Dual knock sensors

[JustCallMeOrlando]

Hey all,

I'm getting an Adapronic soon, and I'm keen to use the knock control on it. It has one knock sensor input and given that the 1G has two knock sensors, one at each end, I was looking for some way to bridge them. I've asked on the Adaptronic forums and gotten one response, but i'm asking here for some other ideas.

I'm thinking something along the lines of bridging them using diodes, but unfortunately electronics isn't really my forte.

Ideas?

Cheers,
Nathan

[oldcorollas]

a knock sensor, being a piezo, gives an AC signal when vibrated?
would you need to do anything at all the join them? i spose the signal from one could excite the other but... diode will.. whats the word... rectify the output?

[JustCallMeOrlando]

Yeah, it's just like someone tapping on a microphone, I dunno. C'mon, I know there are some electronics geeks kicking around here.

[chris davey]

would there be that much of an advantage over using one? I would suspect that the sensor at the front would be able to hear any knock from the rear of the engine anyway. I have one of my knock sensors wired to my logger and I just used the front one. I would whack the other one on there for some testing as well but have run out of inputs.

[whatthe?]

Arent the two on the standard loom running off the same wires anyway? I'm pretty sure the 2 in the 7m are? My SAFC only had to be merged into one wire for knock reading....

[cambelt]

I think you will find the 7m gt has kn1 and kn 2 inputs to the ecu.

[whatthe?]

I think you will find the 7m gt has kn1 and kn 2 inputs to the ecu.

You may be right, it's been a while since I wired it up last. Should grab the wiring diagrams for the SAFC and have a look I spose...

[whatthe?]

The TSRM shows one KNK pinout for 7M, but the SAFC wiring diagrams show 2, Hrm. The SAFC diagrams also show two KNK inputs for 2 gens of 1GGTEs. Ah well, me not know then

[CrUZida]

TSRM has KNK1 and KNK2 you goat.
http://www.cygnusx1.net/supra/library/TSRM/d/D_34.gif

[whatthe?]

I suck at teh interweb. I was looking at the 7MGE page . Going back to my bourbon and coke now

[oldcorollas]

Yeah, it's just like someone tapping on a microphone, I dunno. C'mon, I know there are some electronics geeks kicking around here.

how do you join two piezo microphones together?

[Toobs]

Does a knock computer sense knock via amplitude or frequency?

If it's frequency you run the risk of the two frequencies cancelling each other out, however, this probably won't occur as the cylinders won't be firing at the same time.

If it's voltage (amplitude) then you may have a problem using the diode as it will cause a 0.7v voltage drop or 0.3v if you use a germanium diode and they will most likely interfere with each other without the diodes.
You could probably build a circuit to feed the knock computer the signal from whichever sensor shows highest the highest voltage drop without causing too much of a drop... this would require more thought than I can deal with right now though

[JustCallMeOrlando]

how do you join two piezo microphones together?

NFI. The guy Adaptronic creator suggested putting them in series.

[oldcorollas]

Does a knock computer sense knock via amplitude or frequency?

both i think... filtered for frequency, and then analysed for amplitude above a threshold??
or something like that not sure how toyota actually implemented it tho..

a frequency generator, and engine simulator and a bit of time with a scope would sort it out

[camrygt?]

The one I wired just used the one sensor.

No go with the diode the outputs from knock sensors are to small and the forward voltage drop in the diode is likely to totally remove the signal generated by the sensor.

Piezo knock sensors have both a frequency which they resonate at (tuned to particular engine) and a amplitute related to the intensity of the knock. Good knock amplifier circuits look at the frequency/amplitude and the time within the engine cycle to determine knock

If you really want to use both sensors a microphone mixer cicuit will work. Basically need a few opamps to built that but joining the two outputs of the sensors directly together will probably do the same thing, there is as mentioned some risk of the two signals being out of phase and cancelling each other out. I would just hook up the one.. at least at the start.

Have you seen the new software for the adaptronic? its now has configurable background noise for a heap of rpm points so you cancel the background noise from the sensor and the adaptive tuning for timing is finished so in theory these should at least to some extent tune timing themselves now as well as fuel.
There is a wizard for the crank config now as well just tell it the number of teeth on pickups, no of cylinders, angle of first tooth after reset pulse and whether you want a dizzy or coil packs, press OK and you done