Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Mosfet ultrasonic power amplifier

Status
Not open for further replies.

MexicoMac

New Member
I am looking for a circuit diagram for a high power ultrasonic amplifier. Needs to drive an underwater transducer rated at 50 watts. Currently using a single HexFet driving a 1:25 toroid to 1000vpp, but I need to increase the duty cycle past 1% for more output power. I think a dual HexFet design with a dual DC power supply would work best.
Any ideas would be appreciated.
Thanks,
 
RONV,
Yes, I have seen this one before. It IS the type of circuit I'm looking for. Unfortunately, it does not produce enough power.
Thanks,
Mac
 
If you double the supply voltage then the output power is 4 times higher.
 
hi MM,
Is your transducer a 40KHz version.?
Also what PRF rate do you want the TX to pulse at.?
Which type of RX unit are you using.?
E
 
If you double the supply voltage then the output power is 4 times higher.
I am currently using 36 volts, but hope to use something below 60v, to keep the electrocution factor down to a minimum.
Also, 60 volts is the maximum rating for the ZVP2106 & ZVN2106 so would need to keep voltage well below that.
Mac
 
To increase the power you must increase the signal voltage and maybe electrocute people.
The signal voltage can be increased by tuning the capacitance of the transducer with a series inductor.
The amplifier that was linked here shows a bridged amplifier that effectively doubles the signal voltage at the transducer.
 
hi MM,
Is your transducer a 40KHz version.?
Also what PRF rate do you want the TX to pulse at.?
Which type of RX unit are you using.?
E
I use transducers of various frequencies. I dip them all to find the exact resonant TX & RX frequencies (they are different).
The TX frequency is computer generated then fed to a buffer amp before driving the PA section. I don't need receive function.
My pulse repitation rate is currently 1% duty cycle. IE: 1ms ON, 100ms OFF, but I would like to increase it to 2 to 5 % to increase average power.
Mac
 
To increase the power you must increase the signal voltage and maybe electrocute people.
The signal voltage can be increased by tuning the capacitance of the transducer with a series inductor.
The amplifier that was linked here shows a bridged amplifier that effectively doubles the signal voltage at the transducer.
How would we know what size inductor to use? A tunable coil would be neat if we could guess on an approximate inductance.
The Amplifier shown amplifies the positive and negative portions of the square wave, which does double the signal.
http://www.discovercircuits.com/DJ-Circuits/ultra40khzxtr1.htm
Mac
 
The capacitance of the 40kHz transducer is probably stated in its datasheet or it can be measured. Then the inductance of the inductor is simply calculated.

I had a 240VAC fan that would not run on my 120VAC electrical voltage. I tuned its inductance to my 60Hz mains frequency with a series capacitor and then the fan ran fine and got 220VAC.
 
The capacitance of the 40kHz transducer is probably stated in its datasheet or it can be measured. Then the inductance of the inductor is simply calculated.

I had a 240VAC fan that would not run on my 120VAC electrical voltage. I tuned its inductance to my 60Hz mains frequency with a series capacitor and then the fan ran fine and got 220VAC.
OK Course! It must be getting late! Haha.
Use the good old formula 1/2pi x square root of LC, but solve for L.
Here is a sample Transducer which has 4000pf capacitance.
https://www.steminc.com/PZT/en/bolt-clamped-langevin-tranducer-40-khz
Other brands list Series and Parallel capacitance. Which one would I use? Neat on your fan.
Mac
 
I use transducers of various frequencies. I dip them all to find the exact resonant TX & RX frequencies (they are different).
The TX frequency is computer generated then fed to a buffer amp before driving the PA section. I don't need receive function.
My pulse repitation rate is currently 1% duty cycle. IE: 1ms ON, 100ms OFF, but I would like to increase it to 2 to 5 % to increase average power.
Mac

hi MM,
I used a 'burst' method of driving 200KHz and 30KHz transducers.
The transducer pulse generator was from a square wave osc, tunable over a narrow range in order to get the optimum power out from the transducer.
The pulse or burst width generator was adjustable from 100uSec thru 10mSec, this enabled a level of power control in the acoustic pulse.

