does I pay attaention to it. It will go off if you hit the rev limiter as that seems to duplicate knock. I don't believe the thing out right, more I use it as a base line tool meaning that If I see it working differently than normal then it indicates something has changed for ex. bad fuel. It's just one more tool in the box but I wouldn't rely on it to control boost or timing.

Rx1M5

RunninRX1,
It is sometimes refreshing to read a post with the level of knowledge that you have for electronics.I took my time to read what you wrote and I learned some overview of knock sensor design.Thank you for taking the time to sit down and write all that

Rx1M5 said:

I've been running the Parker box for three years now. Generally mine does not go off so when it does I pay attaention to it. It will go off if you hit the rev limiter as that seems to duplicate knock. I don't believe the thing out right, more I use it as a base line tool meaning that If I see it working differently than normal then it indicates something has changed for ex. bad fuel. It's just one more tool in the box but I wouldn't rely on it to control boost or timing.

Rx1M5

Click to expand...

I think that is the key here.


Especially after reading the explanation post up here.

RunninRX1 - You may post the PhD master on the subject if you like, this is great info. A lot of us enjoys reading advanced techo from time to time, keep it coming.

Anything else you guys want a better understanding of just let me know. See if I can help.