Hacking the CP-2

[George Schmermund]

Stan - Thanks for pointing out potential drawbacks to the heating of powerlets. Here's something more to add to the list. One of the most important, though often overlooked concepts in thermometry is this: 'A thermometer only reads it's own temperature'.

[Nvreloader]

Stan
Both the 150 & 160 have pressurized tubes, after the Co2 cylinder is punctured the gas fills the tube etc.

On the 150 there is a bare metal tube that the Co2 cyl fits into hanging out in the weather,
On the 160 it is the same, except the cylinder tube has a wooden stock (normal rifle configuration),
for some insulation from the weather etc.

The custom cap should be able to fit either air gun/pistol.
As far as the bulk fill, wouldn't that be a bigger problem, ie, a bigger Co2 tank to deal with in the cold weather,
not sure I am following along correctly with the "bulk fill" note?

George
Thank you for your input, and you'll not be held liable also.
The instrument I have reads temps via a multi meter gauge, with a "K" probe,
I have tested melted lead and have tested it on the wood stove for temps at various locations,
for setting the best location of the automatic off/on Piezo fan switch etc.

Thank you,
Don

[WhatUPSbox?]

Don,
Here is the train of thought on the heating approach and the bulk fill. Note, this is just brainstorming, no engineering calculations or drawings.
I was assuming bulk fill where you use the tube volume, not the external paintball bottle (QB79) approach (for the paintball bottle w/burst disk, I think you just bond on some kapton heaters, mount a couple of TCs, slip on a koozie, and tune the PID controller).

For the tube only type bulk fill, I was thinking that you could make a custom cap that includes a heater/TC probe that is internal and immersed in the CO₂. I thought it was easier since you have an unobstructed internal volume.

For the 160, the thought was that you could use the volume of the outer powerlet to have a more robust probe that also retained the inner powerlet. Fewer shots but hey if it is cold outside that may be OK.

For the 150, it may need to be just a longer, hollow cap that is heated and the cap has some internal porting to let the CO₂ circulate somewhat.

The battery and controller would need to get packaged. For the 160 either under the barrel or in the stock like George suggested. For the 150, it is harder. perhaps a LiPo battery under a custom grip.

Like I said, just a train of thought....plenty of Kaboom potential that would need to get addressed.

[George Schmermund]

The new KZQ design is working better than anticipated. The sensitivity is ~ 10X higher than the previous generation. Some unexpected signals have shown up in the measurements that arise from the much lower resonance  frequency of the transducer as a whole.

This much lower resonance has started me down another wayward path heading toward developing an interesting new way to build a short period seismometer. If the resonance could be reduced to ~ 1 HZ with a Q of .7 it may turn into another product. I have no idea of what the market might be, but I know that at least I want one!

The DSO images are from some tests I did today. The first one is of a pellet at ~ 465 fps that includes the half sine wave impulse along with the resonance sine wave. The new KZQ was attached to the back panel of the trap. The second one is of a double pellet shot with the KZQ detached from the back of trap. This is pellet stacking with one shot. The 2 pellet velocity was ~ 330 fps.

[WhatUPSbox?]

Seismometer...an interesting thought. Somewhere I have a geophone, they measure velocity. I wonder if I put on a Milk Dud target plate, and suspend it properly, the velocity peak should be analogous to a ballistic pendulum. 

[George Schmermund]

Your geophone might work if you can attach it to a large mass and very compliant spring. The qeophone would also have to be able to work in the horizontal position and move with a linear stroke. They're usually limited to only a couple of mm of displacement though. Damping is also an issue. If you used the modal hammer calibrator as a model for a short stroke linear displacement device it might work.

The project would certainly be educational and worth the effort in my opinion.

[WhatUPSbox?]

Educational - yes, starting geophone 101

I set up a quick configuration to see how the geophone reacts to an impact, especially in a horizontal configuration. The geophone is clamped behind the target board and a supersized milk dud (modelling clay) catches the pellet. The assembly weighs 226 g (3488 gn) and is roughly horizontal. The pellet is a 7.5 gn wadcutter, fired hand held, maybe a foot from the target.

The results were relatively consistent, about +/- 5 %. It takes about 500 micro-sec to get to the voltage peak (7.44V). I assume this is the path through the milk dud. The two traces show the short and long time traces for two shots.

I have not tried to calibrate this with a chrony yet but I saw somewhere that 20V/m/s is in the ballpark for the geophones. Using that number, the 7.44V and the mass ratio of 7.5 gn to 3488, calculates out to about 570 ft/sec. Probably a little high but a starting point.

