How can I use a pressure sensor to detect drum hits for my lighting rig?

[hardyboy16jm]

Im trying to build this lighting rig, so that every time I hit the bass drum of my drum set, the rig can light up. So I need a way to sense if I hit the drum, and then send an electrical signal to a microcontroller so the lights can light up (I don't need help with the lights, just the sensor). I asked my physics teacher and he said use a pressure sensor. He kind of suggested this on the fly, and now its Christmas break and I can't talk to him about it. So if you can elaborate on how to use a pressure sensor to sense if I hit the drum or suggest a different way to sense the drum hit, I'd greatly appreciate it! Also please give as detailed as explanations as you can! Thanks!**

 

[Simon Bridge]

Microswitch on the drum pedal.

glue a small mirror to the drum surface and bounce a laser beam off it and on to a light sensor - when the drum vibrates the reflection flicks on and off the sensor.

Per the pressure sensor - when the bass drum sounds, it makes a pressure pulse.
So does everything that makes a noise - but the drum will probably have the highest peak pressure (you will need to experiment to find out - but you can hear the bass beat over everything else right?)

 

[Chestermiller]

Hire a roadie.

 

[Simon Bridge]

Coat the drum with a triboluminescent material?

 

[berkeman]

Microswitch on the drum pedal.

Best idea of the bunch, IMO.

 

[Simon Bridge]

Microswitch on the drum pedal.

Best idea of the bunch, IMO.

It's how it's done in cars ... I usually annoy people with these suggestions.
Everyone else seem to think up the sort of stuff that would have Rube Goldberg cry out: "Stop, it's too complicated!"

How about:
Small child inside the drum with a knife switch.
Check local child labor laws.

 

[Pythagorean]

Im trying to build this lighting rig, so that every time I hit the bass drum of my drum set, the rig can light up. So I need a way to sense if I hit the drum, and then send an electrical signal to a microcontroller so the lights can light up (I don't need help with the lights, just the sensor). I asked my physics teacher and he said use a pressure sensor. He kind of suggested this on the fly, and now its Christmas break and I can't talk to him about it. So if you can elaborate on how to use a pressure sensor to sense if I hit the drum or suggest a different way to sense the drum hit, I'd greatly appreciate it! Also please give as detailed as explanations as you can! Thanks!**

It depends on what you want. If you want the intensity of the hit to correlate with the intensity of the light, take your professor's advice. You would have to find a pressure sensor that doesn't interfere with (absorb) too much of the kick of the pedal though, and you would have to tune the assembly to the pressure ranges you will be providing.

If you just want it to be a binary yes-no process (hit or not hit), Simon suggested the easiest route.

 

[DennisN]

Putting a switch on a pedal is a good and simple solution, but it need to be cleverly done to minimize the wear and tear from a frenetic drummer. I would probably try this solution myself.

Otherwise, it could probably be done with a microphone connected to an equalizer, i.e. a low-pass filter set to a cutoff somewhere between 50 and 100 Hz, or a bandpass filter set to between 50 and 100 Hz. The output from the filter will of course probably need to be amplified and perhaps made into a pulse by some circuit (depending on what input the light rig requires).

There's also the risk that a bass guitar will interfere, since the frequency ranges are ca:

• Bass guitar 30-500 Hz
• Bass drum low range ca 50-100 Hz

In this case, some clever microphone placement/soundproofing/attenuation may be necessary. Low frequency sounds are more difficult to attenuate than higher frequencies. This construction would probably be fun to do, but it requires more work and testing.

Otherwise, there are (and you can google for it):

• "kick drum trigger"
• "acoustic trigger pedal"

and you can also check out
http://www.drummagazine.com/plugged-in/post/proven-tips-for-triggering-kicks/
http://en.wikipedia.org/wiki/Trigger_%28drums%29
http://recordinghacks.com/2010/09/07/bass-drum-microphones/

 

[Pythagorean]

I have no idea of the forces present, but I envisioned something like this, simply fixed to the drum where the striker hits:

http://www.trossenrobotics.com/store/p/6496-1-5-Inch-Force-Sensing-Resistor-FSR.aspx

It's pressure range is 1.5 to 150 PSI, and my gut tells me that the hardest striker hits are going to be bigger than that, quieter hits may be somewhere in the range. Another consideration is whether it will interfere with the sound quality.

 

[dlgoff]

It depends on what you want. If you want the intensity of the hit to correlate with the intensity of the light, take your professor's advice.

Exactly. I would use an existing or install a new Air Vent Grommet, perhaps like this one:

http://www.drumfactorydirect.com/bmz_cache/4/47667936106e9578c9d7d81f74970ee0.image.250x250.jpg

and then insert a Pressure Transducer, perhaps like this one:

You will need to determine which transducer (pressure range) to use. Then the fun begins; interfacing it to your micro-controller and scaling it's output (maybe with an op-amp).

