HOWTO: Cheap Chronometer

[Ragnarok]

Supersonic results.
Not too shabby.

BECAUSE OF THE WAY IT WAS DONE!

I said CAN, not WILL - please, read what I've said. The way they did it, although slightly harder, is the only appropriate (but still not ideal) way to do supersonic sound-chronying. Your suggested method would not be appropriate.
The other reason I say they're inappropriate is that a sonic crack is just more garbage on the sound recording to wade through.

Also, that bullet is supersonic under the scientific definition, so there's less drag issues.
Most of the time when most people here (including me) say supersonic, they don't actually mean it. They mean "exceeding Mach 1", compared to the definition of supersonic as "all airflow over the object is in excess of the SOS".

Generally, to be supersonic, you need to be closer to Mach ~1.2.

With most projectiles, when people say they got them "supersonic" they're actually transonic, a horrible, horrible region for drag, where some airflow is subsonic, some supersonic, and a greater fraction of velocity is lost over much shorter distances, even compared to the considerable drag of supersonic flight.

Although simple, there are many cases where sound-chronies are simply not appropriate. Most of our near Mach 1 needs are amongst those.

[boyntonstu]

Supersonic results.
Not too shabby.

BECAUSE OF THE WAY IT WAS DONE!

I said CAN, not WILL - please, read what I've said. The way they did it, although slightly harder, is the only appropriate (but still not ideal) way to do supersonic sound-chronying. Your suggested method would not be appropriate.
The other reason I say they're inappropriate is that a sonic crack is just more garbage on the sound recording to wade through.

Also, that bullet is supersonic under the scientific definition, so there's less drag issues.
Most of the time when most people here (including me) say supersonic, they don't actually mean it. They mean "exceeding Mach 1", compared to the definition of supersonic as "all airflow over the object is in excess of the SOS".

Generally, to be supersonic, you need to be closer to Mach ~1.2.

With most projectiles, when people say they got them "supersonic" they're actually transonic, a horrible, horrible region for drag, where some airflow is subsonic, some supersonic, and a greater fraction of velocity is lost over much shorter distances, even compared to the considerable drag of supersonic flight.

Although simple, there are many cases where sound-chronies are simply not appropriate. Most of our near Mach 1 needs are amongst those.

Are you saying that 2684/2763 = 0.97 is not accurate enough for the average user here?

I seriously doubt it.


BoyntonStu

[Technician1002]

The average velocity of the bullet over the distance of 50 yards is then
Vbullet = d / tbullet = 150 ft/ (0.0559 s) = 2684 ft/s.

BoyntonStu

Just for grins.. Take your time to travel 150 feet.. 0.0559 seconds and invert it. To detect the point in time, you need accuracy several orders of magnitude greater than the time interval. Otherwise, the analysis is much like aligning a couple piles of trash dumped on a curb to find just how many mm they are from each other. The sound event is too low in bandwidth to obtain accurate interval time. Two bangs arrive to the mic. The waveform from the first is still in peak amplitude when the second arrives, much weaker. S/N of the second is lost in the noise of the first.

Math is easy, Posted for those new.. Frequency = 1/time. Even CD quality recording with an upper BW limit of 20 Khz does not provide enough samples to measure this.

In my case hearing a marshmallow land softly somewhere near the target doesn't record well.

Here is the actual recording of one of my air cannon discharges. Note the waveform at 0.05 seconds after the start.

[jimmy101]

Why don't you use the sound card and a mic as the chrono?

A bang and a thud.


Analyze the audio.

Done!


(You could place the mic midway between the muzzle and the target to automatically compensate for the speed of sound.

Not a perfect measurement, not exactly muzzle velocity, but good enough?

BoyntonStu

Yep, done it that way as well. Works OK, and is very easy to do. Indeed, my son used that technique to measure the muzzle velocity of a gauss rifle (and got the velocity of each stage in the rifle) for a science fair project.

