Re: Design Suggestion Requested

Joerg wrote:
Artist wrote:
Joerg wrote:
Artist wrote:
Artist wrote:

I have contacted the manufacturer of the piezoelectric driver. It appears I got some information wrong.

It is a narrow band device. So it may work to square wave pulse it and depend on its narrow band to filter. But the Ritec RAM 10000 does synthesize a sine wave.


The 10MHz version in the plot on page 7 looks like regular wideband PZT to me. A narrowband transducer would ring out for a long time.


Mechanical loading shifts the resonant frequency, but it will still be narrow band.

Tech support could not tell me what kind of electrical load it will present. They will get back with me on this.


Interesting, that data is usually supplied with transducers. You can measure it. Since you seem to work in a university/research environment you could ask around if someone has a HP-4191A or another impedance analyzer. Best to bring the transducer there because those analysers are very heavy.

Power amps are usually of very low output impedance. While the spec may say 50ohms that usually means you should not load it with anything less at high power levels to avoid blowing the final stage transistors.


I am not necessarily going to duplicate the specs for the Ritec RAM 10000 because it is capable of an enormous amount of pulsing power which might not be needed.


Just do the pulser plus hydrophone test versus the Ritec. If you get the same results (or close enough for the scientist) life will become much easier :-)


Your are right about the bandwidth. The information I got from tech support was wrong. It is a wide band device. The scientist wants the frequency to be adjustable to between 200kHz and 500kHz in single cycle pulses that pulse at rate of between 50Hz and 100Hz.


Ok, that sounds much closer to standard practice.


The scientist will need to drive it with a sine wave.


Did he say why? This will make the whole setup quite elaborate and expensive. In my 23 years of ultrasound I've never seen a reason to drive a transducer with anything but pulses. Sometimes they get low-pass filtered a bit but not for functional purposes, that's to pass EMC.


I measured its current today with an oscilloscope using a 1 ohm current sampling resistor. So the oscilloscope would not be exposed to high voltage I measured the ground return current in the coaxial shield. At 500kHz the pulses showed +/-500mV across the resistor. At 200kHz it was +/-200mV. This indicates the load is highly capacitive which is to be expected.


It should not be extremely capacitive at its operating frequency range.


The current measurement showed the load is nonlinear. Impedance went lower with increasing voltage.

The Ritec monitor output showed the voltage went up to 1kV. This means at 500kHz I must deliver 500 Watts of peak power. My original plan was to use a DSS chip to generate the pulsing sine wave, buffer its output to higher current, and the step it up with a pulse transformer. But with a typical output of a DSS being 0V to 5V, and assuming the pulse transformer is 80% efficient, and unity gain for the buffer amp, I would need 500A into the pulse transformer. I need a better way to do this.


I'd seriously discuss the sine wave requirement with the scientist. To me a single-cycle sine doesn't make much sense. If he insist on it you'll have a huge project on your hands.

You can't really generate 500A at the 5V level. Well, you could but it'll be a bear. This is done differently, just like designing and budgeting an RF power amp for radio stations. First you need to find which power transistors are out there at reasonable cost and most of all, available from stock. Look at transistors for 500W-1000W transmitters but sit down before reading the prices :-)

Then look in their datasheets what voltage they like to operate at. This will most likely be in the 20-50V range. Then you'd have to build a multi-stage RF amplifier that can take the miniscule DDS signal and create a stiff RF signal at the 20-50V level. This gets transformer coupled to the transducer so you can run the RF amp at a point where it doesn't get loaded past its design limits. Use (almost) the full available voltage swing at max power since that avoids excessive currents, less heat.

Transformers are nearly free of loss but now for the bombshell: To obtain a reasonably clean sine wave this huge amplifier will have to run with some quiescent current, at least in a mode that RF guys call AB2. This means it'll be generating a lot of heat that you'll have to get rid of via heat sinks and fans.

Another option might be to buy a ham radio transistorized power amp, the biggest honking one there is. But you'll have to heavily modify it because they do not operate below 1-2MHz, meaning new transformers. Also, they usually won't offer gains of more than 20dB so you'd need a smaller amplifier before it.

Not that I want to discourage you but this is going to be a serious project. And it won't result in a small unit.


The reason for the sine wave is that the system works with what the Ritec RAM-10000 produces and so that is where the confidence is. I was unable test what it would do with a square wave because the Ritec RAM-10000 does not have an option to output this and the lab's function generator was not powerful enough to produce a measurable response in the system.

It is not practical to step up from 5V and 500A with a transformer. So right now I am looking at power op amps from Apex for driving a transformer:
http://apex.cirrus.com/en/products/pro/areas/PA139.html#PA141_open
If I can drive a sine with one of these or others like them I can reduce the winding ratio of the transformer and the current required at the transformer's input. I am not sure any of these from Apex are up to the task because all of them have severe output voltage swing compromises when driving capacitive loads. The data I have suggests the transducer capacitance is 158pF, and this must be multiplied by the winding ratio of the transformer to arrive at an approximation of what the power op amp must drive. Even if one these can do the job you would still be right about the size of the hardware. I might end up giving up on the sine wave and take a chance on the square wave.

--
To reply directly remove the sj. from my email address. This is a spam jammer.
.

Loading