High voltage amplifiers

[gaddra]

I'm having trouble finding a suitable way to drive some piezo actuators.
I have a 0-4.096v input signal, and I need a gain of ~55. Normally I would use something like the Supertex HV256, but I need a higher speed than that can provide. The output will updated at a rate of 100khz, with a maximum slew of less that .1V/us.
Current draw is very small (nanoamp range), and I will have 64 channels so size is important too.

Anyone have any ideas?

 

[chuckey]

So you need 4.1 V X 55 = 225.5 V output? . At 100 KHZ a cycle is 10µS, so at a slew rate of .1V/µS, the signal would only get to 1V before the input disappeared. I would think a slew rate of 200V /µS would be more like it.
What is the capacitance of your transducer? From this and the wanted slew rate you can calculate the capacitance charging current and then the likely source impedance. then you can see that you will need a voltage amplifier with a moderate current output and moderately low output impedance. A little bit of inductive peaking might extend the frequency response with out causing too much ringing. I would think that a pair of 2W high voltage 5W transistors with an IC amp as a front end and to apply voltage feedback to, might just work.
Frank

 

[gaddra]

So you need 4.1 V X 55 = 225.5 V output? . At 100 KHZ a cycle is 10µS, so at a slew rate of .1V/µS, the signal would only get to 1V before the input disappeared. I would think a slew rate of 200V /µS would be more like it.
What is the capacitance of your transducer? From this and the wanted slew rate you can calculate the capacitance charging current and then the likely source impedance. then you can see that you will need a voltage amplifier with a moderate current output and moderately low output impedance. A little bit of inductive peaking might extend the frequency response with out causing too much ringing. I would think that a pair of 2W high voltage 5W transistors with an IC amp as a front end and to apply voltage feedback to, might just work.
Frank

Thanks for the reply.
The voltage swing will not change that fast, each step (at 100khz) the output will be changing some x amount (<1V) in order to converge on the solution (closed loop system), so a slew of 0.1V/µS should suffice.

Can you clarify what kind of topology that is, and what I should search for to find out more?

 

[Analog Ground]

A major manufacturer of HV amplifiers suitable for driving piezos is Apex Microtechnology. See https://www.apexanalog.com/ They have many app notes and design tools for piezos and capacitive loads. You mention current load in the nanoamp range. However, your current consumption is probably going to be the dynamic current of charging your capacitive load with each output voltage change. I would be somewhat surprised if this dynamic current was only in the nanoamp range. I used Apex Micro parts for a project with 188 actuators in a piezo deformable mirror. The PA-95 amplifiers worked as advertised. My experience with Supertex was not as good. In any case, Apex has a more extensive product line. Finally, design for the case of the control system running amok and oscillating. This situation can really stress the amplifiers and load. In other words, current limit!