Circuit design help!

[cmangle]

I'm picking any engineer/small~circuit~design Ozzie's brain for some help!

I want to create two panels that house 10 piezoelectric transducers each that will act as an intrusion deterrent when an alarm condition exists.

These transducers operate at 2.5khz sine or square wave but need a signal generator circuit to supply that waveform to operate. I would like to just apply 12vdc to the lot but I think that would create "prawns on the barbie"!

Basically I have a 12vdc battery source that is solar charged(recharged) at a location where there is no electricity.

I believe all i need is a timer circuit with an SCR?? to handle the output, but that's as far as my design knowledge goes. Attached you will see a spec sheet for the transducers and a physical picture of the devices themselves. I got a deal on 90 of them.

Any thoughts suggestions designs ideas pointers?

thanks in advance, chris

[Reschs]

Silicon Chip did a project using Piezo Transducers a while back.

If you can find at least the year of the project I scan and send.

[Uncle Fester]

No scr just a transistor, but you need and inductor otherwise you won't even scare a fly away.

Can be as simple as this:

[Uncle Fester]

Ahh...forget that !

I just remembered I got something much better I designed myself



Read the bit about the inductor for your piezo bank. Depending on the quality of the inductor the voltage may exceed 30Vpp especially if you only load it with one piezo to test it. Using all 10 or 20 piezos will drop the voltage significantly. With quality I mean the physical size and design not the value. Those little ones that look like 1/2 watt resistors should be OK but you might want to use a few parallel to get more power on the whole bank. Good if you could measure the peak voltage.

Make Rf adjustable so you can tune it to the resonance frequency of the transducers so it really hurts the ears.

[Skepticist]

You'll need to generate the square wave to drive them and it sounds like 2.5kHz is their natural frequency. A plain square wave at constant frequency and amplitude would be a bit dull so you could modulate it a bit to make it as disturbing as possible which, I assume, is the goal. Just a couple of 555 timers and a driver transistor might be all you need IE a 2.5kHz oscillator gated by a slower oscillator through an open collector driver . Actual settings might take a bit of experimentation to get the best effect.

[Uncle Fester]

I have somewhere a more professional design that belonged to a project I did decades ago but did not go into production. I have full copyright. The output signal is exactly like Skepticist describes. It a single 4069 though not two 555s. There is a bit of experimentation required to get it right.
I have to dig deep to find it. So give me time till tomorrow. I have to go to bed now.
It has all the correct values for those 2.2kHz sounders.

[cmangle]

thanks nomeat and Skepticist it's been a LONG time since I've done any bread boarding!

I'm going to hate to admit it but I'll probably need some kid gloves here . . 8^(

[Uncle Fester]

You just have to whack a standard CMOS 4069 into the centre of your bread board and push the little caps, resistors and some bridges around it and you got a warbling very annoying alarm signal for a buck worth of parts (well at least where I got them from)






If you build the transistor output stage shown, please keep in mind that it was designed to run off 2 AA batteries.
I would just use the MOSFET output shown in my former circuit instead.

You can try the transistors (emitter of T7 is on GND) but R41 should be then 1k and the transistors should be able to handle 60-80V in case you run it without all the piezos attached but if you are careful you could get away with the good old 2N2222 becuase you don't want to get over 30Vpp anyway.


This circuit was insanely loud with a single Piezo and the purpose built transformer. They may be available on Ebay but a single inductor (or a few parallel) will work too.
C27 was added to hit the resonance peak. The sounder was in a little cavity with a lid that increased the resonance even further. You should experiment with C28.

Edit: Please keep in mind that this was part of a much larger design I made decades ago. Today I would just use a tiny $1 PIC micro and a MOSFET and I am done. ....and have all kinds of special features like alarm timeout, automatic enabling and disabling under certain conditions, battery low voltage detection, blah, blah

[cmangle]

No scr just a transistor, but you need and inductor otherwise you won't even scare a fly away.

Can be as simple as this:

what would the value of the inductor/coil be and I assume it would be variable/tunable?

[Uncle Fester]

Try 4.7mH and 10mH . They are not variable. They look like resistors and the colour coding is similar. So get a few to play with. Hope you got a cheap electronic store nearby. Two in series will double the inductivity, two parallel will halve it but it will increase the current they can handle. Finding a match for you piezo is experimental and I have no clue about using 20 piezo. All I know is you will need more current, that is why I suggested several in parallel to find a better match.
If you got a bit of lacquered copper wire you can have a go at winding it around a bit of ferrite or even a screw although the Q-factor might suck but experiments are fun.

[Uncle Fester]

BTW I think you might have to ground the connection between C and R on that circuit.
The two inverter (or gates) circuits I use are more reliable and you have plenty on a CMOS chip anyhow.
Make sure you use standard CMOS. They have a supply voltage range between 3-15V so you don't to worry about stable power supply. Place a 100nF ceramic cap right near the IC between your +12V and GND. Ground unused inputs on the 4069 or whatever chip you are using.
Just giving you the kid gloves you mentioned

[cmangle]

A friend suggested this circuit, but I believe as you said above it would produce a steady wave that is not too annoying. Can I add a second 555 (dual 555) to modulate it to a warbling output?

