Chronograph for bullets

[Rleo6965]

The proto board version was 10MHz and the perf board version is a high harmonic, I fail to understand how the capcitance of P board made this large difference.

If you see underside of proto pcb. there's a strip of etch for each ic pin connection. This also have capacitance that affect high frequency signals. So cut with blade that etch that was not needed to reduce capacitance of etch.

Also try to position crystal, capacitor and resistor osc component closed to pin of ic or use jumper as short as possible. Much better not to use ic socket for high frequency signals.

 

[CDRIVE]

153 MHz almost certainly means that the oscillator is not being controlled by the crystal. If it was, you'd be seeing 10MHz, 20MHz, 40 MHz, 80MHz, etc.

I'm surprised, but a little more capacitance between the gate's input and ground may help.

However I'm not going to profess any super powers when it comes to crystal oscillators.

I'm more surprised than you but not because of spurious oscillation. That's not that uncommon. Constructing a working 153MHz Osc. is very difficult to do if your were trying too! I believe this more likely to be a measurement error.

Chris

 

[wannabegeek]

I'm gonna bring my Perf Board circuit back to the scope and double check the freq not only using the automated measure but by checking the period. I'll also bring the entire counter project board to see how that affects things....

I'll take pictures of the traces at various time scales.

Also, I'm gonna bring the proto board version of the clock...

It occurred to me that the crystal used in the proto board version was not the same one used in the perf board version.

I'll likely end up soldering a new clock circuit together using Rleo's advice.

wbg

 

[wannabegeek]

Got some answers and some new questions....

My perf board circuit had a mistake in it's wiring...how'd that get there ?
It would never have worked correctly....

I fixed the mistake but haven't put it on the good scope yet. My funky scope shows a 10kHz signal coming from it....weird. I think I need to start over anyways....

Built the same circuit on my protoboard and checked it...it works great.
Same schematic except i changed the phase shift caps from 100pF to 68 pF.

I'm going to try and save a few parts off of the original perf board circuit and build another
one and test it...I forgot to take pics tho.

At least for now I have a working perf board model of my Counter-Timer !

I'm also needing to revisit testing the actual mechanical detectors. I'm not sure that the foil sandwhich will work reliably...but I hope so b/c it's easy to implement.

I spoke with a friend of mine getting his Phd in computational fluid mechanics and he agrees that how the air affects the velocity of the bullet is not obvious. His rough guess is that the velocity will fall off sharply at first, then get linear, like a squared system, where F_drag = k*v^2. That cold be all wrong though...bullets are shaped to form a stream line for oncoming air, but air is quite viscose so the other two forces could be high. They are namely, the drag force due to shear and the drag caused by the vortex induced turbulence ( these things spin fast !).

Once I get a working foil detector, I will start taking data with my .22 Cal air rifle and try to build a mathematical model.


Cheers to all,
wbg

 

[wannabegeek]

Hi all....still thinking about this project...my grad class in physics started....one year out of school and I never was a 'smart' kid in class...so I'm a little blown away with work and all..


However, I now have access to more equipment and a good electronics mentor !

My PCB version of the 10MHz clock isn't working but I did 'fix' it several times so perhaps I just need to start fresh. The proto board version works. I am suspicious that the 74HC AND gate needs a particular input signal. I tried to scope the output of the AND gate going to the counter input but (home funky scope) scope loaded the signal and I couldn't see it.

I tired putting a cap in parallel with the clock output and used the x10 probe setting and then I could see the signal. I suppose now I need to use a better scope. The scope at work is only 300 MHz so that still may not be enough to see if the signal is dirty. My electronics teacher suggested that I research the requirements of the 74HC gates b/c I may need to further clean up my clock signal for good performance.

I am pretty sure I'm not gonna be able to test my project with real bullets anytime soon...so I'm just gonna work with my pellet rifle in my basement for the rest of the design. I doubt it is going to reach sonic speeds, where the characteristics of the
forces changes....oh well.

