Ash,

What was the frequency you found worked best with your setup?

Fixing the duty cycle won't be much of a problem, although I don't know if it would make that much difference anyway if it changed a few percent.

Sure, I'd be happy to design the circuit. Let me know what the fo frequency is and what range you want.

..99

Ash,

What was the frequency you found worked best with your setup?

Fixing the duty cycle won't be much of a problem, although I don't know if it would make that much difference anyway if it changed a few percent.

Sure, I'd be happy to design the circuit. Let me know what the fo frequency is and what range you want.

..99

poynt99 , sure thing and thanks man, ill get the precise frequency for you and post tonight, will have to wait a few hours.

Ash

It seems the XR2206 and ICL8038 are either extremely difficult to find or very expensive, which makes them a bad choice for sine generators.

Does the drive need to be sine wave?

It would sure simplify things if a triangle or square wave would do the job.

Could you test with your generator set to square wave instead of sine and see if the results are similar?

.99

@ Ash & Poynt99

I've found the XR2206 in stock at the following:

Mouser Electronics
$7.79 - http://mouser.com/search/ProductDetail.aspx?qs=H%252bP%2fcHWii3cm1aCPniRXbw%3d%3d
$ 7.29 - http://mouser.com/search/ProductDetail.aspx?qs=H%252bP%2fcHWii3dGXk0dx06osA%3d%3d

JDR Microdevices
$3.25 - http://www.jdr.com/interact/item.asp?itemno=XR2206


Future Electronics
$2.62 - http://www.futureelectronics.com/en/technologies/production-products/test-measurement/function-generators/Pages/3238037-XR2206CP-F.aspx
$2.60 - http://www.futureelectronics.com/en/Technologies/Product.aspx?ProductID=XR2206DFEXAR4075065


-Duff

Duff,

I checked these two chips earlier today and knowing then that current consumption is of prime importance in this application, I was hoping that square waves would do the job (as Ash and Andrew have just confirmed), so that a CMOS chip could be proposed for the oscillator design.

The 2206 and 8038 consume about 5mA with a supply of 10V, or about 50mW of power. The CMOS oscillator I've agreed to design shouldn't use any more than 100uA (900uW with a 9V source) or so when idling (i.e not driving any load) , so the only real power used from it will be to drive the MRA--the 80mW output with 600 Ohm output impedance.

Good to know that these chips can be found still at a reasonable price. The ones I found were in the range of $50 ea.

.99

Hi poynt99, this sounds great man, should be fine we will keep this MRA for your circuit, am going to wind a bigger one next weekend, strangely, the more winds the higher the impedance, so it went down to the audio range after bigger windings. The first one was 70Khrtz, this one went down to 16-20Khz

Ash

poynt99, thanks for the info, in the document you will notice Norm talk about how the golden mean ratio is critical in the Trafo winding and to always keep it in mind when designing . It strangely also does not produce the same power with out the 1.618 ratio, maybe this can explain the pyramids as tapping a resonant frequency..hmmm

Edit, whoops sorry missed the Q, check out the PDF, the winding and resonant  frequency is listed there 

Ash

poynt99, thanks for the info, in the document you will notice Norm talk about how the golden mean ratio is critical in the Trafo winding and to always keep it in mind when designing . It strangely also does not produce the same power with out the 1.618 ratio, maybe this can explain the pyramids as tapping a resonant frequency..hmmm

Edit, whoops sorry missed the Q, check out the PDF, the winding and resonant  frequency is listed there 

Ash

I was thinking something along the lines of this:

- The piezo element has it's own resonant frequency--somewhere in the ultrasonic region, probably around 40kHz.
- The combined LC resonance of the piezo and transformer inductance is at some optimum frequency.
- The magnet material itself will have its own specific resonant frequency (magnetostriction) which can be calculated by its dimensions if it is some regular form, which in this case it is not. It could be determined however by wrapping a heavy gauge low turn coil around the magnet and sweeping it to find the frequency at which it "sings" the loudest.

If resonance is the key to ou, then aligning the electrical and mechanical resonances in the unit should produce the best results.

Norm seems to have done something along these lines, but I'm curious how he determined his magnet was at a 8kHz resonance.

Ash, it is possible that in your first tests at ~70 kHz, you were operating the device at a harmonic of the piezo resonance, and now at 17kHz, you are at a sub-harmonic. Perhaps 32-35 kHz would give better results? Notice that Norm and Joel were using 32kHz.

Just some thoughts.

.99