Sam and all,

Would anyone care to share wisdom and experiences on their use of knock sensors. Brands, sensor mounting location, false readings, costs, etc...?

I think I'd like one of these on the '68 turbo Dart at some point.

Thanks very much,

Lou

_________________
Home of Slant6-powered fun machines since 1988

I think(!) you will have difficulty finding a sensor that can discern between tappet clicks and ping.

_________________
一期一会
Too many people who were born on third base actually believe they've hit a triple.

From what I've heard from Sam and Ehrenberg, this is not a problem. Maybe with loose valves or certain cams?

Lou

_________________
Home of Slant6-powered fun machines since 1988

Just sensors or timing control systems? From my limited exposure on the subject the only system I would buy is the J&S. http://www.jandssafeguard.com/

_________________
Joshua

I am not necessarily interested in realtime feedback control right now, just a knock sensor/gauge. I'll look at that, though.

Thanks, Lou

_________________
Home of Slant6-powered fun machines since 1988

I see these Alot on the Turbo Grand Natiotinal / GNX V6 Buicks - They Have a knock Gauge with the Swirly arrow 6 logo on the Backface of the gauge.. I think they are aftemarket Not Original GM parts..


Maybe hit up a turbo buick forum??

I am sure Tom/Shaker223 can thow a few ideas in the ring too.

Greg

_________________
http://www.youtube.com/hyperpack

Hence my desire to build a fuel injected hydraulic lifter slant.

The ordinary knock sensor senses vibrations within the 4-8 kHz range, ususally around 6 kHz, and may give false readings in an engine with mech. lifters, it is often just one lifter out of adjustment that causes problems.
An alternative may be to adapt parts from a modern engine management system, that use the spark plugs as sensors using ion-sensing. It is used in - among others - Saab and actually also some cheap Chevy models!
It is not my project with the highest priority, but I am looking into how it may be used in a simplified setup for use in older cars.

I had a post in a different thread where the subject was touched.
I wonder if some similar sensing techniques were used by Dr. Jacobs, maker of the Jacobs Electronics ignition systems, I once read som rather vague descriptions of his work. Some of you may have some info on that.

Olaf.

_________________
Aspenized

I don't know that a dash display is going to be able to differentiate between mechanical noise and knocking. The J&S system can tell the difference between mechanical noise and knocking because it only listens for knocking during a window when the knock can occur.

I'm friends with Chris Jacobs and could ask him if he's done anything with knock sensing.

<edit>

I e-mailed Dr. Jacobs. I'll report back.

_________________
Joshua

I would (and do) have a great deal easier time putting a great deal more faith in what Sam says than in what Ehrenberg says (or prints). That is because Sam doesn't guess or make crap up when he doesn't know the answer to a question.

I do think it would be worthwhile to do some careful testing across the whole RPM range with the timing retarded, high-octane fuel, and cold spark plugs to make sure there would be no possibility of ping, monitoring the timing and comparing it to the timing with the knock sensor disabled, to make sure mechanical noise isn't triggering the knock sensor. Alternatively, monitor the signal from the knock sensor under those "no possibility of ping" conditions across the RPM range to make sure it's not being triggered.

One of the benefits of a knock sensor -- its ability to "hear" ping that isn't audible to the driver -- is also its potential downfall on a mechanically noisy engine, and I think most of us have had the experience of hearing valvetrain clatter at certain RPMs even with the lash correctly adjusted. And there's also the additional valvetrain noise on a cold engine before the valvetrain has reached operating temperature; I think(!) you'd probably have a fairly- to very narrow valve lash tolerance before you'd create cold-engine driveability issues due to false triggering of the knock sensor by a cold valvetrain. I suppose one workaround for this if the tolerance is impossibly narrow (can't adjust the lash tight enough to avoid cold-engine false triggering without making 'em too tight on a hot engine) would be to rig up a thermostatic cutout such that the knock sensor is out of the circuit or ignored until engine coolant temp reaches normal working temperature.

It might be good to progressively loosen the lash and re-run the monitored test across the RPM range until you see the knock sensor start to be triggered, so you know the limits of your valve lash tolerance.

_________________
一期一会
Too many people who were born on third base actually believe they've hit a triple.

I have heard from those that drive factory setups with knock sensors in conjunction with EFI engine management systems which retard the timing upon sensing preignition, that the knock sensors can make their cars run very poorly when they do not use premium fuel. This is regardless of other complicating factors such as valve noise. This seems to be more a function of the nature of preignition, and the amount of timing that must be pulled to stop it once it starts. It seems to take about 10 degrees of timing pulled to kill preignition that has worked into its hysteria point.

So, even if your set up works like a charm it has the potential of killing performance. I suppose better that than the engine though.

I have a Jasper knock gauge, which has never seemed to give me a false positive. I have always heard preignition when indicated on the gauge. At the very least it has not sent me false alarms. However, I cannot guarantee there has been no inaudible preignition which the gauge failed to pick up. Have I been appropriately vague on that?

This gauge has a series of LED's around the dial face. The first event lights the first LED to the left. Each time an additional event is recorded, another LED lights up. When pre-ignition begins, there is a clockwise sweep across the gauge. The last LED to the right is red, and is a warning to let off at that point. I mounted the sender right in the side of the block in a hole I drilled and tapped. My valve train is fairly noisy, and yet the gauge seems to ignore it. Once in awhile it will indicate a single event upon start up. I did get mine off of a web site that is dedicated to Grand National devotees. It has the Buick turbo logo on it.

Would I use it in conjunction with an ECU that pulls timing? I suppose so. But I would want to fix it so the events never occurred. In other words, I see this as a tuning aide and not a closed loop part of the engine management system.

Sam

_________________

For what it is worth, the Ford MAF based sequential fuel injection system i am basing my EFI conversion on doesn't use a knock sensor. So I guess I really don't have a good reason to prefer a hydraulic motor, other than that it is just easier to not have to adjust the valves...

Here is a link to a short and relatively good intro to ion-sensing systems, Making sense of ion sense technology.

Olaf.

_________________
Aspenized

According to my "other brand buddy" during the mid-late 90's Corvette engines used a knock sensor and mechanical lifters in "certain applications", this would be an interesting research project to find out what year and how much that sucker costs...

FYI,

-D.Idiot

Thanks everyone for your responses. I feel confident I can test for valve clatter background on a given sensor.

Sam, thanks for your encouraging experiences. If you have a source or PN's for this Jasper stuff, that would be great. I'll give a look online...

Cheers,

Lou

_________________
Home of Slant6-powered fun machines since 1988