[SOLVED] Ultrasound project - pulse width, current

[ToKra]

Hi everybody, I need help with my project.

I want to excite my ultrasound transducer with short square pulses with amplitude around - 185V and width about 440 ns. The transducer resonant frequency is 2,25 MHz with 50 Ω impedance.

What is confusing me:
A) How to select square pulse width?
Basicly: resonant freq is 2,25 MHz so the pulse width should be T = 1/f = 440 ns?

B) How much current I will have to provide? Is it given by Ohms law -> I = 185/50 = 3,7 A?

My confusion come from this: All ultrasound pulsers that I have found cant handle so high current. Most of them generate pulses around +- 100V. With common transducer impedance of 50 ohms the current is still to high(Ohms law idea).
Such as Supertex TC6320 app note here: https://ww1.microchip.com/downloads/en/AppNotes/AN-H53.pdf
Unfortunately they have no info about short time max current in datasheet.

Dou you know about any pulser that I can use?

 

[Warpspeed]

Its a fair bit of power 185v x 3.7 amps = 685 watts.
For pulsed lasers and pulsed active sonar and pulsed radar applications, the usual way to generate very high power narrow rectangular single pulses is with a pulse forming network.

Do some internet research on pulse forming networks, its a very simple thing to implement, and very robust and reliable once its going.

Here is a calculator to help with selecting suitable values of L and C.
http://www.nessengr.com/techdata/pfn/pfn.html

 

[CataM]

I want to excite my ultrasound transducer with short square pulses with amplitude around - 185V and width about 440 ns. The transducer resonant frequency is 2,25 MHz with 50 Ω impedance.

What is confusing me:
A) How to select square pulse width?
Basicly: resonant freq is 2,25 MHz so the pulse width should be T = 1/f = 440 ns?

Period is 440 ns but your width = ON time you said is 440 ns, meaning off time is 440 ns, resulting in 880 ns period.

B) How much current I will have to provide? Is it given by Ohms law -> I = 185/50 = 3,7 A?

Yes, but in square shape, just like the voltage.

 

[c_mitra]

Your transducer may shatter if used without load and way beyond the design power. The transducer just acts like a resonator tank.

 

[ToKra]

Thank you for you replies!
I want to add some details:
I need only few pulses per second/few seconds. Since the device is only for experimental purpose - it will simulate ophtalmology A-scan on agarose fantom.

Suitable DC/DC design 20V/200V 10 mA can be found on Texas Instruments web. More info in link:

http://www.ti.com/tool/PMP10077

I am about to use pulser circuit shown in the attached picture. The transducer is excite by discharging capacitor. It would be good to use discharge resistor for C_HVV. C_DISCHARGE capaciance should be higher than capacitance of transducer, but I don´t know how many times higher.

The transducer capacitance is unknown for me, because Olympus support did not respond to my question. In literature and papers it´s said that the capacitance of transducers is mostly up to 500 pF ( few can get up to 3000 pf ).

Is it possible to measure the transducer capacitance with suitable multimeter, or more complex measuring has to be done?


Currently I am trying to solve how to drive the MOSFET. In some papers from IEEE database the Schmitt triger with current amp is used. But I want to generate only few pulses on command. Use of microprocessor would be ideal (Xmega is optional) + driver IC, but for know I don´t know if the Xmega will be able to generato so short pulses on button command, or command from PC.

 

[c_mitra]

Is the transducer is also used as a detector?

Is it possible to measure the transducer capacitance with suitable multimeter, or more complex measuring has to be done?

Yes, you can measure the DC capacitance using a conventional multimeter. But these transducers (I presume that the one you are using is piezoelectric in nature) are resonant devices- they have both L and C and perhaps the DC capacitance will be only a guide and nothing more.

- - - Updated - - -

A) How to select square pulse width? Basicly: resonant freq is 2,25 MHz so the pulse width should be T = 1/f = 440 ns?

If the resonance freq is 2.25 MHz, then the period is 440 ns and the pulse width should be 220 ns. Others have already pointed out this.

B) How much current I will have to provide? Is it given by Ohms law -> I = 185/50 = 3,7 A?

As the voltage increases (large dV/dt) rapidly, the transducer takes large current (capacitive) initially but it falls off due to the reactance of the transducer.

When the voltage is shut off, the transducer produces an induced voltage (like an inductor) and this oscillation will continue for a few cycles.

The current obtained by using Ohms law is not meaningful in the present context.

 

[FvM]

Did you notice that the Supertex application note implements a push-pull driver which is necessary to drive an actual square wave to the piezo tranducer? Your circuit can only generate a single exponential discharge pulse. It may be appropriate for your application, but looks at least unusual compared to typical piezo applications. Did you consider the intended pulse form for your application? I remember that single exponential pulses are usual in ultrasonic thickness measurement. But the transducer would have an absorptive backing in this application so that it's self resonance is strongly dampened.

High pulse currents with low duty cycle aren't necessarily a problem for your transducer. There are however absolute limits set by
1. the piezo depolarization voltage
2. the mechanical strength of the piezo material

 

[ToKra]

Yes, I know about supertex pulser and their app note. Their pulsers is only for lower voltage and current -> can´t use their product.

I want to use this topology because it´s easier solution. Here

it´s mentioned that it is possible tu use pulser topology with one active element and capacitor. Main aspect is the capacitance value of discharge capacitor. If the capacitance C1 is low then output is a spike with sharp negative front and slow positive front as you mentioned above. But with large capacitance of discharge capacitor it is possible to produce almost rectangular pulses. Basic ideas is explained in the paper and also on biosono web here: https://www.biosono.com/ElctLgd/ElctLgd.php?id=ED_UP

Also driving one MOSFET might by easier.

Negative rectangular pulses are recommended by transducer manufacturer.

 

[c_mitra]

Negative rectangular pulses are recommended by transducer manufacturer.

That's interesting; I would have assumed that the two sides of the transducer are symmetrical and one positive (magnitude of the voltage difference between the two sides) pulse will produce a compression wave and a negative pulse will make a rarefaction wave. I thought both are equivalent.

 

[FvM]

But with large capacitance of discharge capacitor it is possible to produce almost rectangular pulses.

True the real transducer impedance assumed in the link. But is it realistic?

Negative rectangular pulses are recommended by transducer manufacturer.

Could be because piezo tensile strength is larger than compressive strength.

 

[c_mitra]

Could be because piezo tensile strength is larger than compressive strength.

The static strength of most of these minerals are extremely high (rather similar to quartz or glass) but the dynamic strength (shock wave) is rather poor. But anyway I do not know the material of this particular transducer but is likely to belong to the TiO3 /ZrO3 class having similar crystal structures.

 

[ToKra]

True the real transducer impedance assumed in the link. But is it realistic?

According to paper mentioned above - yes. I will use push-pull topology if the capacitor version won´t be good enough.

The static strength of most of these minerals are extremely high (rather similar to quartz or glass) but the dynamic strength (shock wave) is rather poor. But anyway I do not know the material of this particular transducer but is likely to belong to the TiO3 /ZrO3 class having similar crystal structures.

Olympus dos not mention the material in datasheet and the support didn´t anwer my question about transducer material.

I will mark this question as solved, because main questions are answered know.

Thank you