Circuit Design

[Jo_Vision]

Hello, I need some help with the setup of a circuit. I am a mechanical engineering student and I am working on a mechanical system that employs the use of a sine wave at 1kHz, and the signal would need to be amplified to about 400-450V. It will be employed on a 24V DC system. I just need some help on how i should approach setting this circuit up. I have been looking at a 8038 waveform generator IC, but i feel like there might be a simpler way to get what i want. I am not looking for complete schematics on it. I was just wondering if someone could get me headed in the right direction in terms of components and a general setup. Thanks for the help guys.

[dlgoff]

Welcome to PF Jo. 400-450V sounds like a lot for an audio signal. What is you application? Vacuum tubes and transformers come to mind.

[Jo_Vision]

Its not audio. It will be used to power piezoelectric actuators on the inside of a plane wing. We are trying to keep everything integrated into one circuit if possible in order to avoid taking up a lot space also to keep weight down because its on a plane.

[dlgoff]

Well 1kHz is in the audio range. Are you trying to shake the planes wings off. Just kidding.

To get 400 volts your going to need some sort of power supply or use a transformer to step up the voltage.

Here are some High Frequency Transformers (http://www.thomasnet.com/products/transformers-high-frequency-87180451-1.html)

[Jo_Vision]

1 kHz is the frequency of the voltage

[Okefenokee]

If you don't care too much about the accuracy of the frequency output then you could rig some op amps in an astable configuration like this (http://www.national.com/nationaledge/jun04/article.html).

You know, I would have thought that someone would make a simple oscillator in the 1kHz range for audio applications but I didn't see any in a search of Digikey.

There are plenty of more options but they are going to involve more circuit design. That 8038 function generator chip looks pretty good. You're not going to get a much more simple solution than a chip with those useful "typical application" schematics already done for you. I would go with it.

In any case, you're still going to need to follow the signal generator with an amplifier that can provide a large current and then you'll need a transformer to step the voltage up as dlgoff said.

[Jo_Vision]

thanks for the replies fellas. Fortunately the voltage required by these piezoelectrics is on the order of milliamps, so the current shouldn't be to much of an issue.

[berkeman]

thanks for the replies fellas. Fortunately the voltage required by these piezoelectrics is on the order of milliamps, so the current shouldn't be to much of an issue.

You mean the current required is small, not the voltage. Probably just a typo.

[Bob S]

For 24 volts dc to 450 volts ac, the simplest transformer is probably a ferrite toroid or pot core with roughly a 1:10 turns ratio (push-pull). The weight for the core varies inversely with frequency, so a higher frequency than 1 kHz may be preferable. Also, the output is usually a square wave, so a sine wave will be a more complex circuit.

Bob S

[Okefenokee]

Fortunately the voltage required by these piezoelectrics is on the order of milliamps, so the current shouldn't be to much of an issue.

Be aware that if you use a transformer to boost the voltage then the current requirements will increase on the source. For example, if the transformer boosts the voltage 20x on the secondary side and the secondary side's load (piezoelectrics) draws 10mA, then the source (your generator) on the primary side will have to supply 200mA.

[Jo_Vision]

Yes that was a typo. Fortunately it looks like we will have about 10 or so amps at our disposal, so we should be in the clear as long as we can keep the number of piezos to a reasonable amount. As far as the transformers that were suggested, could you use an integrator to produce a sine wave from the square wave output or would the voltage be to high?

[berkeman]

Yes that was a typo. Fortunately it looks like we will have about 10 or so amps at our disposal, so we should be in the clear as long as we can keep the number of piezos to a reasonable amount. As far as the transformers that were suggested, could you use an integrator to produce a sine wave from the square wave output or would the voltage be to high?

If you integrate a square wave, you get a triangle wave. The 8038 IC that you mentioned in your original post is obsolete, but there are plenty of ways to make a 1kHz sine wave:

http://www.google.com/search?hl=en&rlz=1T4GGLL_enUS301US302&q=sine+wave+generator+ic&aq=3&aqi=g10&aql=&oq=sine+wave+generator&gs_rfai=

.

[skeptic2]

If you integrate a square wave, you get a triangle wave. The 8038 IC that you mentioned in your original post is obsolete, but there are plenty of ways to make a 1kHz sine wave:

http://www.google.com/search?hl=en&rlz=1T4GGLL_enUS301US302&q=sine+wave+generator+ic&aq=3&aqi=g10&aql=&oq=sine+wave+generator&gs_rfai=

.

And if you integrate that triangle wave, you get something very close to a sine wave.

As mentioned above, there are many methods for getting a 1 kHz sine wave preferable to double integration.

[Mike_In_Plano]

Sounds like a good application for a Royer Converter. It takes reactive loads very well, is extremely efficient and if well designed, it will tune into the larger system.