Audio/Sound Counter

Hi Everyone. Iv'e just registered and am looking for some ideas on counting sound pulses. The application I have involves counting detonation of explosives and displaying the total. The explosives would detonate around 5 mSec's apart and could be several hundred. Application is for mining and opencast blasting.

Hope someone has some ideas.

 

The regulars will give a better answer, but...
The method I would use is a microphone for recording only loud sounds, linked to counting IC's and the result to 3 single number LED displays.
I have a metronome that I built for my wife's piano playing (her request lol) that i'm about to ask a question on here about and it follows a similar method. An LM555N pulses to a 74LS138 and 74161 counting stage, with a bar of 8 LED's for the position count. Even simple circuits like that one can count incredibly quickly and are stable if built with care. Replace the 555 with the mic, the bar LED's with an answering machine type pair of 2 digit readouts.
The problem areas would seem to be the 5mSec slight gaps and the reliable counting because of that by using a microphone.
A vibrations sensor would seem to offer another solution, but the vibrations may become also muddied by those quick successive detonations.
If the circuit was connected to the actual firing box, then the pulses for triggering the explosions could be used a lot more cleanly.

 

Thank you very much for your suggestion. I shall certainly investigate this. The problem with tying into the firing box is that although the impulse is sent to the shock cord lines, that doesn't necessarily mean that the explosive has detonated, ie. if 100 charges are loaded and detonated, I need confirmation by count, for safety reasons. There may be a break somewhere in a shock cord line, causing a misfire on a particular charge, or the detonator may go, but the booster may not set off the charge. But this is food for thought.

I understand the 'LOL' bit. My wife is a concert cellist and plays for the Johannesburg Philharmonic Orchestra. She also uses a metronome for teaching, so I know what that's about.

Thanks once again

Bobby

 

Yes indeed, picking the right words for some suggestions can be the difference to seeming wonderfully supportive and being an ass! Examples of mixed messages being - "hey honey, lets convert the shed at the end of the garden to be a music practice room", or "that new CD you have for training is very good, would you like these decent headphones to listen to all the nuances, rather than have to have the volume on the speaker turned up ?"

Forward from the hopefully designed and built circuit you have in mind for the project, it could be imagined that a fired count and an actual count would be useful. Seeing how many detonations didn't detonate, by comparing readouts.
Thinking a little, is there a clear start and stop to the sound of each detonation ? I'm thinking that proximity and echo's might be large players for false counting. The threshold level of the count triggering could be troublesome.

 

One important signal to track is the firing impulse. There should always be a bang shortly afterward. If not, your squib did not detonate, and you will have that one identified by the number of the impulse.

 

With only 5mSec between blasts using one microphone you are going to get lots of overlapping sound pulses and wildly inaccurate counts.

Sound travels less that 2 meters in 5mSec also the sound pulse from each blast will get stretched and smeared travelling from the source to the microphone.

You might be able to do something with a widely spaced array of microphones and some clever signal processing but a simple set-up of one mic and a pulse counting circuit isn't going to work.

 

Thanks Alex

I agree with you totally, which is why I asked the question. I think I may
have to tie into the 'shot-exploder', and pick up the impulses. The only problem
with that is that, it may well initiate, but if there is a broken shock cord,
this would never reach the detonator and trigger, leaving behind, an unsafe
charge. The detonator, itself, may go, but maybe not the booster for the main
charge. I would prefer not to get into FFT, etc, but find a relatively simple solution, if possible.
I know it's not an easy one. Again, food for thought, and I thank you for your input.


Bobby

 

As an overview concept, how about this
(Again, i'm only familiar with what i've seen on TV for demolitions etc)

Each charge has a fuse type link, that link gives a connection to Ground. Each time a charge fires, the fuse link breaks and momentarily a connection is made to positive voltage, as the explosive goes off. The resulting pulse is then used as the counting stage pulse.
Only those that fire would create the count, those that don't would still have their links in place.

 

Cool project!

Sound is about 1mS per foot travelled. However if you are sure that there is always 5mS between pops it should be do-able enough. Normally the blast matrix is from the face across, then the next row behind the face etc so with the mike mounted at the front of the face at some distance (the obvious position usually used for cameras etc) there should be always 5mS+ between pops.

Unless you have walls around to give reflections? Typically reflections have much less high freq component so high pass filtering should get rid of most early reflections.

I'd like to hear how this turns out.

Theres some military apps out there that use a square of microphones to identify gunshots and then tell the distance and direction to the sniper based on the received sound. You might be able to google for that?

 

Darn double post glitch sorry.