[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Knock sensing



I'm guessing that if it's lifter noise (and brand new lifters at that!) then it'll make the noise irrespective of throttle position.
So, if it's only making the noise under full throttle it's detonation.

I'd back off the ignition a couple degrees in the 90-100-110 KPA range from 3500 to 5000 as a first step.

I know that MSnS-e can respond to knock input but you have to add additional circuitry.
http://www.glensgarage.com/catalog/megasquirtnsparkextra-daughter-board-only-p-47.html
or http://www.viatrack.ca/

I have no idea how the knock output would look on a datalog but would assume it's only "on" or "off". There is no "intensity" of knock that MS can register.
Basically the MSnS-e control strategy is to remove "X" degrees of timing if it senses knock then return the timing at "Y" degrees per engine cycle. X and Y are user settable.
So, all MSnS-e knows is if you have knock or not and then responds with retard and subsequent advance once knock goes away.

The audio DIY knock detection info is here:
http://www.autospeed.com/cms/article.html?&A=1353
http://www.autospeed.com/cms/A_0353/article.html

It's amazing the stuff you can hear with these. Coolant gurgling, turbos spoiling down, lifters clattering. Unfortunately detecting knock amongst all that takes some practice.

From: "C Boyko" <roccit_53@xxxxxxxxxxxxxxxxxxxxxxx>
To: <scirocco-l@xxxxxxxxxxxx>
Sent: Tuesday, August 28, 2007 9:05 AM
Subject: Knock sensing


> Okay, I have a perfectly good knock sensor sitting in a box, and I'd like
> to wire it up to a light on the dash, so I can "see" any knock. (This is on
> my MSnS car, an 8V).
> Once in a while I hear a light noise that sounds like lifters, and may well
> be (since at this point my oil pump is still an unknown quantity), but it
> occurs at a time when it could also be knock (high rpm like 4500+/high
> load). Obviously I'd like to avoid knock.
> Also, for the 'Squirters, what does knock "look like" on MegaLogViewer when
> reviewing datalog graphs? AFRs in the suspect areas look fine (13:1), but
> it may be a spark advance thing. Oil pump will get swapped for a known
> hydro version next oil change, so that'll resolve that as an option for the
> noise, I just don't want to be doing any damage in the meantime.
> 
> Also, I got some piezos from Jim R a long while back, who's got the recipe
> for that little listening device? It was mentioned on list a long time ago,
> but searching the archives is just too painful for me. Please take pity. :)
> Cathy
> 
> 
> _______________________________________________
> Scirocco-l mailing list
> Scirocco-l@xxxxxxxxxxxx
> http://neubayern.net/mailman/listinfo/scirocco-l
>