Sine Wave Oscillator with Multiple Phases

Thread Starter

bte

Joined Aug 15, 2012
9
For a project I'm working on, I need to generate a sine wave at about 3.3Mhz, then I need to get phase shifted outputs at 60 degree phase intervals so I have a phase output of: 0,60,120,180,240, and 300. Alternatively, it could just be phases 0,60, and 120, then I could run it through a modified full wave rectifier to get the other half of the sine wave phases.

These outputs will drive a chain of saturable inductors to form a pulse, which will be stepped up to high voltage later. I'll probably need to amplify the sine waves to get enough power to push into the inductors and step-up transformer.

Any ideas?
 

Ron H

Joined Apr 14, 2005
7,063
For a project I'm working on, I need to generate a sine wave at about 3.3Mhz, then I need to get phase shifted outputs at 60 degree phase intervals so I have a phase output of: 0,60,120,180,240, and 300. Alternatively, it could just be phases 0,60, and 120, then I could run it through a modified full wave rectifier to get the other half of the sine wave phases.

These outputs will drive a chain of saturable inductors to form a pulse, which will be stepped up to high voltage later. I'll probably need to amplify the sine waves to get enough power to push into the inductors and step-up transformer.

Any ideas?
I don't have a phase shifter for you, but note that if you invert 0°, you get 180°. Likewise, 60° --> 240°, and 120° --> 300°.
 

Thread Starter

bte

Joined Aug 15, 2012
9
I don't have a phase shifter for you, but note that if you invert 0°, you get 180°. Likewise, 60° --> 240°, and 120° --> 300°.
Yes, that's what I was thinking to save on phase shifting too many times. Just need a way to shift at 60 degrees.
 

Ron H

Joined Apr 14, 2005
7,063
I can design the phase shifter for you, for low-level signals on the order of 1V p-p. Would that be useful?
Can you use surface mount parts?
 

Thread Starter

bte

Joined Aug 15, 2012
9
Yes, I would most appreciate that. I don't normally use surface mounted components, but I can work with it.

Thanks for your time!
 

#12

Joined Nov 30, 2010
18,224
I'm staying out of this because I have weird ideas like, with a fixed frequency, a simple RC circuit will provide a predictable phase shift.
 

Ron H

Joined Apr 14, 2005
7,063
I'm staying out of this because I have weird ideas like, with a fixed frequency, a simple RC circuit will provide a predictable phase shift.
Yeah, I thought about that, but there are attenuation considerations, which you could fix fairly easily, if the frequency were truly fixed. The OP mentioned "about 3.3MHz". I thought this might mean that some experimentation would be in order.
Right now, I'm thinking that two cascaded 1st order 60° allpass networks might be the best.
 

Ron H

Joined Apr 14, 2005
7,063
Bte, can you use square waves? If so, the simple solution is a ≈20MHz oscillator and a 3 stage Johnson counter. The phase shift is exactly 60°/stage, independent of frequency.
If you need sine waves, you can add 6 matched passive lowpass filters.
 

Thread Starter

bte

Joined Aug 15, 2012
9
It may be better to just use square waves, in that case I'll just use a flip flop chip to make the "phases" and drive the flip flop from a signal gen. Finding high voltage mosfets to handle over 1.2kv is pretty tough, the switching times need to be extremely quick too. Lower voltage mosfets would work better I think, the current requirement is minimal as well.

My original approach was to generate sine waves, feed the sines into an amplifying stage to get a 24v sine wave, compress the waves into pulses using saturable reactors, and step the voltage up with an air core transformer to whatever voltage needed. (1.5kv - 10kv)

I wish there was an analog approach, digital circuits and HV don't really mix well.
 
Top