| piezoelectric sensor that has a forward bias of 2.33V from the ECU. When the sensor is vibrated, the piezo element produces a voltage signal similar to that of a microphone. The more it is vibrated, amplitude-wise, the greater the swing in its voltage output from this baseline voltage. Im inclined to say that you may not be familiar with piezoelectric materials - you might want to google it - pretty interesting stuffs. :) You are correct that it wont miss the sound of detonation, but what I am getting at is that it will hear a lot more than detonation and mistake it for detonation, thereby pulling a lot of timing advance in instances where there is no detonation at all. FWIW, I built a turbine rotating group balancer unit using a piezoelectric buzzer, a comparator circuit, and a timing light's electronics. It is pretty impressive what kind of voltage potentials you can get from piezoelectric transducers. Basically I have used the piezo transducer to produce a voltage signal which correlates the balance of the rotating part. The comparator circuit has one input that sets the threshold level and the other input is from the transducer. When the transducer voltage exceeds the threshold voltage, it turns on the comparator's output which fires the timing light circuit. The heavy side of the rotating part is at the bottom of the rotation at that moment and it creates a stroboscopic effect showing the heavy side.
An old post from another forum I visit:
I was going to use the circuit I built for the optical tachometer but decided to leave that circuit alone and build a new one. I took a trip up to Radio Shack to pickup a few odds and ends and went through the component cases and actually found everything I needed to build the balancer - sweet score at the Rat Shack! So I picked up several piezo buzzers, an LM339 Quad Comparator IC, a few 10K panel mount POTs, some nice machined aluminum knobs, a few switches, a fuse holder, a fan, a few TIP120 darlongton transistors, and well, I picked up about $350 worth of stuff which also included a few packs of rechargeable batteries, a soldering station, and other misc items I use around here. I put together the comparator circuit and linked in the piezo buzzer element to test its operation - was pretty interesting to see that even with just minimal tapping of the piezo, it would generate ~8V potential. I suspect that this transducer will work out very well in this application given that the LM339 is capable of seeing about a 2mV difference between inputs - I can only imagine what small amount of mass would be necessary to produce a 2mV output on the peizo. For kicks I blew a few pulses of air from my mouth onto the piezo and it would trigger the output on the comparator. I took apart the extra timing light I had here to get the guts out and also picked up a different xenon bulb, one of the u-bend style as the timing light had a long straight tube that would have required a larger diameter mounting tube. An aluminum plate was machined to hold the piezo element directly below the shaft center on the swingarm and I used 3M double-sided viscoelastic adhesive foam to mount the piezo, electrically isolated. In the bottom of the swingarm I drilled and tapped a hole for a small screw to act as the pin for transfering the load to the piezo. I left one of the small springs in place to provide support for the turbine/shaft so the piezo wouldn't have to bear the full force of the rotating part. At the other end of the swingarm I installed a small screw that can be tightened down to contact the rear baseplate and prevent the piezo from being damaged when installing/removing shafts and for storage. This screw backs the swingarm from the piezo and leaves about a 0.060" gap between the screw and piezo plate. I also received the extruded aluminum case to put the guts into and got everything mounted up - using standoffs for teh electronic bits inside. I cut out an endplate for the housing and left a tab for mouting the strobe light housing. I turned down a piece of aluminum tube I had here so that I could install the lens that was on the timing light and cutout a backplate and installed a grommet for the wiring. A piece of teflon was used inside the case to electrically insulate it from the ~600V that the strobe receives from the strobe circuit. I mounted up a shaft and started playing around with balancing one of the T04GT turbine/shafts. I have spent about an hour with it, adding some aluminum tape to the heavy side in intervals. There is a process to using this rig: Start with low rotational speed and with the sensitivity set to the lowest setting. Increase the sensitivity until the strobe fires. The heavy side will be the lowest point.
Leave the rotational speed and sensitivity adjustments alone, turn off the balancer. Remove material on the heavy side. Reinstall the group and turn the balancer back on. I have found that in running the shaft at the same speed, each time material is removed, the sensitivity has to be turned up a little more. I kept going like this until I was at the max sensitivity and then I increased the rotation speed a step at a time and left the sensitivity alone. I would just raise the rotation speed each time until the strobe began firing. The way the electronics work is pretty simple. The comparator has two inputs, one is connected to the piezo and the other is connected to the adjustment knob. The adjustment knob allows you to set the voltage threshold that must be exceeded by the piezo at which the comparator will turn its output on. The output is connected to the strobe circuit. When the heavy side of the turbine/shaft is bottom dead center, the piezo will be at its maximum voltage output. The sensitivity adjustment allows the threshold to be adjusted to just a few millivolts below the peak output voltage of the piezo. It is interesting to see the response of the setup when you lower the threshold voltage - it will actually appear as though the heavy side is moving. The trick is to turn the sensitivity up just enough to where the strobe doesnt fire at all and then back it down just a hair until the strobe begins firing again. AT this point the sensitivity is set dead on and the strobe will be firing just when the heavy side is low. Im going to continue playing with it and I Will build a calibration tool to double check the accuracy of the machine. I will turn a shaft and intentionally install a small straight piece of wire sticking out of the side of the shaft. I will be able to spin this up and see in closer detail how accurate the machine is. I will also be able to do this with a known amount of mass to determine empirically just what kind of precision I can balance down to. So, enough with the chitty chat, here's the pics!
 
These last two shots were taken without the camera flash and with the balancer running. You can clearly see the stroboscopic effect here even without motion video! 
So now Im going to get these two turbines/shafts balanced up and then make an additional shaft for balancing the compressor wheels. Unfortunately although this process is considered dynamic balancing, it isn't truly a fully dynamic balancing rig as I would need to have two piezo sensors on both journals and spin the assembly with everything connected. I have plenty of the extruded casing left over and the comparator circuit actually has four comparators (I'm only using one), so perhaps sometime down the road I could make the balancer v.2 with twin piezo, twin strobe, etc etc.. But for now I think this will work out sufficiently. I'll just have to get the T04GT back together and see if I Really am getting good balance! I'm pretty happy with how this balancer project has turned out so far - it has come together relatively easily, although there were a few hiccups here and there, but I dont give up very easily and I've been prett determined to do this and there isn't much going to stop me when I have my mind set on something. I'll report back once I have balanced one of the wheels and also tried spinning up a known off-weighted shaft for calibration.
[ ashspecz.com ]
[ agpowers@bellsouth.net ] Enthusiasts soon understand each other. --W. Irving.
Are you an enthusiast? If you are out to describe the truth, leave elegance to the tailor.
Albert Einstein
|
The knock sensor isn't a resistive device - it is a - 
