Hacking the CP-2

[George Schmermund]

The KZQ transducer in v1.? has now been upgraded to v2.0. Testing so far with the v1.? has been very instructive in regards to resonances. The main goal from the outset was to be able to measure the energy of a pellet at the target accurately and without ambiguity. To that end v2.0 now has a single resonance of ~ 52 kHz which is easily seperated from the impulse energy measurement using a low pass filter. The signal of interest is now contained within a time window as an impulse that will be relatively easy to quantify.  There is still work to do in fine tuning the instrument, but the new version is a big step from where it started.

 The photos show the assembled transducer. The stand alone weight is 3.5 lbs. It's the same size as v1 and uses the same front energy absorber (not shown) as described in an earlier KZQ post.

[George Schmermund]

The next step in setting up the KZQ for further testing was to mount the device in a box that could provide a stiff mounting arrangement and add extra mass to it as a system. The box is actually a version of the quiet trap that I posted as a construction project some time back. The box is 13.5" square and 3.5" deep. This package is less deep and lighter than the original design and weighs in at 20 lbs. The original trap design was posted here:  https://www.gatewaytoairguns.org/GTA/index.php?topic=127358.0

The photos show the KZQ mounted in the trap with and w/o the front impact absorber.  I'll try to post some measurements in the next couple of days.

[George Schmermund]

It might be instructive at this point to show the difference between the impulse signal from the KZQ going straight into the scope and first running the signal through a charge amplifier. Keep in mind that the signal generated by the KZQ's piezoelectric crystal is actually a charge of only a few Picocoulombs (pC). Therefore any resistance in the signal path will begin to bleed this charge off the crystal while the measurement is still being made.

A charge amplifier has an extremely high input impedance compared to the scope's input. It also has a low impedance output signal that can easily drive the scopes input impedance load without loosing signal information.

As can be seen in the screen shot channel 2, being the charge amp's output, shows a good representation of a real world half sine pulse.  The trace below it is a stored output of the KZQ crystal when it's loaded down by the scope's input without the benefit of the charge amp.

The cursor arrows indicate the time window where the signal can be integrated to quantify the pellets energy after striking the impact absorber. Obviously the lower trace is not useful for this purpose. Note that neither of the traces show any signs of extraneous resonances or other interfering signals coming from the v2 KZQ!

It should also be mentioned that the stimulus for the test signal is being generated by one of my nifty modal hammers with the compliant tip installed to simulate the front impact absorber.

[George Schmermund]

The testing and improvements on the KZQ continues and has now progressed to the v2.5 level of development. Each incremental step in the measurements allows a commensurate reduction in machining complexity. This will all lead to a more reliable and reproducible device.

In the meantime I thought that it might be useful and possibly interesting to review the reason for attempting to design the KZQ in the first place. I understand that there are different levels of consciousness involved in the process of hunting. My personal goal within this activity is to humanely dispatch rodents that become problematic on our property. I don't hunt them as such, but if they interfere with the quality of my life, on our property, I'm going to aggressively interfere with theirs. I have no other quarry at the moment. Errant rabbits are captured in 'catch and release' traps and then set free in the local canyons for the coyotes.... just supporting the food chain.

That being said, I'll report that coyotes are starting to come up out of the canyons around our neighborhood and are becoming aggressive towards pets according to the neighbors. I haven't seen evidence of them on our property yet, but if I do I'll address their trespassing the same way as I deal with the rats.

Should this situation come to pass I'll need something more substantial than a CO2 powered gun to stand my ground with. This will most likely push me over the edge into the world of PCP armaments. The ability of the KZQ will be even more relevant in this case.


 

[WyoMan]

George, a word here if I may... don't try to dispatch coyotes with a pcp. Coyotes are wily, anxious, and don't generally allow you a good, close shot. Get a 30-06 if you can legally use it...

Wyo

[George Schmermund]

Thank you for the advice. I'll have to check with the city to see what level of protection I'm allowed to use in this situation.

[lloyd-ss]

George,
I am picking back up on this very educational thread and absorbing what I can. Your unique approaches, throughout this thread, are appreciated and admired.  I believe you are testing and documenting characteristics that have not been done before.

