Detect rifle shot

There are electrochemical integrators that I've seen used to put an "hour meter" on phonograph pickups. Perhaps that sort of thing could be adapted to your needs.

I also have in my desk a piece of piezoelectric film that is wired to a neon bulb... flexing it lights the neon... takes no power quiescently.

...Jim Thompson

Could you keep track of how many cartridges you use?

-------------------------------------------------------------------------------- I am using the free version of SPAMfighter for private users. It has removed 602 spam emails to date. Paying users do not have this message in their emails. Try SPAMfighter for free now!

Robert,

Why not use a cheap pedometer? They've already solved the hard problems. If you want to get fancy, pull the electronics and embed them in the butt.

For heavy use on a range, this simple solution should to the trick.

If it's a casual-use gun (more toting around than shooting), the number of false "steps" will be significantly high - in that case, maybe there's a way to adjust the sensitivity on a pedometer, rather than reinvent it. Or, even simpler, just mark the ammo boxes for when to stop & clean (e.g., when you break open a new case).

Cheers, Richard

Interesting problem. Looking for a self-contained/non-invasive solution, with minimum modifications/interference of/with the weapon, is a good idea.

You'll need to determine whether or not false detections might be a problem, e.g. from dropping a weapon, or slamming the buttstock into something, or even vertical storage on a hard surface while in a moving vehicle, et al. If it can be made to not be a problem, by using a sensor that makes it easier to be sure you're counting actual shots, so much the better, and you'll basically just need to be looking for a threshold crossing, or a switch closure.

Using an accelerometer, if false detections might be a problem, you might want to also be able to somehow compare the acceleration's "signature" to a known average for the type of weapon (and ammunition) being used, to weed out false detections, especially if the system is to be active at all times. Maybe you could integrate the pulse, and/or differentiate the leading edge, etc etc, and compare some things to programmed values, and/or running averages, or something like that.

With a mechanical switch instead of an accelerometer, you would have some sort of mass, and spring and damping forces, in the mechanics of the switch (if it's kind of a "restricted pendulum", conceptually), to worry about. Maybe you could compare "switch closed time" to an average or a preset, to weed out false detections. You'd also have to watch out for switch bouncing. And the needed mass, and spring and damping forces, might vary considerably with type of weapon and ammunition. Probably should just use an accelerometer that already has appropriate dynamic range.

It might be "less messy" to use the sound that's transmitted through the body of the weapon.

Or... maybe there's a measurable EMP (ElectroMagnetic Pulse) that's generated when a shot is fired.

Or... since this is one-dimensional, maybe "roll your own" and get something like a slide potentiometer, put some springs and some mass on the slider, and read the deflection profile by sensing the current through the pot for a fixed voltage across it. But, in that case (especially, but all of the others, too), you might have to be careful to make everything light enough and fast enough, in order to make sure it would also work reliably for automatic or rapid-firing weapons. And remember that they could be oriented in any direction, relative to gravity, and, in general, might themselves be accelerating (but, we can hope at least, not at anything approaching the recoil's acceleration).

I think I'm still leaning toward favoring trying to use the internal sound/vibration of the shot, rather than trying to reliably sense the recoil motion.

Good luck!

Tom Gootee

The standard method, is a differentiator on a piezo sensor. If the circuitry is low frequency, battery life is not a problem (think how long a digital watch runs on a cell, and this is driving a display all the time as well). Systems doing this for some types of gun, have been around for over ten years.

Best Wishes

Sure. That's how it is currently being done. But the method has a few disadvantages, so an automated method is worth investigating.

A barrel typically lasts 2,000 shots. It is easy to lose count on the way, especially if you have more than one rifle.

Isn't there some way of measuring the wear directly, such as a micrometer?

-- Joe Legris

That's a good idea. I'll get one and crack it open to see if I can find something interesting. It will, of course, be way too sensitive if left unmodified, but there may be some useful information to find.

The device must be able to separate normal handling, such as transporting, cleaning and maintenance from actual shooting, so it must be somewhat more sophisticated.

It will not only be used to count number of shots in a day, but also how many shots on one barrel, or perhaps even on the rifle. Therefore, it should be able to last for months on one battery. That's why I am hoping to find a solution where the complete electronics can turn itself completely off, and store the count in some sort of non-volatile memory.