 

[Pythagorean]

Good call. That would remove any detectable interference with sound quality that might have come from a pressure patch.

 

[hardyboy16jm]

It's how it's done in cars ... I usually annoy people with these suggestions.
Everyone else seem to think up the sort of stuff that would have Rube Goldberg cry out: "Stop, it's too complicated!"

How about:
Small child inside the drum with a knife switch.
Check local child labor laws.

I like the child idea, it's probably the best idea I've heard so far. But just in case... how can I use
the microswitch so I don't break the switch in any way? I like to hit the pedal pretty hard, and there are points where I'm playing at peak speed and control where I end up hitting the pedal even harder, so I can't be breaking the thing mid-song. Thanks!

***Also I could be mistaken on how the switch works, so please correct me if I'm wrong: it's just a matter of pushing the lever or whatever you'd call it (probably actually called the switch), right?

 

[Simon Bridge]

If you really like to pound that switch - then you may be better off just building your own industrial strength one.
You just need two wires and something for the contacts.

You press a pedal to hit the bass right?
I'm thinking two wires touching when the pedal is depressed - the advantage is that this is flexable so if you slam down hard you only bend the wires and the thing springs back.

A simple switch is cheap so you can play around for your particular kit.

 

[DennisN]

This is a fun project IMO, something I could very well have tried myself (but I haven't - I don't play drums, I use drum machines and sampled drums ). It's good with pictures to go along with the discussion concerning the construction, so, a kick drum pedal:

http://drumsetguru.com/wp-content/uploads/2009/06/DW_5000.jpg

and more can be found with a google picture search for "kick drum pedals".

I've have actually got two more ideas (pretty easy I think) for a solution, but I've got to think them through for a while, so I'll be back.

EDIT: When I looked closer on the picture, it seems the horizontal axle (colored red) might be a good place to place a switch, that is, a switch between the axle and the casing for the axle. The motion of the axle should be less dramatic than motions elsewhere, methinks...

 

[DennisN]

Ok, here comes my two other ideas, one with medium difficulty and one very, very simple:

1. Medium simplicity:
A photodiode (perhaps partially enclosed) and a LED placed beside the pedal. Attach, let's say, a metal bar to the pedal itself, which will be blocking the space between the photodiode and LED when the pedal reaches its bottom position.

2. Very, very simple (actually somewhat embarrassingly simple , but I love simplicity):

Attach a string to the horizontal axle and spin it (is this the right word in English?) a couple of times around the axle. At the other end of the string, attach a conductor A (some metal thingy). At a proper distance below conductor A, place a conductor B (another metal thingy), which is fixed, i.e. attached to the floor/pedal bottom/axle casing or whatever. Attach the conductors to your circuit/light rig. When the axle turns, conductor A will be lowered to connect with B, and then raised again. This is embarrassingly simple, but it should be very durable and should be easy to make fail-safe.

EDIT 1: Concerning the string thing (2), this construction could be enclosed in a fixed plastic tube to eliminate/minimize any unwanted horizontal motion.

EDIT 2:
Update on the string thing: Three conductors are probably even better. Conductor A (on the string) is electrically disconnected in the normal position. At the bottom of the plastic tube, two conductors, B and C, are placed, which are connected to the light rig. When the axle turns, A gets lowered and connects B to C. This will mean no moving cables, only one moving conductor. So it should be even more fail-safe.

 

[Simon Bridge]

Photogate will trigger twice for each drum-hit.
It's simple enough to get only every second hit to do anything.

In English: you can "wrap" a string around an axle ... but the proper word is one of those annoying ones where the same spelling has several meanings and pronunciations. The proper word is "wind" - pronounced to rhyme with "kind". The past tense is "wound" - to rhyme with round. So you want to wind the string around the axle.

You are right - a string can be used to pull something ... variation on the mechanical switch.
Same problem - hitting the drum hard could stress the string to breaking point: there's a reason why the standard pedal uses a chain.

Even simpler - the pedal itself can be turned into the switch.
The components can be fixed to the pedal and made as tough as you like.
I'd experiment with a microswitch by the tensioner myself.

 

[dawin]

I'm going have to agree with others that Simon's suggestion of just turning your pedal into a switch will be the simplest, and probably most robust, way of doing this.

Unless of course, like Pythagorean mentioned, you want some intensity-related effect. Hope you're prepared to tinker with that one for a little while before it works just right. (Nothing's quick, nothing's easy!)

 

[DennisN]

Photogate will trigger twice for each drum-hit.
It's simple enough to get only every second hit to do anything.

Yes, it will either be unblocked->blocked->unblocked (if the LED gets completely blocked at max kick position)
or
unblocked->blocked->unblocked->(reverse motion)->blocked->unblocked (if the "blocker" just passes the LED during a kick)

In English: you can "wrap" a string around an axle ... but the proper word is one of those annoying ones where the same spelling has several meanings and pronunciations. The proper word is "wind" - pronounced to rhyme with "kind". The past tense is "wound" - to rhyme with round. So you want to wind the string around the axle.

