# Is there an IC that inverts DC==>AC at the resonant frequency of a tank connected to it? (~50V ~40A)

Assuming I used the word "tank" correctly (something that has a resonant frequency) I have been working on a circuit that has 4 leads--two for DC input and two for AC output--such that you can attach something with a resonant frequency to the AC outputs, and the device will automatically adjust the frequency of the AC current to resonate with the tank.

I actually came up with a simple design using just two n-channel mosfets, two p-channel mosfets and four diodes. The principle of operation is that whatever voltage the tank produces, the mosfets direct the DC current to counter it (assuming the tank's voltage is above the threshold of the mosfets.) However, I found it in part by a lot of trial and error, and I was wondering if there was a standard IC that does this that I missed.

• There are lots of oscillator circuits that need only a single transistor. Do a web search and see if one of those meets your needs. If you still need help from us, we'll want to know what is your operating frequency and how much output power do you need? – The Photon Mar 1 '12 at 0:25
• What should I search for. I did not go about designing a circuit without a lot of searching. I could find nothing. I figured it would go under a different name. I call it a "resonator" but that did not yield any fruitful results. – Alex Eftimiades Mar 1 '12 at 0:36
• At this point, I would be fine with any power output. Just to know what the thing is called would be nice. I would like to use it to power an induction heater, so a 40amp tolerence at anywhere from 1khz-1Mhz (that is--whatever is feasible) would be great. I would not need very high voltage--something around 50V should be enough. – Alex Eftimiades Mar 1 '12 at 0:38
• 40 Amp and 50 V is a pretty challenging spec. An off the shelf chip solution is likely to be more like 5 V and 10 mA. I recommend you update your question to include these very important requirements. – The Photon Mar 1 '12 at 5:19
• "I recommend you update your question to include these very important requirements." -- Methinks it's too late; the question has been asked/answered/accepted on the assumption that it's a normal signal oscillator. It would be better to start a new question about a resonant converter. – Jason S Mar 1 '12 at 15:50

This is essentially what an oscillator using a "tuned tank" inductor-capacitor does.

Google Colpits, Hartley, LC resonant, Pierce, ... etc.

Can be done with as little as a single active device - eg a bipolar transistor or a MOSFET or a thermionic valve or ... .

Circuits below can be implemented "in an IC" if desired, but, just using a transistor will often be enough.

Tuned tank oscillator - many images, all with links

Good tutorial - Hartley Oscillator

Hartley Oscillators

Basic Hartley oscillators - varying grounding point.
NB note that this is what AC sees - DC needs a blocking component to stop DC shorting.

.

Shunt fed Hartley:

Feedback coil based from here

Colpitts from here

and again from here

• So where on these circuits would I connect a tank so that it would resonate at the frequency of that tank? I am trying to treat the tank as a black box with a resonant frequency. I want the device to figure out what the frequency is. – Alex Eftimiades Mar 1 '12 at 1:40
• On the so called in this case "Tuned Collector Oscilator" the tank is L1 and the 9 to 180 pF cap. If you dont want a second winding use another circuit. | In the basic Colpitts the tank circuit is the part in grey labelled tank circuit. If you don't want a tapped C use another circuit.| In the Hartley it's the only inductor plus the capacitor across it. If you don't want a tapped inductor use another circuit. | If you don't want any of these work out how get feedback from an output with a pure untapped tank. Doable. – Russell McMahon Mar 1 '12 at 2:54
• Are there any nice variations which don't require a "middle tap" in the LC circuit and can still 'find' a frequency close to the natural resonance of the LC? I'm not sure a circuit that would perfectly match the natural resonance of any connected RLC would be even theoretically possible given that the resistance would cause the phase difference between voltage and current to be something other than the 'ideal' 90 degrees; the circuit would have to compensate for that, and in so doing 'pull' the frequency slightly. Still, I wonder how close one could come to 'ideal'? – supercat Mar 1 '12 at 16:25