For me, the big practical problem with sound is the often complex patterns caused by echoes, the gun resonating after launch etc. But it works well enough that the paintball community sometimes uses it (google SoftChrono). As others have posted, there are also issues if you are near the SOS.

The PT setup is so cheap and easy, and the data is much easier to analyze, that's why I prefer it. $2 for 2 PTs, some scrap PVC pipe, sacrafice a crappy old pair of headhones for the cable and submini stereo plug. The MIC input on the PC supplies the power for the PTs. A freeware software package, like Audacity, to do the recording and display the wave form.

[Ragnarok]

Are you saying that 2684/2763 = 0.97 is not accurate enough for the average user here?
I seriously doubt it.

Trust me on this one. As far as aerodynamics and drag are concerned, although I'm far from a world expert, it is clear from your misunderstandings that I know the subject in substantially more detail than you do.
You shouldn't take any offence at that however, because I've studied the subject to an epically obsessive level.
Although I don't recommend trying to compete with my ballistics knowledge, I would still suggest that you read up on the basics.

The important fact that you're missing is that drag is different for different things. Bullets have high ballistic coefficients, they hold velocity well. Most spudding projectiles have low ballistic coefficients, they do NOT hold velocity well.

Let's take a "macro" variant of a sound chrony - a guy with super stopwatch skills, and two markers 100 feet apart. You are at marker A with a golf ball and a ping pong ball. You will now throw each to marker B.
You take the golf ball, hurl it, and about 3 seconds later, it passes marker B. Average Speed is ~33 fps.
You take the ping pong ball, and because it's lighter, hurl it much faster than the golf ball - but because of drag, it slows down so much it hits the ground long before marker B. It never reaches B, and average speed is recorded by stopwatch man as 0 fps.

Like the difference between golf balls and ping pong balls, bullets have much less drag than spudding projectiles. While the difference is 0.97 for a bullet over 150 feet, for other less perfect projectiles (read: almost all spudding projectiles), it will be much greater. There are some projectiles where the difference would be 0.5 or worse.

With the right ballistics knowledge, you could make an estimate at this factor, and get back to roughly the right result.
While I'm perfectly capable of this myself, I wouldn't wager on most of the forum being able to do it particularly easily (yet, at least. )

[jimmy101]

Just for grins.. Take your time to travel 150 feet.. 0.0559 seconds and invert it. To detect the point in time, you need accuracy several orders of magnitude greater than the time interval.

I must not be following you. 1/0.0559 sec = 17.9Hz. Throw in a factor of 100 in frequency to get to a 1% sample error and you get 1.79KHz sample rate. Typical sound card has another order of magnitude resolution on top of that. A sound card shouldn't have any problem with the basic time frame. Dampend sine wave original signal, mutliple signals, echoes etc. may be a problem.

Two bangs arrive to the mic. The waveform from the first is still in peak amplitude when the second arrives, much weaker. S/N of the second is lost in the noise of the first.

Your phasing doesn't sound right. Say the gun resonates at 300Hz (WAG), time to first zero crossing is 1/300/2=1.7mS. But the impact times is at 55.9mS. Of course, it depends on where the mic is and the speed of the ammo.

At Mach 2 with a mic halfway from muzzle to target, muzzle blast gets to mic at time X, sound from round hitting target gets to mic at time 2X, where X is (muzzle to MIC distance)/(SOS). I think you still have 55.9mS from muzzle blast to "ding". The sound card shouldn't have any problem measuring a duration that long. Again, still have problems with digging a weak dampened since wave (ding) out of a strong dampened sine wave (bang).

Put one mic at the muzzle and on mic at the target. Or, a mic at the muzzle and tape a piezo transducer to the target (might be able to get away without any amp for the piezo). Sound cards in PC (but usually not laptops) have stereo left and right line inputs. Since the mics are well separated, and sound intensity drops off as 1/r², it should be possible to mostly eliminate the unwanted sound from the two channels. Probably need a preamp for the mics to get them to work with a line input.