Some thoughts: I was generous in sizing the milk dud (did I mention hand held and point blank?). It is not clear how much of the milk dud is accelerated in that first 500 micro-seconds  as opposed to getting stretched. Now that I know the geophone may be useful, Increasing the mass of the target plate (metal) and reducing the thickness of the milk dud should reduce this effect.

Doing some tap tests on the assembly and the geophone should be interesting, as well as running some tests in the vertical configuration.

Can't let George have all the fun.

[George Schmermund]

Stan - It's good to see that you're back in harness at your ballistics bench. I'm looking forward to seeing what you find using your geophone.

[WhatUPSbox?]

I did a little more work with the geophone as a sensor. I made a target block out of some cast aluminum. For initial characterization, I used a pendulum mounted weight to repeatably strike the target. There is also a laser light gate to measure the speed of the weight just prior to impact.

What worked: The light gate and the calculated speed based on drop height were consistent with each other and repeatable....The law of gravity is safe (kudos Sir Isaac). With modelling clay at the impact, the geophone output (peak voltage) for the pendulum weight strikes was repeatable and consistent with what a reasonable sensitivity might be for the geophone. The pendulum weight had the same momentum as a 7.4gn pellet at 370 ft/s so it is a reasonable test.

What didn't work (yet): If you don't use the modelling clay at the strike point of the weight, or for actual pellet strikes, There is a higher frequency in the output that makes getting a good peak value unreliable. This frequency can be moved around some but is present if the target experiences a hard impact (the pellets cut through the clay) and a resonance somewhere in the system is excited. The options are a little constrained because as you soften the impact point you slow the response to where the geophone first mode starts affecting the peak voltage.

Not quite ready to give up on the geophone, there are still some mounting options to try as well as tuning the impact material. As a side interest, I may try using the light gate to measure the target velocity. This wasn't the original objective but diversions are not unheard of in this thread. Also looking forward to George's piezo results.

[WhatUPSbox?]

Finally got some hopeful results with the geophone. I tried a number of mounting and testing configurations trying to get rid of the high frequency content that is present in pellet shot data that you see in post 466. I could get rid of the content in the relatively low velocity impacts of the pendulum weight (post 468), but the high velocity impacts of the pellet excited the higher frequency, no matter how I configured the clay buffer or the Geophone mount. I thought it was the assembly but then I tapped just the geophone itself and it does have an output mode in the 2 KHz range (note vertical and horizontal orientations behave similarly).

I switched to using thin LDPE bags as the breaking medium for the pellet (something Lloyd used in his high velocity tests). The target configuration is in the photo below. The PVC tube just provides containment for several plastic bags. The clay is there to primarily hold the PVC. The aluminum block provides the mass and the geophone (brass color) is clamped to the back. I also attached my microphone block to the back to provide a shock indicator. Everything is softly suspended.

The second image shows output that still has the high frequency content. This was from a shot that still had a bunch of clay inside the tube and less plastic bag so the pellet still hit the clay. The third image shows the final output without the high frequency content. I had taken out almost all of the clay and filled the tube with plastic bags and the pellet stops in the plastic and does not hit a hard surface. The hope was that the pellet deceleration would be slow enough so it does not excite the higher modes but is still well above the geophone first mode around 30 Hz. I could have used the O-scope low pass filter function but since I want to capture the voltage peak right after impact, I did not want to modify the waveform.

Next steps would be to do a number of shots to see if the results are consistent and set up the chrony to get a calibration of the sensitivity.

[George Schmermund]

Your experiments are looking interesting. Are you using a damping shunt resistor across the geophone's output? It would be instructive and useful to fully characterize the device without any additional mass and spring components.

[WhatUPSbox?]

George, I  did some simple tap tests of just the geophone softly suspended. I did include my microphone block (double sided tape mount) just to confirm any higher frequency features.

The geophone output is channel 1 one in each case. The first plot is vertical, the second is horizontal. It was meant to be a quick and dirty confirmation that there is high frequency content in the geophone itself.

It showed that there is higher frequency content to be avoided. I had also tried a saddle mount for the geophone and that had the content as well. This is why I went after softening the impact.

The geophone is marked 400 ohm and measures out at 387

[George Schmermund]

The suspension appears to be rubber bands which would add another spring into the system. Supporting the geophone on some very soft foam would be closer to a free-free condition.