So now, back to photo diodes...!

wbg

 

[CDRIVE]

The scope at work is only 300 MHz ........
wbg

OMG, how spoiled this generation is! ... Only!?!?!? ... There was a time when we would have killed for a 300MHz scope.

Chris

 

[wannabegeek]

Hey guys I'm back...i tried to take a grad level physics course while working full time...that was not the best idea I've had all year....anyways I decided to drop the class and focus of learning practical skills...

Back to the chrono....I've been thinking a lot about using LED's. I know it can be done and I've thought of a way to deal with the solid angel problem of detection. To collimate the light I would set the LED array into a channel. Likewise with the detectors.

But, really, this solution means a lot more work on the detector, more time delay issues and expensive LED detectors with a fast slew rate.

My original mechanical breaking detector is simple, it could work fine and I might actually being able to use it on a real live ammo public range with the approval of the gun club management.

Now there's still an issue of measuring the distance between the printed traces. A micrometer will be accurate enough for the initial position, but the bullets will likely deform the trace detectors before breaking them. With a distance on the order of 1 or 2 cm, which is the thickness of the cardboard target hanger, this can cause serious systematic error in the measurement.

Further more, I will need to CAD up a my simple trace pattern and have it manufactured which might be cost prohibitive.

I would like to think about the distance problem more and proceed with getting a CAD schematic of the trace design. The problem here is, I've never made a schematic with a CAD program before.

I have access to Windows 7 through Virtual Box on my Linux machine. I may be able to get a student copy of the appropriate CAD software or borrow it...but I'm really not sure where to start with this stage...I 'v seen some circuit printing companies have a free web based software.

Suggestions...ideas....?

thanks!
wbg

 

[wingnut]

This suggestion may already have been given, but in school we would regularly work out the theoretical speed of a bullet by firing it into a block of wood suspended from a string (like a pendulum)

Ek before = Ep after where Ek is kinetic energy of bullet (.5 x m x v x v)
and m is mass of bullet in kg and v is velocity in m/s.
Ep = potential energy or mgh where m is combined mass of bullet and block.
g = 9.8 and h is vertical height the block and bullet combined swing up to.

v(bullet) = square root of (2 x (mass bullet and block) x 9.8 x height)/ (mass bullet)

This works on the principle of conservation of energy, that all kinetic energy of the bullet goes into potential energy in raising bullet and block.

 

[wingnut]

My other suggestion would be this....

Set up a camera on a tripod with the shutter open and a dark background.
Set up a strobe light.
Fire the bullet, and if the strobe is fast enough, two (or more) images of the bullet will appear on the frame.
Measure the distance between the two images and divide by the time between each flash.

 

[duke37]

"This works on the principle of conservation of energy, that all kinetic energy of the bullet goes into potential energy in raising bullet and block. "

But energy will be lost as the bullet enters the block. Surely it should be done by conservation of momentum.

 

[wingnut]

Duke - I think you are right about using momentum

Using the law of conservation of momentum when bullet impales itself into block, combined with the law of conservation of energy as kinetic energy of bullet and block convert into vertical height, then...

V(bullet) = m(bullet & Block) x square root of (2 x 9.8 x vertical height) / m(bullet)

 

[BobK]

The only loss of energy would be as heat (from the friction of bullet entering the wood.). I have no idea how much that would effect the result. My gut feel is that it is small enough to ignore.

Bob

 

[duke37]

Bob

A bullet is intended to transfer its kinetic energy into the target so I would think it is large

Duke

 

[wannabegeek]

welcome to my mile long thread...

This project is more of a learning tool for me and the method of computing the velocity is the one I'm using.... A ballistic pendulum is certainly not a modern way to do that.

A friend suggested a clever idea that would requite a lot of calibration:
Use a solid state pressure sensor or crystal gyroscope to measure the Impulse
momentum transfer of a bullet striking a large block of iron.

My method is interesting because a person could potentially shoot through several thin 'detectors' and track the velocity of the projectile as a function of distance. Then one could spline between points or fit a function to the curve. Then statistical quantities could be extrapolated without firing thousands of shots. I think....I want to try anyways...