Regarding your latest creative challenge, the KZQ force/impulse transducer is quite intriguing. Are you satisfied with the calibration of the unit? You seem to really like solid, empirical data, and are not satisfied with "relative" measurements.
I truly believe his kind of brain challenging work keeps us young and vital.  You are going to be around forever. 

[lloyd-ss]

Hi George, I was reading about your thoughts on pressure transducers, their cost, size, responsiveness, etc. I built a few pressure transducers a while back but did not test the responsiveness of them. Your idea of a piston arrangement that you tap with a hammer is great and I think I will give that a try.
But if you are interested in trying to make one, and I see that you don't shy away from that, I started with Merit SMD series sensor dies. They came already mounted and compensated. Approx .28" square and .08" thick. Four wires must be soldered to them and then exited out through a pressure-proof seal. Magnet wire and epoxy worked well once the process was figured out. One of the sensors has 4 connector pins instead of wires. It was connected to an AD623 instrument amp and an A-D convertor. with a reference and excitation voltage. Here are pics (not that great) of 2 designs. One of the designs screws in like an SAE o-ring plug. The other has a 10-32 screw with a .050 dia hole thru it to bring in the air pressure.  Making them was probably more trouble than they were worth, but they actually did work well for the intended purpose.
In both designs, free dead space was held to a minimum.

[WhatUPSbox?]

Lloyd,
I found a spec sheet on their site for one of the other configurations, using the same technology. https://meritsensor.com/assets/documents/pdf/BP-series.pdf It mentioned a frequency response of 1200 hz. That may not be fast enough for a breech sensor. Have you tried just tapping the sensor housing to see if there is a response rise time?

Found an interesting, slightly vintage, discussion on the topic. https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19700010011.pdf Their dynamic calibrator is interesting, I think it would work well with George's accelerometer calibrator.

Thanks for the idea

[George Schmermund]

It's nice to see that Lloyd and Stan are keeping track of things here in this wilderness of airgun experiments! Stan's search juju is right up there with the best I've seen.

Lloyd's pressure sensor build is interesting. Characterizing its frequency response and other specs would be useful and important depending on it intended application.

The KZQ calibration is moving along at my usual distracted pace, but progress is being made. The current effort is to determine the uniformity and reproducibility of its impact surface. I've been playing with various methods of pellet impact simulation using the B&K 2515 Vibration Analyzer and the 4 ch DSO to evaluate these impact signals simultaneously. All of my B&K instruments are vintage, but they work well and still calibrate to the original specs. In general, accelerometer, force, and pressure transducers have very relaxed specs compared to audio (lunacy) requirements. I take comfort in knowing this when I use these hallowed instruments to make measurements with.

[WhatUPSbox?]

George,
Are you going to integrate the force response by using the math within the DSO? I have not tried that yet with mine.

Lloyd,
I also found an interesting paper on optical pressure sensors. The paper is about biomed applications at relatively low pressures, but the approach should be applicable to higher pressures and would yield tiny sensors
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4541926/ Section 3

[George Schmermund]

Stan - The testing that I'm doing at the moment uses the serendipitous confluence of the clean half sine wave output of the charge amplifier impulse response and the DSO's collective measurement screen when the measurement button is pressed. No need to get tangled up in the math functions section of the scope. The DSO's measurement panel has an area calculator that can use just the positive half of a single sine wave (it doesn't know that's what I'm giving it) and calculates the area (integral). The signal from the charge amp is also going to the 2515 which then measures the frequency spectrum of the same half sine wave. The numbers are remarkably close for being this early in the game!

[George Schmermund]

It might be of interest and some help to those following the KZQ project to see the general scheme that is being used for these measurements. The images are from the Endevco website and show the basics of modal impact testing signals. One image demonstrates the typical output of an instrumented hammer when measured in the time domain as would be seen on the DSO. The other image is what the same signal looks like when it's measured in the frequency domain on the vibration analyzer.

The overlaid plots in each image are results from the different tips that can be interchanged on the hammer. As can be seen the tips can determine a different range of outputs. The KZQ is in essence an instrumented hammer operated in a reverse mode where the pellet is generating the force to be measured. I'm probably preaching to the choir, but it may be helpful to some folks out there.

