Post by thread:[PIC]: Good Vibrations (excitations)

Piclisters: I have two problems: 1. Selecting a decent piezo speaker. I am using a 17mm piezo speaker from Murata. Piezo's are cheap, so I don't want to go to another technology. It's a fine cheap speaker ($0.35 to 0.50US in high volumes), but it's not very loud. I'm driving it with 9V instead of 5V, which helps a little. More voltage does not help above 9V, IMHO. Anybody got a louder piezo speaker in ther bag of tricks??? 2. Nice tones: (= Good Vibrations) I'm driving the speaker with three sets of tones, to indicate three conditions. OK, four tones to be exact. One brief burst of 25 cycles of 2000 hz to indicate a button press, a longer burst of 2000 hz to indicate condition A, a burst that goes in steps from 1000 to 1500 to 2000 hz to indicate condition B, and a long low tone at 600 hz to indicate the process is finished. Freq response of a piezo is not flat, and probably about 600 to 3000 hz at most with nasty peaks. I try for tones under 3000 hz to accomodate older people with high frequency hearing loss. I am just using Square waves to excite the speaker (= Excitations). Nobody likes my tones, they say they are discordant. I'm looking for a "nice" sounding sequence of tones. Don't need code, I can do that just fine, just need some experiences with what "sounds good" to users. (P.S. they all sound good to me, as the raucous din of the crying babe sounds good to the parent...) -- Lawrence Lile Sr. Project Engineer Salton inc. Toastmaster Div. 573-446-5661 Voice 573-446-5676 Fax -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

Lawrence Lile wrote: > > Piclisters: > > I have two problems: > > 1. Selecting a decent piezo speaker. I am using a 17mm piezo speaker from > Murata. Piezo's are cheap, so I don't want to go to another technology. > It's a fine cheap speaker ($0.35 to 0.50US in high volumes), but it's not > very loud. I'm driving it with 9V instead of 5V, which helps a little. > More voltage does not help above 9V, IMHO. Anybody got a louder piezo > speaker in ther bag of tricks??? The way to get louder sounds from a piezo transducer is to get more power into it. Increasing the drive voltage is one way. Choosing a piezo element with more capacitance is another. You may or may not be able to get a piezo with more capacitance in the same size case, however. Driving the piezo at its resonant frequency is another way to get much louder sounds. There are sometimes more than one resonant frequencies. One resonance will usually be much louder than the others. {Quote hidden}> > 2. Nice tones: (= Good Vibrations) I'm driving the speaker with three > sets of tones, to indicate three conditions. OK, four tones to be exact. > One brief burst of 25 cycles of 2000 hz to indicate a button press, a longer > burst of 2000 hz to indicate condition A, a burst that goes in steps from > 1000 to 1500 to 2000 hz to indicate condition B, and a long low tone at 600 > hz to indicate the process is finished. Freq response of a piezo is not > flat, and probably about 600 to 3000 hz at most with nasty peaks. I try for > tones under 3000 hz to accomodate older people with high frequency hearing > loss. Since the piezo transducer is very peaky in its frequency response, driving it with several different frequencies is a problem. You might be able to choose the frequency of your tones that match these peaks. A problem I see with this is that these peaks can vary significantly from transducer to transducer and from model to model. How the speaker is used also affects the loudness. The mounting can affect the resonances. Just like any speaker, the box it is in has its own frequency response and resonances. > I am just using Square waves to excite the speaker (= Excitations). Square waves are perceived as having a harsh sound. Try putting a resistor in series with the piezo. This resistor forms a low pass filter which will round the edges of the square wave. This will reduce the amplitude of the harmonics in the waveform. Removing these harmonics will reduce the power delivered to the speaker and therefore the loudness. Small piezo transducers typically have about 0.01uf of capacitance. With a 5K ohm resistor the -3db point will be about 3KHz. > > Nobody likes my tones, they say they are discordant. I'm looking for a > "nice" sounding sequence of tones. Don't need code, I can do that just > fine, just need some experiences with what "sounds good" to users. If more than one tone is sounded in sequence then they should be chosen to have "musical" frequency relationships. I'm no musician so I will not try to suggest what tone sets to use. Maybe try a little Bach? :-) -- Rich > > (P.S. they all sound good to me, as the raucous din of the crying babe > sounds good to the parent...) I have had similar complaints when using the piezo device for both making sound and as the timing capacitor in an oscillator. This created a mix of the resonance frequency of the piezo element and the frequency of the oscillator. This causes a pair of frequencies that are not a nice musical chord, they sound quite bad. {Quote hidden}> > -- Lawrence Lile > Sr. Project Engineer > Salton inc. Toastmaster Div. > 573-446-5661 Voice > 573-446-5676 Fax > > -- > http://www.piclist.com hint: The PICList is archived three different > ways. See http://www.piclist.com/#archives for details. -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