Thanks! I started with "wound" but I got uncertain about it .

You are right - a string can be used to pull something ... variation on the mechanical switch.
Same problem - hitting the drum hard could stress the string to breaking point: there's a reason why the standard pedal uses a chain.

Yes, that's true. When I now think about the string thing in the long run, the "wear and tear" will most likely have an impact. It might be possible to do with good material however. Something for the experimentalist, I suppose .

Update on my photogate suggestion:

I think a photogate positioned to trigger motion of the axle would be better than a bar connected to the pedal. Here's what I mean (detail in blue rectangle):

Pros 1: Will be very durable and fail-safe if done correctly.
Pros 2: With the right components and design I think one might even be able to make it respond to intensity (more kick intensity means faster motion which means faster blocking). Something for the experimentalist again, I suppose.
Cons: Requires electronics knowledge.
Notes: The construction will need to be powered.

 

[hardyboy16jm]

I just realized sort of a problem with some of the suggestions (obviously not any ones fault but mine for forgetting to mention this)
When drummers play, the pedal doesn't always go back to starting point since we keep our foot on the pedal. i also sometimes try to keep the beat by slightly moving my foot which makes the pedal move, but not necessarily hit the drum. There are also times, where I actually keep the beater of the pedal touching to drum.
I also sometimes hit the pedal then immediately leave the pedal to go to a different pedal or something like that where it would make the pedal sway back and forth.
So now that I think of it, the photogate idea and switch idea might not work. So the sensor can't really rely on any pedal movement (unless anyone has another idea). But the way I see, the sensor needs to rely on the actual hit of the drum

 

[Simon Bridge]

The drum only sounds if the pedal is pressed beyond a certain point though - so the switch/photogate is positioned at, or just past, that point. The small size of the switch will help you there.

This is one of those cases where it helps to be physically present.

 

[DennisN]

I just realized sort of a problem with some of the suggestions [...] (snip)

I see no problem with what you wrote if you do a correct photogate construction and correct positioning (except one, see * below).

The drum only sounds if the pedal is pressed beyond a certain point though - so the switch/photogate is positioned at, or just past, that point. The small size of the switch will help you there.

This is one of those cases where it helps to be physically present.

Exactly. I said:

Yes, it will either be unblocked->blocked->unblocked (if the LED gets completely blocked at max kick position)
or
unblocked->blocked->unblocked->(reverse motion)->blocked->unblocked (if the "blocker" just passes the LED during a kick)

So, what I meant was: position the "blocker" (moving bar) and the photogate so the blocker reaches and blocks the gate when the axle reaches its "max kick position" (i.e. when the beater touches the skin).

If you are uncertain about the idea, or maybe feel it would be too hard, you could try Simons original idea. Some kind of mechanical switch somewhere. The same principle will apply here; the objective is to close the circuit when the beater touches the skin (or just before).

In this case, I would probably try something positioned at the axle/axle casing, like two conductors A and B, where A is fixed to the casing and B is moving with the axle, quite similar to the photogate suggestion. Two metal "thingys", one firm (A) and one flexible (B), perhaps?

(*)

There are also times, where I actually keep the beater of the pedal touching to drum.

You mean you keep the pedal pressed down with the beater touching the skin? If so, this might be a problem with a simple gate/switch construction, since the gate/switch will be closed at this point. It could be solved by

1) repositiong the gate/blocker so the gate closes somewhere just before "max kick position".

and/or

2) An electronics circuit can solve this by sending one short pulse only when the gate/switch closes. Your beat will have been completely digitalized .

 

[vk6kro]

You can avoid switch damage by having it at the pedal rest position. So it is compressed when the pedal is NOT pushed.

To do this you need to get a normally closed, single pole double throw switch.
You compress it by taking your foot off the pedal so this opens the switch. Hit it and the switch closes.

Then it doesn't matter how hard the pedal is pushed, because the switch is nowhere near the pedal.

However, aren't most drum hits achieved with drumsticks?
If you wanted the lights to react to those, you might consider using microphones or getting an output from the microphones that are already there.

There are plenty of circuits for converting music to a light display.

 

[AlephZero]

This is one of those cases where it helps to be physically present.

This might help, as a reality check on some of the more "theoretical" suggestions.

https://www.youtube.com/watch?v=3lkZPJES45Q

 

[Simon Bridge]

Kinda helps. I also know a drummer so I can get a close look at his kit ... it's just that I'm also not getting paid ;)
I still think a mechanical switch will work in the situation in the vid...

the normally-closed approach is not so hot as the pedal is not always released to the same position - a photo-gate beam running close to he drum head may work - if the drummer consistently pulls the mallet-head back beyond the lip of the drum (which is the closest you'll get without drilling holes) - but one could be placed to be blocked by the mallet-arm close to the axle.

