Method 2 make burst then short circuit

[EE4life]

Hi All,

I am trying to make a system which supplies a large burst mode AC signal to a piezo and then short circuits the terminals once the burst is over( here in the literature).

The system has a Function Generator attached to a High Power amplifier which is attached to the leads of the piezo. I know how to program the function generator to give burst signal of a certain number of cycles, but how can I short circuit the terminals after the burst is done? The measurement is too fast to short circuit by hand.

Frequency: 20kHz
Voltage (out of the amplifier): ~50V
Piezo impedance: 10 Ohms (resonance drive)

Thanks in advance.

 

[Baluncore]

It is better to make a single test jig for such a test procedure. The test jig would have an H-bridge of four mosfets to provide the high power drive to the transducer. It would also have a pair of mosfets used to disconnect the H-bridge from the fixed voltage power supply rails.

An H-bridge will provide a differential voltage across the transducer of 50V with only a 25V supply. As the gates of the mosfets must remain within 30V of their source voltage, the H-bridge has an advantage in that it permits the mosfet gates to be driven directly.

The test procedure would be;
1. Turn on the mosfets that connect the power rails.
2. Cyclically turn on and off the diagonal pairs of the H-bridge to drive the transducer with a high amplitude signal.
3. Turn off the supply rail mosfets, turn on all four H-bridge mosfets. The transducer is then short circuited by the H-bridge, which is disconnected from the power supply.

 

[EE4life]

Thank you for your through reply. Unfortunately, I do not have a DAQ which can manipulate output signals at 20kHz+. However, it is a great idea.

The researchers in the literature said that they short circuited the terminals at the amplifier. They did not mention how they achieved the short circuit condition.

Could I use the following setup? Use a normally-on mosfet with (AC to DC rectifier on it) between the terminals at the power amplifier. I will hook the function generator up to the rectifier, which will turn off the mosfet when the burst is turn on, therefore allowing the signal to reach the transducer. When the burst is done, the mosfet will turn on and the circuit will be shorted.

 

[Baluncore]

If you short circuit the amplifier input you will pull the amplifier output to zero volts.
If the piezo was “shorted” with a 10 ohm resistor, or a 10 ohm amplifier, then transducer current could be monitored as voltage across the resistor.

Shorting the piezo with zero ohms will result in a loss of external voltage signals. You will then require a current sensor to monitor the loss of energy in the shorted transducer. The 10 ohm impedance of the piezo will limit the current circulating through the zero ohm short circuit. But at the instant the short circuit is created there will be a current spike as the parallel capacitance of the piezo transducer is discharged.

If you are interested in the response to a single transient pulse then shorting the transducer will need to be synchronised with the pulse generator.

If you are interested in the response to a pulse train then detecting the drive burst will be difficult. One 20 kHz cycle will take 5 usec which will be similar to the expected response time of a stimulation detector.

A low impedance square wave stimulation could be detected more easily than a sine wave stimulation from a 10 ohm impedance generator.

Unfortunately, your reference is behind a 2,000JPY, (US$20), pay wall so I cannot access the full text.

 

[EE4life]

Try this referece See figure on page 5. I am using an occilloscope, laser doppler vibrometer, current probe, amplifier, and function generator and computer

I have a current probe to measure current, so I do not need a resistor.
I will set the oscilloscope to read a few milliseconds of data after a trigger from the function generator signal.

I think if I do what I suggested (use rectified AC signal from function generator as the logic for the mosfet between amplifier terminals), when the burst stops the MOSFET will turn on and the circuit will be shorted. The number of cycles of the burst is between 200 and 1000.

I am going to measure the decay of vibration and current to determine the mechanical losses of the transducer/material. So I am not concerned about the details of voltage while the transducer is being driven..

 

[Baluncore]

Thanks for the reference. Things are much clearer now.
Your rectifier detector will work, but it's design requires more information because it must avoid a lock resulting from a false detection.

Can you please identify the make and model number of the available function generator?

Is your burst always a sine wave burst, or can it be a square wave? (All harmonics will be rejected by a resonant device).
How much time is available to decide if the pulse train has finished and the piezo can be shorted?
How much time jitter can you tolerate in the start of shorting relative to the final cycle in the burst?

 

[EE4life]

Sorry for the late reply. I think I got it. 2 mosfets back to back can act as a good AC switch. Thank for your help.