PDA

View Full Version : Sensor for arduino-based bullet impact counter


CHS
06-03-2010, 2:28 PM
I'm working on a project right now that's arduino-based for counting bullet impacts on cardboard and using that information to develop an interactive target system.

My biggest hurdle at the moment is finding a sensor that's cheap enough, and reliable enough that it can be taped to a piece of cardboard and actually measure the bullet impacts as they hit and then pass through the cardboard target.

Initially, I was using a Piezo speaker element as a knock/force sensor, but it was extremely unreliable, only counting the impacts 2/3 of the time. I know it's not the Arduino's clock rate, because we hooked it up to an O-Scope and got about the same results. I mean, it's a speaker, so it's not exactly designed for working in the opposite direction.

I'm looking for new ideas of sensors to use. I'm thinking about a little electret microphone, but I'm wondering if that will be TOO sensitive, and record the concussions before the bullet impacts.

There are also piezo vibration sensors. I'm wondering if those might be more reliable than re-purposing a piezo speaker.

I'd like to keep the sensor cost at or under $5 since it WILL take a bullet eventually, and needs to be cheap enough to replace to make the whole project worth it.

So... Ideas/suggestions?

Dr Rockso
06-03-2010, 2:41 PM
I've used this style before with good results. Might want to give it a try.

http://www.sparkfun.com/commerce/product_info.php?products_id=9199

Also saw a LabView based video target system with some (presumably more expensive) accelerometers attached to drywall and armored against accidental hits. Video was projected onto a piece of drywall and bullet impact locations were triangulated from accelerometers.

CHS
06-03-2010, 2:45 PM
I've used this style before with good results. Might want to give it a try.

http://www.sparkfun.com/commerce/product_info.php?products_id=9199


Yeah, those are the piezo vibration sensors I was talking about. Digikey has them without the mass/weight and in a larger size for like $2.40/each.

I just wasn't sure if they would be a better choice and more reliable than what I have been using. I was leaning towards picking one up regardless. Maybe I'll order some from Digikey.


Also saw a LabView based video target system with some (presumably more expensive) accelerometers attached to drywall and armored against accidental hits. Video was projected onto a piece of drywall and bullet impact locations were triangulated from accelerometers.

Yeah, I really wanted to do that, but I don't think the Arduino has a fast enough clock cycle to actually do that kind of triangulation. Would be so awesome though.

Unfortunately, I am in NO WAY qualified to write a triangulation routine. Heh.

stan
06-07-2010, 7:49 PM
dude you're doing arduino work? badarse. i have a duemilanove sitting on my desk feeling lonely.


what about using something tougher for your target? seems like cardboard is so soft that it wont produce a significant vibration, compared to something like steel. think about how much energy the bullet imparts to the cardboard (very little, the bullet hardly slows down), and now you're trying to sense some of that energy after it is transmitted across the cardboard (crappy transmitter).

if nothing else, maybe you could use something at least more rigid like thin MDF. what about that?

regardless, i think increasing the signal is going to be a lot better than trying to make a better sensor. your signal:noise ratio is probably a nontrivial issue here.

gadjeep
06-07-2010, 7:52 PM
1/4" maple plywood would make a good substrate. Cheap, light, easy to work with and holds up better than mdf, drywall or cardboard.

stan
06-07-2010, 7:54 PM
1/4" maple plywood would make a good substrate. Cheap, light, easy to work with and holds up better than mdf, drywall or cardboard.

the MDF i'm thinking of is what ikea uses for drawer bottoms and dresser backs. its like the thickness of cardboard, but is MDF. would probably be a lot less work to cut than maple ply.

JDay
06-07-2010, 9:11 PM
the MDF i'm thinking of is what ikea uses for drawer bottoms and dresser backs. its like the thickness of cardboard, but is MDF. would probably be a lot less work to cut than maple ply.

Isn't all the Ikea wood made out of compressed paper? I remember reading an article about this.

CHS
06-08-2010, 8:37 AM
dude you're doing arduino work? badarse. i have a duemilanove sitting on my desk feeling lonely.


Yup! These things are great!