The shunt resistor can be used to load the geophone's coils and reduce or eliminate overshoot and ringing. A damping factor (DF) of .7 is about ideal and can be found experimentally or from the manufactures specs.

The other detector adds mass to the geophone for the free-free measurement, but could still be helpful in the overall scheme of things at this point. In theory the housing and magnet of the geophone are anchored to the earth and follow the earth's movement while the spring suspended coil assembly, due to inertia, remains in a quasi-stationary position with the coils generating a velocity following signal. https://shop.raspberryshake.org/product/geophone-earth-sensor-cut-open/

I'm sure that you're well aware of all this stuff, but it might be helpful info to anyone trying to follow our continuing follies.  It looks like you're making good progress. Please continue to post your experiments.

[WhatUPSbox?]

Thanks George,

I used the rubber band suspension because it had a fairly specific suspension mode in the 1-2 Hz range and I was looking for KHz features. I'm a little cautious about loading the geophone circuit. Since I'm looking for peak amplitude somewhere around a msec in, I want to avoid filtering in that timescale.

I took the configuration shown in the last post and did some shots for repeatability. I used a P17 which puts out around 400 ft/s with a 7.5 gr wadcutter. The PVC snout is stuffed with 3 LDPE (newspaper) bags and each pellet is removed between shots.
For a string of 10 shots, I got a peak voltage average of 2.86 V and a standard deviation of 3.3% The first image below shows a typical clean trace. There were a few in the string that had some ripple on them.
I then took one of the bags out and got the second image which shows that the pellet tapped the block (I could see a small indent in the clay). If I'm reading the trace right, I think it shows the higher resonance being excited and affecting the waveform before the basic signal has reached its peak. That is what I'm trying to avoid. I think I'll use a longer PVC snout so I can get some margin in my baggie brake.

I think the next step is to set up the chrony and see if this all comes up with the right answers. I'll use one of the crosman pistols that has a power adjuster.

The geophone I'm using, I bought surplus long time ago (it has a Reagan era date stamp). I just ordered a couple more surplus units to see if they behave the same. I had forgotten how much fun they were to play with.

[George Schmermund]

The original KZQ device that morphed into a seismometer boondoggle as a substitute for the geophone front end on the Raspberry Shake instrument has taken another turn. The educational market won't have any money to spend on new lab course instruments anytime soon so I'm back to the POI airgun devise.

The KZQ in it's previous incarnation was being designed to measure the POI energy as an impulse number. This was going to introduce a new metric that would probably face considerable resistance in my targeted(?) market for the device. I've since come up with a compact dual diode array scheme that would bring the metrics back to a conventional set of numbers similar to what a generic chrony would provide. I'll report on progress as (if?) it progresses.

[George Schmermund]

Here are a couple of photos that show how the diode arrays are going to be assembled. They will be mounted in a pellet proof housing and finally mounted into a trap configuration about the same as the original KZQ system.

[WhatUPSbox?]

Interesting, looking forward to the design

[RScott]

Wow.  I've just spent the better part of two days reading with great interest the exploits of George and his peer reviewer / collaborator Stan.  I am blown away by their deep technical knowledge, intense curiosity, and ability to tee up a question and work it.

When the barrel harmonics work morphed in the KZQ quest, I became concerned that we will not see much more insight on the harmonics beyond Replies #366-369.

I learned a huge amount through that point - enough for me to conclude that the hammer-valve interactions and pellet movement dynamics are likely the biggest drivers of vibrations and barrel swings.  And that damping vibrations might be a better strategy than timing the vertical barrel swing to improve pellet direction consistency.  And enough to get me interested in putting sensors on the muzzles and receivers of my PCPs to see if I can correlate the x-y patterns and shot groupings, and then test ways to tame them.  (I've already used audio techniques to determine their shot development times, and maybe can adapt that equipment for motion sensing)

So a big thanks for all your efforts even though a summary of findings and recommended approaches for future barrel harmonics hobbyists might never appear.
I hope George and Stan are well and enjoying their latest exploits.

[AmBraCol]

It was a bit of a shock to get the notice of a reply on this thread, hadn't seen it or thought of it in some time, life's been too busy.  I hope you share your own journey and experiences with us, there is still a lot to learn.

[WhatUPSbox?]

RScott, I'm glad you enjoyed the thread.
I took a brief look at muzzle motion on the last page of this thread https://www.gatewaytoairguns.org/GTA/index.php?topic=176445.220
I was looking at hold sensitivity but harmonics would be part of the motion.