OK Lawrence my suggestions have been inserted in yr text > Piclisters: > > I have two problems: > > 1. Selecting a decent piezo speaker. I am using a 17mm piezo speaker from > Murata. Piezo's are cheap, so I don't want to go to another technology. > It's a fine cheap speaker ($0.35 to 0.50US in high volumes), but it's not > very loud. I'm driving it with 9V instead of 5V, which helps a little. > More voltage does not help above 9V, IMHO. Anybody got a louder piezo > speaker in ther bag of tricks??? Increase the efficiency. A Sounding Board never goes amiss. Think back to your science demos at school .... start a tuning fork vibrating, little sound, then hold against a surface, lots of sound. The equipment case can make a good sounding board (depending upon the material used to construct the case of course). Increase the drive voltage. Drive at the resonant frequency > > 2. Nice tones: (= Good Vibrations) I'm driving the speaker with three > sets of tones, to indicate three conditions. OK, four tones to be exact. > One brief burst of 25 cycles of 2000 hz to indicate a button press, a longer > burst of 2000 hz to indicate condition A, a burst that goes in steps from > 1000 to 1500 to 2000 hz to indicate condition B, and a long low tone at 600 > hz to indicate the process is finished. Freq response of a piezo is not > flat, and probably about 600 to 3000 hz at most with nasty peaks. I try for > tones under 3000 hz to accomodate older people with high frequency hearing > loss. I am just using Square waves to excite the speaker (= Excitations). > Using different frequencies means you are not going to operate at the resonant frequency and therefore not at peak efficiency. Why not just use 1 bleep, two bleeps, short bleep, long bleep etc etc all at the same frequency ? > Nobody likes my tones, they say they are discordant. I'm looking for a > "nice" sounding sequence of tones. Don't need code, I can do that just > fine, just need some experiences with what "sounds good" to users. > Probably because you are driving it with a square wave, all those harmonics. One possibility to improve this (and to increase the drive voltage at the same time) is to use an autotransformer. The piezo speaker is connected across the full winding and the drive voltage is applied to the tap Regards Chris -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

At 12:44 PM 11/20/00 -0600, you wrote: >Piclisters: > >I have two problems: > >1. Selecting a decent piezo speaker. I am using a 17mm piezo speaker from >Murata. Piezo's are cheap, so I don't want to go to another technology. >It's a fine cheap speaker ($0.35 to 0.50US in high volumes), but it's not >very loud. I'm driving it with 9V instead of 5V, which helps a little. >More voltage does not help above 9V, IMHO. Anybody got a louder piezo >speaker in ther bag of tricks??? I'd suggest an electromagnetic speaker, smaller (12mm diameter) and you can get 85dBA at 10cm with a couple of volts drive. Cost can beat the above depending on the exact requirements and quantity. Contact me via e-mail if you are interested. If you must use a piezo, are you driving it push-pull? That will greatly increase the loudness. {Quote hidden}>2. Nice tones: (= Good Vibrations) I'm driving the speaker with three >sets of tones, to indicate three conditions. OK, four tones to be exact. >One brief burst of 25 cycles of 2000 hz to indicate a button press, a longer >burst of 2000 hz to indicate condition A, a burst that goes in steps from >1000 to 1500 to 2000 hz to indicate condition B, and a long low tone at 600 >hz to indicate the process is finished. Freq response of a piezo is not >flat, and probably about 600 to 3000 hz at most with nasty peaks. I try for >tones under 3000 hz to accomodate older people with high frequency hearing >loss. I am just using Square waves to excite the speaker (= Excitations). > >Nobody likes my tones, they say they are discordant. I'm looking for a >"nice" sounding sequence of tones. Don't need code, I can do that just >fine, just need some experiences with what "sounds good" to users. Here is one I have used to indicate the end of a process. It sounds a little less harsh than this, in fact, this is a simulation. It is used to indicate the end of a process, so it is intended to be very noticable without being unnecessarily annoying. Because of the resonances you'll probably have to fiddle with a real speaker to get realistic results. http://www.speff.com/beeper1.wav Best regards, =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= Spehro Pefhany --"it's the network..." "The Journey is the reward" spam_OUTspeffTakeThisOuTinterlog.com Info for manufacturers: http://www.trexon.com Embedded software/hardware/analog Info for designers: http://www.speff.com Contributions invited->The AVR-gcc FAQ is at: http://www.bluecollarlinux.com =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

