GM knock sensor on honda ?


New Member
Jan 17, 2003
I have a question that is off any topic related to TR's. But I feel that you guys are best suited to answer it or at least give me some good input.

I know a little about TRs ( my favorite is the "limited" b/c it sleeps so well ) and my friend has an extremly nice GN ( 87' 32K miles ). He has told me a little about the GM knock sensors and how they work in general.

Here are a few things I know about the GM knock sensor.
#1 if the sensor isn't torqued right on TRs it will not read properly.
#2 It can be connected to a "knock alart" and/or knock guage.
#3 Its a voltage deal, not a frequency. deal like my TT stealth.
Correct me if I've wrong on any of those.

Now I saw a pic of a GM sensor mounted on SOHC honda motor, here is a link to the pic among other pics . In that topic the owner of the car uses a standalone FMU to retard timing when the GM sensor sees knock ( just like the TRs ). I my case ( and 90% of other turbo honda owners ) I don't have a standalone FMU, hense "auto-retarding" isn't possible, BUT to have a knock sensor would be a great tuning tool for Sohc honda guys ( Dohc honda guys have a stock knock sensor ).

If you look at the pic on the thread above you'll see that the sensor is mounted around the front of the engine by the header in his case, do you guys think that would be a fair place to put it ? Or should another place to mount it be found ?

Do you guys think this is a good idea or is it just a waste of time ( b/c of something I'm unaware of ) ?

And a quick plea for two things, #1 please don't ask, argue, or comment about the thread above, its not my car, and I don't care if it makes 800hp or 80hp...
#2 ANY suggestion would be greatly appersated ( my spelling stinks, I know ).

If this Thread belongs in another section of the forum, I'm sorry and pleae move it.

you guys are killing me, someone has to have some good info about putting a GM knock sensor on something other than a TR..
GM knock sensors are something like a bandpass microphone tuned somewhere between 2-10Khz (that's from memory so don't anybody crucify me if I'm off a little!) depending on the engine they were destined for.

The resonant frequency is more closely related to the bore size than anything else. Try to find a sensor and matching ESC module from a donor engine with the same bore diameter as the engine you will be using it on.

Also, on your points above...

The tighter the sensor is torqued the more sensitive it is.

Caspers Electronics has a knock gauge but I don't know if anybody else sells a similar setup.

The sensor output is monitored by the ESC which switches a line from 5V to ground when the amplitude crosses a threshold.

Charles Brooks
I wouldn't be surprised if the knock system he is using is the J&S Safeguard.

J&S also sells a knock guage similar to Casper Electronics.

I believe that John from J&S posts on this forum from time to time.


Mikey is correct. The GM sensor is connected to a J&S unit. I haven't been to his site in a few years, but at one time, he had pics of the unit mounted in the vehicle in question.

Larry has quite the gift of gab. I was under his spell for about a year.

The J&S unit, on the other hand, is not BS.

By the way, we have a unit for the GN, in case anyone is interested. The Buick knock control system is good, but some gains can be made with an individual cylinder control system.

Here's a link to a recent review from John Banks on the UK Subaru forum. He's claiming 23hp, by letting the J&S retard individual cylinders:
Originally posted by ttawd3s

If you look at the pic on the thread above you'll see that the sensor is mounted around the front of the engine by the header in his case, do you guys think that would be a fair place to put it ? Or should another place to mount it be found ?

There is lots of theory about exactly what causes the acoustic noise related to detonation.

The one that I find most pausible is that it's the rattling of the piston in the bore from the sudden pressure rise, that actually makes the noise.

That said, then you want to mount the sensor actually on the block. Perferable in the region of the block below where the wrist pin at TDC. That seems to be the most common GM strategy.

And with the threads of the sensor perpendicular to the bore axis.

But, I'm no expert in the field.
On the orginal J&S boxeswit the inidividual cyl retard , heard had problems on the buick v6. Any truth to these rumors?
Correcto-mundo. Our early systems intercepted the EST signal, which is a one channel signal.

The unit would work fine, as long as we kept the retard below ten degrees, or operated in "retard all" mode.

In the "individual cylinder" retard mode, retarding one cylinder robs dwell time from the next cylinder. Above 4000 RPM, we had to limit the retard to ten degrees or less.

Our three channel system does not suffer this limitation. In fact, the unit retards timing by extending the dwell, which increases the spark energy on the retarded cylinders.