Two questions though, (1) Is it compatible with my piezo transducers and (2) will it drive 10 of those devices (current draw)?



thanks for all of your input

[cmangle]

As far as current requirements for the 20, I should have said that I would build two separate circuits for EACH of the panels!

I appreciate the clarity, it's been a long, LONG time since Electronic Lab class!

What do you think about the 555 circuit in the link I posted above?

[Uncle Fester]

The timer can handle up to 200mA so I reckon 10 piezo should be fine, but you can not place an inductor across it unless you use an open collector transistor as output.
Play with the timer if you want and see if it is loud enough for you with 10 piezos and no inductor.
Two timers in series, one around 5Hz the other 2.5khz, look up specs to calculate that.
You need to create a triangular wave (roughly) with an RC filter from the 5Hz timer that is connected to pin 5 of the 2.5kHz timer

[cmangle]

When adding the inductors/coils and the open collector xsistor how would they be added to the circuit above and which pins feed which pins on the dual 555 arrangement?

And an example of the values for the RC filter would be ?

Can you believe they closed the Radio Shack close to where I live?

I'm going online parts hunting later on this morning?

[Uncle Fester]

The output transistor is done like here:
Just replace with npn transistor, emitter to GND, base to pin 3 with the 3.9k resistor, collector to the sounders that have the coil parallel to them and then to +12V.
...but try first if the 11-12Vpp you should get out of the timer is loud enough for you. You can only safely double that with the coil because of he 30Vpp limit which is 6dB so about 93dB maybe.
They really only get loud if you make an acoustic resonance chamber for them, but you have 20, so this might be a different story.

Pin 3 is the output of the timer and is connected to pin 5 which is the control input of the other timer.
I would try a 100k resistor from pin 3 of the 5Hz timer to a 10µF electrolyte capacitor with it's negative to GND of course. Then from the junction of the 100k and positive of the 10µF connect with a variable 1M resistor in series to pin 5 of the 2.5kHz timer.
Now you have two adjustable pots, VR1 on your circuit for the centre frequency and this new added one we call VR2 for the modulation depth. However it will also influence the centre frequency.

First you should use perhaps 0.0047µF for C1 or even 0.01µF (instead of 0.001µF) for your 2.5kHz generator. The value in your circuit is way out of range. Then start with VR2 set at the middle and adjust VR1 until it sounds the loudest. Then adjust VR2 until it wobbles between an acceptable audible range and correct with VR1 again. There is a way to DC decouple the filter to slightly simplify the adjustment procedure but I would have to draw it up.

... or you could just try my original design in post 8#

...and then i just got another idea and maybe the simplest and best:

Seeing you are building this twice, why don't you just build a single timer for each panel and tune one a slight bit above and the other a bit below 2.5kHz.
You will probably never get them perfectly tuned any how.
That should create a nice intermodulation like a wobble, especially as it is also coming from 20 different sources creating phase shift effects between your ears. Don't have the panels and the individual sounders too close together.
I am sure you will find a point in detuning them that will make your ears bleed

[cmangle]

ok thanks . . . off to the bread board!

I did a google search for the nearest Rat Shack and i couldn't believe it!

Radio shack corporate has closed 1700 plus stores! ( I can only assume the franchised ones are safe!)

I think Mouser or Digikey (here in the US) will be my parts source(s)!

For all of your input you have a Jim's Steaks, Cheese steak on me. But I don't think I can get that cheese steak to you in time for it to still be edible!

[Uncle Fester]

Don't worry, my wife does those things better and trust me you would never even notice that there is actually no steak inside them

Our version of Radio shack was called Dick Smith.
They started dying here over 10 years ago and now it is just another obsolete Phone and Flatscreen dump store.

Dick Smith calls himself a supporter of Australian produce but I call him a hypocrite.

Most People today have become plain consumers. Creators are a dying species, at least in the western world.
The kids today simply have no time. Too busy with social networks and gaming or actually working like slaves to buy their own car with 16.
I was trying to BUILD my own car back then. LOL.
... and of course a total electronic maniac and earned money by building HiFi systems for my richer school buddies when I was 16.

[cmangle]

Just to be clear, the modifications and circuit adjustments you mention above and the posts before(#10, #11, #14, #16) are to your simple circuit you posted . . . .



Or are your mods and substitutions to the link I posted with the 555 timer circuit here . . .



I barely remember Kirchoff, Thevenin and Norton . . . that was many many years ago, so dusting off the brain for this circuit SHOULD have been simple! 8^(

[cmangle]

It's late here and I'm counting sheep as i type . . . . that being said there is no VR1 in your posted circuit! duh!

Thanks once more!

So, I can get an equivalent Jim's Steak cheese steak, in Oz, that the wife cooks, and it will be just as delicious with "NO MEAT"?

hmmm, that sounds like a story Crocodile Dundee might spin, just might have to do some travelin . .