I'm doing the dev work on a Duemilanove. I've also got a Mega standing by. Once I get this stupid thing running properly then I'll load the code onto an RBBB and build a permanent enclosure.... Or five.


what about using something tougher for your target? seems like cardboard is so soft that it wont produce a significant vibration, compared to something like steel. think about how much energy the bullet imparts to the cardboard (very little, the bullet hardly slows down), and now you're trying to sense some of that energy after it is transmitted across the cardboard (crappy transmitter).

if nothing else, maybe you could use something at least more rigid like thin MDF. what about that?

regardless, i think increasing the signal is going to be a lot better than trying to make a better sensor. your signal:noise ratio is probably a nontrivial issue here.

Well, the idea here is to keep things as cheap as possible. One of the advantages of cardboard is that it doesn't compress very well, so it should be pretty good at transmitting shock. If I have to go with something even harder like a cheap 1/8" or so MDF, I'll probably do that.

I've got a couple bridge rectifiers on the way, and some proper piezo film vibration sensors. I had realized after all my testing that I completely forgot that a piezo will output +-90v. I've got a nice diode in my protection circuit, so I'm pretty much just cutting out ALL negative voltage. The bridge rectifier should take care of that.

I've also got a TI datasheet here with some example circuits using opamps that I need to look over.

I'm going to be working on this a lot more tonight up at my buddy's hackerspace.

The frustrating thing is, I've got code that WORKS. I'm just pretty sure it's the electronics side of things that's being flaky.

Hopefully some more gear gets here today so I can really work on it tonight.

stan
06-08-2010, 10:18 PM
Well, the idea here is to keep things as cheap as possible. One of the advantages of cardboard is that it doesn't compress very well, so it should be pretty good at transmitting shock. If I have to go with something even harder like a cheap 1/8" or so MDF, I'll probably do that.






i am no expert, and this is just my intuition:

you want the most rigid thing possible. it is going to be nearly impossible to get any significant energy to be imparted to the target without getting thick and heavy. so you need to rely on rigidity in order to transmit the small amount of energy as effectively as possible.

stan
06-08-2010, 10:25 PM
also this just came to me - what about electronic drum triggers?

http://www.activatedspace.com/blog/wp-content/uploads/2006/11/ddrum.jpg

they're designed to do EXACTLY what you need. pick up short impulse from a surface.

ripcurlksm
06-09-2010, 2:54 PM
so you want to have a target that you can digitally track bullet impact and show virtual groupings? does the bullet have to pass through the target?

What about a steel plate with sensors on the back? No penetration and equipment is protected. Just duct tape an iPad onto the back of the steel plate and download the app for that and it will sense bullet impacts (lol)

CHS
06-09-2010, 3:45 PM
so you want to have a target that you can digitally track bullet impact and show virtual groupings? does the bullet have to pass through the target?


Unfortunately, due to the resolution of the Arduino, that's not going to happen. Though it's a really cool idea :)


What about a steel plate with sensors on the back? No penetration and equipment is protected. Just duct tape an iPad onto the back of the steel plate and download the app for that and it will sense bullet impacts (lol)

The arduino and associated sensors are cheap disposable electronics. An ipad isn't :)

ripcurlksm
06-09-2010, 4:28 PM
ok well not sure what you are doing then, but was joking about the ipad, but serious about the steel plate with some sort of sensor on it, but again not sure what its being used for ... good luck!

Barabas
08-23-2010, 12:41 PM
Any progress? Another thread reminded me of your project. Thanks to GrizzlyGuy for the link back here.

CHS
08-30-2010, 8:18 PM
Any progress? Another thread reminded me of your project. Thanks to GrizzlyGuy for the link back here.

Yes! A little progress.

Haven't had a chance to work on it at all in the last month, but I picked up an accelerometer and a buddy of mine designed and milled a little board for it. I got it all soldered up and tested:

http://i335.photobucket.com/albums/m446/bdsmchs/2010-07-27213404.jpg

Now I just need to get back to testing and finalizing the code. But the accelerometer board definitely works. That's an ADXL335 3-axis.