[George Schmermund]

A few posts back Lloyd asked if I was satisfied with the calibration of the KZQ. I have to say that the question is really jumping the gun (pun) at this point. There is still some amount of thought that has to be given as to how to best represent the POI (point of impact) metric and what method(s) give the most reliable numbers.

Generally, all of the 'chronys' that are popular with airgunners today report the projectile's velocity in ft/sec and energy in ft-lbs. The measurements are usually done at or near the muzzle and therefore can be comparatively interesting, but not overly useful at any distance beyond the muzzle.

The point of designing and building the KZQ is to enable shooters to accurately measure the amount of muzzle energy a projectile still has at the POI. To accomplish this in a meaningful way the projectile's mass and velocity at the POI must be converted to units of momentum (a directed force) and timed as the velocity drops to zero. This time to zero measurement then becomes an impulse. The impulse can then be integrated (the area under the half sine wave curve). This new unit is reported in lb-sec.

[lloyd-ss]

A few posts back Lloyd asked if I was satisfied with the calibration of the KZQ. I have to say that the question is really jumping the gun (pun) at this point. There is still some amount of thought that has to be given as to how to best represent the POI (point of impact) metric and what method(s) give the most reliable numbers.

Generally, all of the 'chronys' that are popular with airgunners today report the projectile's velocity in ft/sec and energy in ft-lbs. The measurements are usually done at or near the muzzle and therefore can be comparatively interesting, but not overly useful at any distance beyond the muzzle.

The point of designing and building the KZQ is to enable shooters to accurately measure the amount of muzzle energy a projectile still has at the POI. To accomplish this in a meaningful way the projectile's mass and velocity at the POI must be converted to units of momentum (a directed force) and timed as the velocity drops to zero. This time to zero measurement then becomes an impulse. The impulse can then be integrated (the area under the half sine wave curve). This new unit is reported in lb-sec.

George, That is an interesting approach, and a bit confusing to me, because the concepts of kinetic energy and momentum seem similar but are actually different. So, you are thinking that you will know the velocity of the projectile as it hits the sensor and can therefore calculate the force that it has, and then see if the force-time integral that is measured by your system is the same as what the projectile arrives with?  I hope it sinks in eventually, LOL.
Lloyd

[George Schmermund]

Lloyd - It may be useful as a thought experiment to consider the KZQ as an electronic ballistic pendulum. The force transducer and DSO combine to measure and store the momentum (which is always conserved in any collision) as being the pellet's impulse signature.

The charge generated in the piezoelectric crystal converts the pellet's kinetic energy into potential energy minus any of the usual losses do to capturing the pellet.

There are, of coarse, some things that need to be worked out to standardize the measurements and hardware design to make this a reliable transducer, but it all seems to be coming together using the approach I've chosen at this stage of the game.

I'm open to any and all comments and testable critiques. 

[WhatUPSbox?]

George,
I think your second sentence may be causing a little confusion. The way I understand your planned measurement. The piezo sensor will provide you (after calibration) a signal that is proportional to the force on its surface. The time integral of that trace will be the impulse or the change in momentum of the pellet. Since you are capturing the pellet and not bouncing it off, it is also the momentum (mv) prior to impact. From that you can calculate the velocity and then the energy. There is nothing in the measurement approach that measures the energy of the pellet directly.

The charge generated in the piezoelectric crystal converts the pellet's kinetic energy into potential energy minus any of the usual losses do to capturing the pellet.

I'm looking forward to some of the data.

[George Schmermund]

Stan - Thanks for pointing out the over simplification of the measurement process in my previous post. You are correct that the sensor only provides a signal that is proportional to the force and is not a direct measurement of the pellet's energy. I think the electronic pendulum analogy should still be valid though.

[WhatUPSbox?]

Yeah, now that I know about them, I'm tempted to build a ballistic pendulum. I like the separate conservation conditions.
http://hyperphysics.phy-astr.gsu.edu/hbase/balpen.html

[George Schmermund]

Stan - I hope that you reify your ballistic pendulum temptation. Would you pursue a mechanical or electronic version? Either way it would be more than interesting to compare notes with you on any experimental results.