I can see several places a small switch could be fastened that would get the effect wanted (working off the hammer axle for instance). I'd need to get closer to check the variation in how far the mallet swings for different strokes, but mounting the switch on and with a shock absorber seems the straight-forward approach.

In the vid setup, you'd want two :)

pressure sensor or small child inside the drum would also work... could just use a mic with filtering.

I notice the drum surface moves quite a bit so the small-mirror glues by the edge with a laser bouncing off it would work too.

 

[AlephZero]

IMO by far the simplest way to do this is the same way you make a recording of a kick drum: put a microphone inside the drum. A cheap mike insert stuck to the inside of the drum shell should work fine.

You need a mike insert that will survive very high sound levels, but to pick up the start of each "hit" you don't need good low frequency response. The initial transient sound energy covers a surprisingly large frequency range, up to several kHz. the "obvious" idea of using a low pass filter to cut out unwanted noises might not be the best way to do it.

 

[Simon Bridge]

If it's inside the drum, the loudest thing around should be the drum right?

 

[DennisN]

EQ warning from me:

1. The rest of the drum set WILL produce sounds of all frequencies which WILL reach the microphone. One obvious problem will be the low tom (floor tom) which is similar to the bass drum. But fear not , the rest of the drum set is also very loud, which will be problematic. E.g. cymbals tend to ring "forever".

2. Also, during a live session there WILL be other sounds reaching the microphone, mostly from other amplifiers, PA and monitors. Bass sounds will likely interfere with the dedicated microphone (see my post 8).

This is an extremely common problem during recording sessions (it's usually called "microphone bleed"): http://en.wikipedia.org/wiki/Spill_%28audio%29

So, I say either try something according to the KISS principle, or be prepared to face some serious design challenges.

EDIT: I will post frequency spectrums in a while, since I've got real drum samples on my computer which I will process, so I'll be back soon...

 

[DennisN]

I did a quick spectrum analysis of a bass drum, floor tom and a drum loop with a fill (snare, mid tom and low tom in a live room, i.e. many drum hits):

(Sorry about the colors, I didn't find a way to change them ).

When you look at the spectrum, don't pay too much attention to the differences between the drums in dB, since the drums are not from the same drum set or same recording session. However, if you study the spectrum you understand what problems I am talking about:

Bass drum (blue): the first peak is between 50-100 Hz + more peaks elsewhere.
Floor tom (green): the first peak is at 100 Hz + more peaks elsewhere.
Drum fill (orange): frequencies all over the place.

And, obviously, an amplified bass guitar will also produce peaks all over the place from 30 Hz and up.

 

[DennisN]

This might be an interesting solution:

http://en.wikipedia.org/wiki/Piezoelectric_pickup#Piezoelectric_pickups

See here:

Piezo Pickup in a Bass Drum as a Synth Trigger
https://www.youtube.com/watch?v=_q_75kZTdQg

I haven't got any experience of piezo pickups myself, though, but it might be worth trying, methinks...

 

[DennisN]

By the way, I also found this clip via the wiki page I mentioned before (http://en.wikipedia.org/wiki/Trigger_%28drums%29).

It should be helpful to the original poster, I think.

What are drum triggers? (a demonstration of commercial drum triggers)
https://www.youtube.com/watch?v=dGOFmVNiSH0

 

[vk6kro]

I just realized sort of a problem with some of the suggestions (obviously not any ones fault but mine for forgetting to mention this)
When drummers play, the pedal doesn't always go back to starting point since we keep our foot on the pedal. i also sometimes try to keep the beat by slightly moving my foot which makes the pedal move, but not necessarily hit the drum. There are also times, where I actually keep the beater of the pedal touching to drum.
I also sometimes hit the pedal then immediately leave the pedal to go to a different pedal or something like that where it would make the pedal sway back and forth.
So now that I think of it, the photogate idea and switch idea might not work. So the sensor can't really rely on any pedal movement (unless anyone has another idea). But the way I see, the sensor needs to rely on the actual hit of the drum

Seems like you could use an accelerometer on the pedal shaft.
The actual position would then not matter but a large shift in acceleration in any direction could light the light.

These are available as small printed board modules for a few dollars on EBay.

Possibly a bit of overkill, though. You would need a micro to read the accelerometer.

 

[HowlerMonkey]

drum trigger that inputs to a midi device.

Once the data is in the midi realm, there should already be devices that will actuate switches from a midi datastream.

 

[Baluncore]

I would attach a small electret microphone to the frame of the drum, follow it with a high pass filter and a threshold detector. That would detect the step impulse from a drum hit but not the lower amplitude drum resonance or other sounds present in the local environment.