So it can be done. A supersonic marshmallow is kind of unusual. Heck, any supersonic shot is going to be more difficult.

[boyntonstu]

Just for grins.. Take your time to travel 150 feet.. 0.0559 seconds and invert it. To detect the point in time, you need accuracy several orders of magnitude greater than the time interval.

I must not be following you. 1/0.0559 sec = 17.9Hz. Throw in a factor of 100 in frequency to get to a 1% sample error and you get 1.79KHz sample rate. Typical sound card has another order of magnitude resolution on top of that. A sound card shouldn't have any problem with the basic time frame. Dampend sine wave original signal, mutliple signals, echoes etc. may be a problem.

Two bangs arrive to the mic. The waveform from the first is still in peak amplitude when the second arrives, much weaker. S/N of the second is lost in the noise of the first.

Your phasing doesn't sound right. Say the gun resonates at 300Hz (WAG), time to first zero crossing is 1/300/2=1.7mS. But the impact times is at 55.9mS. Of course, it depends on where the mic is and the speed of the ammo.

At Mach 2 with a mic halfway from muzzle to target, muzzle blast gets to mic at time X, sound from round hitting target gets to mic at time 2X, where X is (muzzle to MIC distance)/(SOS). I think you still have 55.9mS from muzzle blast to "ding". The sound card shouldn't have any problem measuring a duration that long. Again, still have problems with digging a weak dampened since wave (ding) out of a strong dampened sine wave (bang).

Put one mic at the muzzle and on mic at the target. Or, a mic at the muzzle and tape a piezo transducer to the target (might be able to get away without any amp for the piezo). Sound cards in PC (but usually not laptops) have stereo left and right line inputs. Since the mics are well separated, and sound intensity drops off as 1/r², it should be possible to mostly eliminate the unwanted sound from the two channels. Probably need a preamp for the mics to get them to work with a line input.


So it can be done. A supersonic marshmallow is kind of unusual. Heck, any supersonic shot is going to be more difficult.

Good post.

Take a look at the graph here:

http://arxiv.org/ftp/physics/papers/0601/0601102.pdf

It the target hit time is very easy to see.

The noise after the projectile leaves the muzzle is easy to ignore.

BoyntonStu

P.S. Can you suggest a method to take the .pdf graph and convert it to an image for uploading?

[jimmy101]

P.S. Can you suggest a method to take the .pdf graph and convert it to an image for uploading?

Like the attached?
On a PC;
Print screen button (Prt Scrn?), or Alt Print Screen (for just the active window)
Open MS Paint and paste in the screen shot
If you want, copy just the part of interest and paste into a new paint window.
Save the file as a gif.

Looking at the other images in that link reminds why I don't use a mic. You pretty much have to know what the answer is in order to interpret the soundgrams.

[boyntonstu]

P.S. Can you suggest a method to take the .pdf graph and convert it to an image for uploading?

Like the attached?
On a PC;
Print screen button (Prt Scrn?), or Alt Print Screen (for just the active window)
Open MS Paint and paste in the screen shot
If you want, copy just the part of interest and paste into a new paint window.
Save the file as a gif.

Looking at the other images in that link reminds why I don't use a mic. You pretty much have to know what the answer is in order to interpret the soundgrams.

Thanks,

The positive going vertical spike lines of the muzzle and the bullet strike are very similar.

If you know the approximate muzzle velocity you can look at the ballpark region for the strike signal.

The other spikes are not believable.

(One can use a .22 at 1050 fps as a calibration)


To each their own, but 97% accuracy is good enough for me and the price is right.


BTW People used cumbersome ballistic pendulums to measure energy and to calculate velocity.


BoyntonStu

P.S. I really like this forum.

FYI and possible enjoyment:

See my $100 elevator video here:

[Technician1002]

Well done guys. Works for lead bullets. Thanks for the PDF.
Proof is better than theory any day.

On the photo, any OS with screen capture and any good photo editor.
I grabbed it and did it in Ubuntu using the Gimp before reading all the posts. Result is below.