Driving an Ultrasound Transducer

Hello!
Thanks to the help provided in this thread I was able to produce the pulsed ultrasound signal that I needed. (1.5 MHz sinewave, pulsed at 1kHz, 20% duty cycle, 21 V p-p)
The general design ended being as follows: a signal generator that produces the sinewave, a 555 circuit that controls a 4066 which switches the sinewave and finally a LM318 that amplifies the signal to 21 V peak to peak.

I was getting the correct signal by measuring with the oscilloscope without the transducer load. As I connected the transducer, the circuit worked for about 30s and the shut off and some smoke appeared. It seems no big damage was caused.

What happened? What I have read leads me to believe that I need a transducer driving circuit (disclaimer: I am a total electronic noob). I thought the LM318 handeled the "power"part by amplifying the signal, but it seems that this is a current issue, right?

So my question is: do you have any design advice for an ultrasound transducer driver for this kind of signal?

BTW, on the LM318 datasheet on page 9 it shows a circuit for "isolating large capacitive loads". Will that suffice or do I need other completely different driver circuit?

You didn't yet mention LM318 as intended output amplifier for an ultrasonic transducer. You also didn't tell about a transducer impedance, but I would generally doubt that it's able to drive the respective currents.

According to the datasheet, the device is said to be short circuit protected, but there are several points why it may be unable to drive a reactant load safely. You may in fact cause self oscillations by connecting the transducer without a series resistor. In this case, the output will most likely swing rail to rail and forward bias the substrate diodes, possibly causing destructive device latch-up.

It may be necessary to protect the output stage by schottky diodes connected to the supply rails. But basically, an output amplifier with higher current drive capability should be provided.

An ultrasonic transducer needs a signal source set to its frequency marked on the body. Some transducers are tuned to 24 kHz, others to 40 kHz.
Your first step, to generate the signal in a 555 seems to be correct. For 1.5 MHz I think the 555 is not good as the frequency should be stable in operation. I would rather recommend to build either a LC oscillator, or better, get a 1.5 MHz quartz and build a quartz oscillator. Next you will need to build a "transmitter" but make sure it is built into a well screened metal case. You are NOT allowed to interfere with medium-wave broadcasting using 1.5 MHz and around it.
After the quartz oscillator, you need a buffer stage ( a small-power transistor), and then a power stage capable to deliver ~21 Vp-p to the transducer. This can be a MOSFET and I would expect to dissipate several watts.
Refer to amateur 160-meter transmitters described e.g. in ARRL Radio Amateurs Handbook, or google a suitable design. You should monitor voltages and currents during adjustments on your transmitters, otherwise you can expect some smoke again.

Hi.
Don't worry, the 555 only generates the signal to switch the sinewave on and off (the sinewave comes from this signal generator).
Here is a basic layout:

Thank you for showing the schematic!
I do not know LM318 but I know ultrasonic transducers. They utilize piezo-ceramic material; when excited electrically, they "discharge" back and the voltage spikes can easily kill your LM318. I would protect ir with a ~100 Ohm resistor and some Zeners that will limit the voltage kicks to what is LM318 specified for. I use a 24 kHz ceramic transducer and excite it by a MOSFET driven by a 555 but through an output transformer giving ~100 V p-p output with the transducer. The toroid-core transformer nicely drives up the voltage, and the MOSFET says nothing to the spikes- it is rated to >400 V p-p.
If you use the 1.5 MHz single frequency, I would also advise to try the final stage with an IRF 640 or similar. They can be driven with a TTL signal, too. Use an output choke or transformer tuned to 1.5 MHz with the connected transducer.