Piezo, PWM, and MSP430G2553 Oh My...

[curtis63 5]

I have a custom board that is driven with an MSP4302553 chip and can drive a Piezo Transducer which resonates at 25khz.  I am limited to using a single rechargeable Lithium battery which is 3.3v.

 

In order to get the 117dB that I want out of the Piezo Transducer, I need to generate a square wave with a 25khz frequency and 10volts across the 2 Piezo pins.

 

I found a Push Pull library which allows me to get a -3.3v-3.3v range, or a total of 6.6v across the 2 pins on my Piezo.  This gives me about 80dB.

 

I need to boost that voltage by just a little bit after the PWM pins, without losing the 25khz frequency that I'm sending to the Piezo.

 

Does anybody have any ideas for doing this?

 

I'm on a tight budget and want to keep my board as tiny as possible with the fewest components possible.

 

Thanks,
Curtis

[zeke 693]

@@curtis63

 

You may want to consider using a voltage doubler circuit.

[Roboticus 1]

I'd suggest using a small toroid core and wind yourself a little 1::3 transformer, and drive that with a 3.3v square wave.  At 25kHz, you wouldn't need a beefy core.

 

Of course, I'm assuming that you don't actually require a square wave across the piezo element.  If you do, then I'd follow zeke's suggestion to build a small voltage doubler, then you can chop that with a BJT switch to get a pretty reasonable square wave.

[enl 227]

The info you provided is unclear and the numbers you give don't add up.

 

Do you need 10V peak-to-peak? 10V amplitude (20V peak-to-peak)? 10VRMS?

 

If 6.6Vp-p (+/-3.3v) is giving 80dB, then, as piezo devices are pretty linear for constant frequency, the output power will be pretty linear with voltage, whichever measurement is used. 6.6Vp-p square wave has a principal amplitude of (4/pi)*3.3V, and an RMS of about 3.0V. This is what is driving the resonance. If the device is spec'd at 10VRMS, then you should be getting 117dB+20log(3.0/10)=117-10.4dB=106.5dB (approximately 1/10 the power, which goes by the square of the voltage). If the device is spec'd peac amplitude square wave, sine amplitude, or p-p, you should be closer to the 117dB.

 

How did you measure the 80dB? Are you sure you are at resonance?

 

In general, due to the high Q value, driving one of these at resonance requires either careful tuning with a stable source, or a resonant driver, which is how the clock crystals are driven and commercial piezo alarms are driven. I would guess that you are driving off resonance (as the Q can be several hundred to several thousand, you need to be within a few cycles/sec) or you don't have a proper measurement of the actual output.

 

Also note that at 25KHz, you can do a LOT of damage to your ears, even though you can't hear it. Wear hearing protection if you are near the device when operating.

 

@@Roboticus a transformer would be fine here. If even remotely well made, it will reproduce the square wave well. In general, a sine wave if what you actually want, though. All power to drive the resonance. Power that doesn't do the work you want can lead to failure of the device, worst case. For a square wave, roughly 40% of the power is in the harmonics, 60% in the principal. Losing the harmonics is a benefit.