Adventures with MS3 knock module

[Piledriver]

My main issue is my current motor is apparently just about det proof in its current NA form, so beyond abusing the hell out of my motor via doing ~all the things you should never, ever do, trying to get some actual knock to test it, it hasn't seen much use.

At higher RPM, the background noise gets pretty high, but at high RPM you generally don't have to worry about knock (unless boosted)

I've thought a lot about this. What I plan on trying is to take one of the ignition points and set it out of bounds so it will ping hopefully, but at a low rpm and low load. The frequency should be the same at high or low rpm. Hitting a piece of metal fast or slow will create the same frequency just different amplitude. I plan to start looking into knock frequencies of similar boxer engines just to get a very rough guestimation.

http://www.phormula.com/KnockCalculator.aspx

Correct, Steve.
MS3, many J&S Safeguards and even OEM ECUs use the same knock processor chip.

[Steve Arndt]

Typo on phone sorry. Before I meant.

[thedrew]

http://www.vikingspeedshop.com/detonation/

Very interesting read and on point with this topic

[aircooledtechguy]

This is how they're mounted on the P964, I think something similar would work well for the type 4?

It would not be very difficult to make fitting that would mount to the base of the cylinders like this for a VW motor. The real trick is to ensure it has a very flat contact point with which to properly transmit the vibrations(sound) so that the Piezoelectric chip can properly hear the right frequency.

I would bet that someone planning to use OEM 94mm cylinders bored to 96mm could modify the cylinder bases with simple bolt bosses and then have them bored out. Something to think about. . .

[thedrew]

So Steve, a knock window like this?

I believe this is starting to listen at 30 BTDC, and stopping at 5 BTDC

[billz60vw]

no.
it should start at tdc and go 20 to 30 % after.

[Steve Arndt]

That was what I was thinking when I posted that on the previous page. I haven't done the testing myself, just parrot'ing.

[Steve Arndt]

This article shows some graphs of knock after TDC

http://www.avweb.com/news/pelican/182132-1.html

[thedrew]

so more like this:

[billz60vw]

yes that looks better. one thing you can do is try to make rod or bar that will hook to both cylinders on the same side and than mount your sensor to it.

[Michael]

Hey guys,

Can you share what sensor you are using and where I can source some?

Thanks,

[Piledriver]

I bought mine for 2005 Neon SRT/4 (the insane turbo 2.4L thing)... takes the right ones.
No pigtail, takes a common EV1 injector plug directly.
$15.99, part# SU3333
(the link won't work)
IIRC mine were "duralast" store brand, but actually NTK (built under license from Bosch) and much cheaper.
Nothing wrong with NTK sensors at all.

You can probably beat the price on ebay, but you could probably order two at autozone or equivalent tomorrow morning and pick them up at lunch, if they aren't already in stock at the store...
VERY common part.

[Michael]

Ahh thanks for the part number. I'll checkout some local sources to see if they come in at that price. I'm in Canada so I doubt I'll find one at $16.

Good to know at least they are readily available if one from a Neon works. Are there different ones that are tuned to different frequencies?

[Steve Arndt]

Most late 80s in line four golf engines have a wide band spectrum knock sensor easy to find.

[thedrew]

Well its been a few months, and I've been driving around watching my logs when/if the knock retard would fire up.

I'm not sure what I was finding...but it seems like I found a few spots in my ignition map that would cuase a light knock...or at least the system registered a knock and retarded the timing. consistently. and in spots I was in a lean burning area with lots of advance.

I pulled the timing back a few degrees and it would go away. I can't say for certain it was actually a true knock, but it seems to be doing something!