Thank you for the tips.

You can get piezo film fairly easily. It is a plastic film with a metal coating on the two surfaces. You can use conductive epoxy to make contact to it.

I think if you glue it right down onto the PCB, you will still get enough signal from it to turn on a small bipolar transistor. If not you will have to experiment a little on getting the recoil to trip it. Basically the idea is to use mechanical means to make a high pass filter so that normal handling doesn't fire it.

/ !---+---/\/\/--- GND ------+---OO----+---------------------- Counter +---------! / ! !/c -+---+---! NPN ! ! !\e / --- ! R1 \ ^ GND / ! ! GND GND

R1 sets the amount of shock needed to turn on the NPN. R2 is a modestly large resistor. It has to pass the leakage current of the NPN and keep the input of the CMOS NOR at a good solid high. When the circuit fires, it's power connection is taken low by the MML oneshot made for the NORs.

Although it looks at first glance like the connection marked "A" is not needed, I think it is a good idea because the path through R2 forms an RC delay. If the NPN only manages to pull its collector part way down, we want the circuit to quickly respond so noise doesn't get throught to the counter.

If the rifle is a single shot in that each trigger pull fires 1 shot, I would think and optocouple type arrangement to sense the trigger pull would work. Running its output to a oneshot gate arrangment then to the counter. If not a single shot, maybe the same arrangement on the bolt movement. Jtt

ANDing an accelerometer with a microswitch (somehow) connected to the trigger should be a very reliable way of detecting a shot. Instead of putting a m/switch in the trigger u cld possibly put it somewhere along the gun where it wld necessarily be pressed when the user grabs the gun ....

We've got a freebie pedometer that's been "on" (like a wristwatch is) for months, so I think they've solved your battery life problem. You just need to tackle the sensitivity problem.

Good luck!

Cheers, RIchard

Microswitch behind the trigger?

Your post is so typical of the type who knows a little bit of electronics and jumps right in with little to no knowledge of the underlying problem. Then you compound the idiocy by soliciting an equally clueless bunch of geeks who know even less. Rifle maintenance is far more complex than a mere counting of shots. Looks like the powers that be have settled on a barrel attachment:

I suggest acoustically. You could adapt a COTS glassbreak sensor for this application. There are wireless products of this type, the batteries on which typically last 2-3 years (this includes periodic status transmissions). The sensors themselves can be fairly small (even when encapsulated; one I'm "aware" of is approx 1.5x1.5x1.0" including battery.

If its aimed at you, its pretty obvious. ;-)

That's what I was thinking. I'd see if I could get one and glue it to the chamber in an unobtrusive spot. The idea is to get one that is relatively insensitive, so as not to worry about false signals, and have it respond to the sound conducted through the chamber.

That might get in the way when screwing on a silencer.

accelerometers require less then 1mA of power, the larger coin cells (25mm) are 1000mAh, so 1000 hours isn't so bad, don't know if the LCD could withstand the shock

Why do you want to count shots, when that's a really poor indicator of barrel condition?

From what I know there are three main reasons to stop using a given barrel:

(1) Bullets erode the inside of a barrel which eventually increases its internal diameter to the point it can no longer spin a bullet/shoot accurately.

(2) Shooters can get impatient and fail to allow cooldown time which can alter a barrel's crystalline structure enough to deform it permanently, also affecting accuracy.

(3) Each shot "cold works" the metal also altering its crystalline structure, just a lot slower. The structure eventually alters to the point it can no longer safely contain the pressure generated during a shot.

The first is noted with a micrometer as Joe Legris says (at the throat). The second visually and the last, x-ray crystallography maybe?

Are you talking rifles at a range that anyone might use, or your personal arms?

Do you know exactly what's going to be fired through them (each shot will affect a barrel differently even with ammo from the same production run from a known mfgr, much less "shoot what you brung")?

If you want to keep count just as a ballpark indicator of when you ought to check for problems that aren't visible during cleaning, ask the barrel manufacturer how many it's good for and subtract about ten percent given the variation in ammo properties and such things as weather, impatience, etc.

Mark L. Fergerson

PS I've heard "old wive's tales" about disassembling a rifle and ringing the barrel like a bell to see if it's recrystallized. Let us know if it works. ;>)