RCinXS

I'm in need of 2 Knock Sensor Mounts that come from the engine valley of the late model 32Vs.

Need them in order to implement Knock sensing capability to my 16V during the upcoming Spring project.

I've tried the Big three with no luck.

There isn't even a part number for them.

Jim Bailey says there has to be a bunch of blown blocks out there, so I figured the 928 Rennlist community was worth a try.

Anybody have a blown late model 32V block that has them and want to sell them?

Does anyone know where I can get two of these sensor mounts?

Here's a picture of what I'm looking for:

WallyP

The knock-sensing system depends upon time-of-flight of the precise frequency of the sound of knocking from each cylinder to both of the sensors. I would assume that means that sensor location is very critical. I would also assume that the "mounts" are part of the block, and not a separate part.

You might get it to work by just firmly bolting the sensors to the same locations in your block.

RCinXS

Placement, Impedance and Frequency have all been taken into consideration.

The new EMS does not distinguish between the two sensors. It only monitors for a knock.

Thought about drilling and Tapping, however the mounts are removable per Porken.

He has done this on his 32V.

See the attached Picture of Porken's install.

SeanR

Porkens block is a 32v in that picture.

Don Carter

I have extra knock sensor mounts. PM me.

RCinXS

Porken sent the picture. It is an early 32V that did not have sensors.

86.5 I believe.

He assures me that the mounts are removable. And that is how he added the sensor.

Indeed the "plugs" can be removed from my 16V as they have a provision for a hex key to turn them.

RCinXS

Many thanks to all who have replied!

One is on the way to me.

Only need one more.

Rob Edwards

Wally- I believe you 100% on this point, but am curious as to why this is the case- Between the crank position sensor and (on 87-up cars) the Hall sensor plus the presence of a knock, that combination of the three elements aren't enough to tell the EZK which cylinder is knocking? Or (as you say) is the combination of knock sensor distance and the CPS/Hall all necessary? When you log with the ST2 it (of course) tells you which cylinder is being retarded based on knocks so there's clearly some algorithm being used to pull timing- just curious as to why the physical positioning of the sensors is so important. Maybe at high rpm the conduction velocity of vibrations through aluminum is slow enough that that delay is part of the calculation?

RCinXS

Hmmm.

Not to step on any toes, but I just have to ask. It simply peeks my curiosity.

And I am a firm believer in that you can never learn enough.

1: The EZK indicates which cylinder has knocked? I can see this as the EMS compares the knock vs time with the associated spark on a fully sequential ingnition.

2: Or does it actually retard the spark for that particular cylinder? Also can see this as a possibility for the same reason as the above. In this case, does it continue to tailor the spark for the offending cylinder(s)?

3: Or does it retard the timing cycle overall? This is the method used by most EMS computers of today to insure a smooth running engine.

I have a Modified and Supercharged 16V. I will be programming the new EMS to retard overall as in #3 for obvious reasons. So the EMS only needs to know that a knock has occured. All adjustment curves are based on the severity level of the knock(s).

Again, Just curious about how the EZK EMS did this.

Still need Knock sensor mounts.

RCinXS

I've had a couple of reponses and indeed there are dead blocks out there. I have asked Porken to please chime in on this thread, on the removal of the sensor plugs. The block does need to be S4 and up.

John Speake

The EZK knock system is very advanced, even by today's standards. It detects which cylinder(s) knock, and quickly retards the igntion on only the affected cylinder(s) by 3 deg. If there's a second knock it adds another 3deg, up to a max of 9 deg. If the knock ceases the igntion is slowly advanced back to the stock timimg.

This is easily observed with the SharkTuner, in real time using the datalogger.

James Bailey

John it retards for a single knock or for a number above a threshold ? Later cars fault code is stored for "excessive " knocks as I recall.

Lizard928

If you change the mounting point to the center plug (requiring a large change to the intake to work), then you will only need one sensor and therefor plug to implement this idea.

These plugs seem to be in the block with loctite (blue), they can be tricky to remove. I would be tempted just to drill and tap the plugs that are already in there.

RCinXS

Lots of good information. Thanks to all of you.

I didn't know that the EZK was caplable of retarding spark to various cylinders way back in the day. Pretty cool!

Due to the higher than normal compression ratio of a U.S. engine, I need the knock sensors not only to retard the overall timing of the engine under boost, but as an additional safety margin, the new EMS is programmed to bleed off boost if the knocking can not be overcome, as well as some other safety features such as complete ignition and fuel shut off in case of critical failure.

I knew the mounts were removable but I didn’t think that the mounts would be difficult to remove. It would seem that I was mistaken.

Plan is to mount 2 sensors. One in the very front and one where the Air guide would normally be mounted to the rear isolation mount.

This gives very good placement for the knock sensors. Almost dead center of cylinders 1,2,5,6 and the rear one almost dead center of 3,4,7,8.

These are the locations that Porsche used to mount the sensors on the later 32V engines.

Due to the reworking that has been done to the later model air guide and Throttle Body (LH-Jet components) along with the way it is then connected to the intercooler, the rear isolation mount is no longer required.

That would be the easy one to modify for a knock sensor. Remove the Isolator and drill and tap for the larger 8mm bolt.

The front one is a bit more difficult to deal with while it is in the block. Drilling is no problem at all. Tapping however may be difficult.

I suppose if it really comes down to it, I’ll simply need to do it this way.

The adapting of the Cam Home sensor to the 16V for sequential FI took some math to figure out and some machining but I really didn't think the knock sensors would be such a bump in the road.

Just goes to show you, there's always something to learn.

Again, thanks to all who have given advice and thoughts on my little project.

John Speake

It retards 3 deg on the first knock, 6 deg 2nd knock 9 deg on 3rd knock.

That's if the knocks follow each other fairly quickly, if they are more random, the igntion 3deg retard for the first knock may have started to return to stock timing. The retard of 3 deg is activated imediately, but the recovery time back to stock timing if no other knocks takes about 6-7 seconds.

So if timing had recovered to 2 deg retard and there was a second knock, igntion would be imediately reatrded to 5deg by a second knock.

The EZK ECU only stores fault codes relating to knock sensor or Hall sensor faults "implausable signals". It does not store any knock counts.

Diagnostic tester will run a knock count test. They run a test over 10,000 engine firings (about 2 minutes at 3000rpm) and the number of knocks. The pass mark is less than 50 knocks in that time.