> If you must use a piezo, are you driving it push-pull? That will greatly > increase the loudness. Aha! I'm just using a common emmitter transistor with the piezo across the collector and ground. Only single ended. > Here is one I have used to indicate the end of a process. It sounds a > little less harsh than this, in fact, this is a simulation. It is used to > indicate > the end of a process, so it is intended to be very noticable without being > unnecessarily annoying. Because of the resonances you'll probably have > to fiddle with a real speaker to get realistic results. > > http://www.speff.com/beeper1.wav > Nice sound! > Best regards, > =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- = > Spehro Pefhany --"it's the network..." "The Journey is the reward" > .....speffKILLspam@spam@interlog.com Info for manufacturers: http://www.trexon.com > Embedded software/hardware/analog Info for designers: http://www.speff.com > Contributions invited->The AVR-gcc FAQ is at: http://www.bluecollarlinux.com > =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- = > -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

Lawrence Lile wrote: >it's not >very loud. I'm driving it with 9V instead of 5V, which helps a little. >More voltage does not help above 9V, IMHO. It certainly does, I had tiny piezo scream me half deaf at above 30vpp and resonating... but check with the maxum drive specified by manufacturer. However it might be expensive to get 30vpp. (I used a signal transformer and all gates in an 74AC04 connected 3+3 in parallel as an H bridge driver) If you do not already, drive the Piezo in bridge mode, i.e. between two outputs in antiphase, so two 0/5V outputs create 10V peak-peak to the piezo. amybe then you can afford a series resistor to soften the square wave. If the two pins are PIC outputs, you can soften the tone also by having an intermediate step where the two outputs is at same level. Also Piezos live longer if they have no DC voltage, so resting them at zero Volt is a good idea. I have not tested this, but adding a suitable inducor in series wil make an LC cirquit together with the Piezo. It will also soften the square wave drive a lot. If that resonant frequency is the same as the Piexo mechanical resonance then we can get one strong frequency. In your app it might be a idea to try locating the LC resonance between two mechanical/acoustic resonances to try falttening the curve. Maybe also use an R parallel to the L to flatten the curve. A trick I have observed somewhere is to connect one end of piezo to plus, and the other to collector of an NPN with emitter to minus, and now it comes: have an inductor parallel to the Piezo! When the transistor shuts off a large voltage puls is generated across the piezo. Oh, and put a R in series with it all or you will have a lot EMI on the supply. (Piezo is a cap) Also the audio resonance cirquit is also important. Normaly it consists of the air volume inside the plastic piezo casing, and the hole outlet. If you use a minimalist piezo consisting of just the metal plate and ceramic, then you should make some acoustic system like that. Regards /Morgan -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

IMHO, resonance is the way to go. Both with the LC parameters of the Piezo (it being the C of course) and with the cavity. By doing this you can get more volume than you ever wanted. I headed up the Invisible Fence Co. engineering department for about the last 8 years (no plug intended), and we were always looking for ways of increasing sound output of a piezo. However, by putting it in electrical and mechanical resonance was just way too loud. We also tried the inductive kickback trick which was loud but raspy. Resonance will give you a nice pure tone. If you want different frequencies you can lower the effective Q (either electronically or in the mechanics) or play with harmonics. Also remember to use bipolar drive. Easy way to implement is with the 4007 gate array. {Original Message removed}

>good sequence You really need to get one of those organ programs on the PC and find out what 'sounds good'. This mostly depends on who is listening. When in doubt, I formed an ad-hoc panel out of a few people who were around and asked them (by making 4 or 5 versions for each sound). I picked the one they liked best. It was not the one I liked best. I had no complaints for the final sound ;-). As a start, Try to make the spacing between sounds 2 steps (counted on 'white' piano keys). For 3 steps things get complicated. You have a lot of choices. One is a, a, a+8 or a, a, a-8 (where a is the base tone). A+8 is one octave higher (double frequency). Put a short break between each two tones. You want the third tone shorter than the previous two or longer than them (not equal). As a rule of thumb imho avoid tone counts that are divisible by 2 (like 2, 4, etc), unless they are arpeggios or glissandos (go look this up) or otherwise amplitude or frequency modulated. imho there are no cheap piezo speakers that do 600Hz. On the other hand mounting the piezo on a small resonant box will help a lot with loudness (and make the frequency characteristic even more uneven). Welcome to the art of music composition (which is a mystery to me), hope this helps, Peter -- http://www.piclist.com#nomail Going offline? Don't AutoReply us! email listservKILLspammitvma.mit.edu with SET PICList DIGEST in the body