1
\$\begingroup\$

For an ongoing project of studying magnetic properties of various metals we need to use a power supply which can generate 25 V peak to peak, frequency 50- 500 kHz, 5A (PP) current. Frequency would be sine/cosine (Hanning) with tone bursts of 1 to 10.

Being An electronics guy finding it is difficult to move ahead. Any help would be greatly appreciated.


Thought of using an audio amplifier but it may not work since signal bandwidth is limited to 100 khz. How about using a sinewave generator with power amplifier such as OPA541. Slew rate of amplifier is 10v/micro second. Any suggestion??

\$\endgroup\$
8
  • 3
    \$\begingroup\$ Are you looking for design help or a place to buy a power supply? \$\endgroup\$ – Elliot Alderson Mar 30 '20 at 18:39
  • 5
    \$\begingroup\$ This is more like something you would do with a power amplifier (fed from some kind of signal generator) than anything that is normally called a power supply. \$\endgroup\$ – user_1818839 Mar 30 '20 at 18:42
  • \$\begingroup\$ Look at SMPS (switch mode power supply). Frequency can be adjusted with a simple resistor or a pot. You just have to remove the soothing capacitor from the shematic to keep the sine wave. Of course, this is a simplified answer. \$\endgroup\$ – Fredled Mar 30 '20 at 22:45
  • 1
    \$\begingroup\$ What you need is like a really good audio amplifier. A commercial one may work for you at lower frequencies. What load are you driving? What is your budget? \$\endgroup\$ – user69795 Mar 30 '20 at 23:21
  • \$\begingroup\$ Agree about using the switching part of a SMPS. You'll have to remove the feedback loop and run the thing open loop. Also you'll have to design the transformer to work over the range of frequencies you've indicated, 50 - 500 kHz. On second thought, that could be a show stopper. \$\endgroup\$ – SteveSh Mar 31 '20 at 0:14
0
\$\begingroup\$

So you are curious about the transient response of some metals, using 1/2 sin to 5 sin period stimuli.

How about synthesizing the waveforms using a switch_mode amplifier.

There will be distortion products. If your output frequency is a sub-multiple of your switching/chopping frequency, the distortion will appear as a waveform error. Thus I'd suggest your switching frequency be variable-swept, over a small range, and you use data-acquisition-averaging to exclude the distortion.

If minor waveform errors are not important, then don't worry about this.

\$\endgroup\$
3
  • \$\begingroup\$ yes More precisely using it for magnetostriction Study. As Suggested by you and other members. I m thinking of using ad9102 for generating waveform then feeding it to a power amplifier circuit. ckt [link] (elprocus.com/150-watt-power-amplifier-circuit) .This circuit looks Quite bulky and has transformers whereas i Have to use a non transformer arrangement. My input signal to device would be 220VAC, 50 Hz and than it should generate desired output. I welcome your suggestion. \$\endgroup\$ – Rahul Mar 31 '20 at 17:52
  • \$\begingroup\$ Can you explain the requirement for "no transformers"? The circuit that you linked requires +-45vdc. You agree? You will need a transformer to step-down the 220 50 Hz to the correct power supply voltage. This does not mean that the output of you amplifier has a transformer - just the power supply. An additional point; do you know for sure that the amplifier that you linked has bandwidth to 500 kHz? \$\endgroup\$ – xstack Apr 3 '20 at 2:28
  • \$\begingroup\$ i want to avoid Transformer, instead using an SMPS or something similar. The amplifier should have a variable bandwidth up to 500KHZ. this is Must since we need to change frequency ,voltage and current. \$\endgroup\$ – Rahul Apr 3 '20 at 12:39

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.