A note about energy transfer: Different bullets transfer different amounts of energy.
The energy of deformation, is very different in a HP vs a JHP or ball ammo.
Ball ammo typically has very little KE lost to deformation and HP has quite a bit. Of course HP and JHP don't always expand well and they tend to jam pistols.

 

[BobK]

Bob

A bullet is intended to transfer its kinetic energy into the target so I would think it is large

Duke

But still most of it is transferred as kinetic energy right? That is what the swinging wood block is measuring, how much kinetic energy was transferred to the wood block. And what I was saying was that some of the bullet's kinetic energy would become heat, and the rest would become kinetic energy in the wood block.

Bob

 

[CDRIVE]

. Of course HP and JHP don't always expand well and they tend to jam pistols.

From strictly a reliability standpoint; that's why I prefer wheel guns.

 

[wannabegeek]

my thoughts, which I need to consider carefully....I agree with duke and add the following...

That is what the swinging wood block is measuring, how much kinetic energy was transferred to the wood block.
Bob

We might think to use conservation of momentum and energy because energy is not conserved in this system. But if we were, then we'd start with two equations and have two unknowns. When using momentum conservation, we have to keep track of the vectors involved. We don't do that at all here. When I used momentum conservation combined with energy conservation, I always had a lot of information about the velocity and the directions of the particles.

Bullets transfer very little momentum b/c their mass is tiny. Instead they apply huge forces. The bullets KE is converted into gravitation potential in the pendulum mass.

So no, most of the energy in a bullet is not " transferred as kinetic energy", it is turned into force. I think I know what BobK means but we are getting sloppy with terms and I want to force myself back on track.

The bullet's energy becomes force when the pendulum decelerates it. The Work-Energy Theorem states that for a conservative system: integral{ Force (dot) Distance} = Kinetic Energy. Grav potential is a conservative quantity and allows us to use this theorem. When the force of the bullet does work on the pendulum's mass, the pendulum gains Grav potential.

For a hollow point bullet which is designed to expand , a portion of the force is taken to deform the bullet. That force is pointing into the bullet, not the target so that force can't be applied to the target.

This is a confusing topic and good to discuss. My friend's idea to use the solid state gyroscope is smart b/c it can measure the force directly.
If I ever get this working I'd like to try the solid state gyroscope idea.

 

[wannabegeek]

so I 'm a little bit wrong with my previous post...

We probably would use consevation of momentum with the pendulum b/c the mass of the pendulum is not too much greater than the bullet, so that it actually moves.

For an inelastic collision of a bullet at v_1 and a stationary pendulum mass, m2:
m_1*v_1 = (m1 +m2)/v_1+2

let v_1+2 = v_final

v_final = (m1v1)/(m1 + m2)

sub that into the work energy theorem:
1/2 m v^2 = mgh

v1 = sqrt( 2(m1+m2)^2 g h (1/m1)^2 )

Now, since I'm not using that method, I still need to learn how to make my trace detectors...

 

[wannabegeek]

Started working with photodiodes, ordered 10 high intensity IR LED's and 10 side view photodiodes.

I still think that a breakable barrier is best for real bullets, but I wanted to work in my basement with my pellet rifle so I can try to numerically model the velocity vs. distance of the pellet.

I've been experimenting with some Photodiodes I got from Radio Shack, they are not real impressive but I've been able to get some ideas about how to use them.

So far, I've been thinking it might be best, to pulse the IR LED beam with a square wave, and then look for a missing pulse. I tried setting up a missing pulse detector with a 555 I had sitting around and couldn't get it to work properly. Choosing the right resistor and capacitor i think has been the sticking point.

I started with a 1kHz pulse and tried to get a hi and lo output when I blocked the IR beam.
As I understand it, I want the RC time constant to be a little bit larger than the period of the pulse, so RC = 1.3 ms for instance.
So that a lost pulse with allow the Capacitor to discharge fully before the next pulse comes in.