The transducer driver stage was a NPN power transistor driving a 1:3 or 1:5 transformer at 30KHz or 200KHz respectively, supply voltage 12V or 24V depending upon the required acoustic power.

My product applications were for hydro-graphic depth echo sounders, working in water depths of 0.5mtr upto 400mres.

E
Is you application bio-diesel.?
 
OK E,
Not sure what you mean by a burst width generator, but think I'm doing a similar thing.
My computer generated signal is ms pulses of the desired ultrasonic frequency (say 40khz).
I also use 24v. See enclosed diagram. I am working on various applications, but bio-diesel is not one of them.
Mac
 

Attachments

  • New_Option_TX.png
    New_Option_TX.png
    9 KB · Views: 840
hi Mac,
By burst I mean the number of cycles of 40Khz that you gate to drive the transducer.

As your 40KHz is generated by the PIC have you arranged for a method of fine tuning tuning the 40KHz source, while the power amp is running, so that the transducer is resonant?

Eric
 
hi Mac,
By burst I mean the number of cycles of 40Khz that you gate to drive the transducer.

As your 40KHz is generated by the PIC have you arranged for a method of fine tuning tuning the 40KHz source, while the power amp is running, so that the transducer is resonant?

Eric

Hi Eric,
Isn't the number of cycles (burst) of 40khz determined by the length of the ON pulse? A 20ms pulse will have more cycles that a 2ms one (10 times more).
The size of the Burst is directly related to the duty cycle. In order to increase the burst time, we need an amplifier that will take more power. Currently I can't get past a duty cycle of 1% without smoking the heatsinked Fet. EG: 5ms ON and 500mx OFF.

As I mentioned before, I dip all transducers to find their exact resonant TX frequencies and their band pass.
Then I program the CPU to change frequency over the band pass. If the BP is say 2khz wide, I will step the frequency up by 100hz every few seconds.
That way I send multiple bursts to 20 different frequency locations, of which a few should be more effective in ringing the wafer.
What do you think?
Mac
 
Last edited:
hi Mac,
If you post me details of your transformer [ winding ratios and VA rating] and the way you have the transducer connected, together with the tuning capacitor I will run a LTS simulation of the amplifier.
You should be able to get a higher duty cycle without smoking.
Eric

EDIT:
What voltage levels are you using to drive the NMOS Gate.?
 
Last edited:
The capacitance of the 40kHz transducer is probably stated in its datasheet or it can be measured. Then the inductance of the inductor is simply calculated.

I had a 240VAC fan that would not run on my 120VAC electrical voltage. I tuned its inductance to my 60Hz mains frequency with a series capacitor and then the fan ran fine and got 220VAC.

AG:
This formula 1/2pi *square root of LC, is what you used?

Can you discuss this a bit more pls? I know of tank circuits and such, but how does the voltage almost double? Is it a matter of impedance matching for max power transfer?
 
In a series tuned circuit, the resistance of the inductor limits the current since a capacitor has no resistance. My 240V fan had a low resistance but the reactance of its inductance limited its current at 60Hz. The series capacitance I used tuned the tuned circuit to 60Hz which cancelled the reactance of the capacitance and inductance then Ohm's Law increased the voltage and the current in the resistance of the fan.
 
hi Mac,
If you post me details of your transformer [ winding ratios and VA rating] and the way you have the transducer connected, together with the tuning capacitor I will run a LTS simulation of the amplifier.
You should be able to get a higher duty cycle without smoking.
Eric

EDIT:
What voltage levels are you using to drive the NMOS Gate.?
Fet drive level is about 8 volts. Here is my circuit:
 

Attachments

  • New_Option_TX.png
    New_Option_TX.png
    9 KB · Views: 545
I built a cat scarer at my last house. I didn't bother with trying to tune to the ultrasonic horn I had. I drove the secondary of a toroidal mains transformer using a low voltage drive at ~41KHz and used the primary straight across the piezo horn.

Maybe not the most efficient, but it worked effectively and gave quite a nip when touched.

What surprised me most was the frequency response of a 6.0.6-240VAC cheapo toroid was at such a frequency.

It solved my cat problem BTW with me having to suffer the loud click when the device first powered the horn. I could have resolved this with a bit of signal/startup logic but never bothered in the end.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top