Generate High Frequency, High Data Rate BFSK

[richas]

I would like to Generate BKSK with a center frequency at about 1GHz and a data rate of 1MHz. We are lacking the Uber expensive signal generators and so I am looking for an alternative.

Perhaps a DDS evaluation board that would allow the toggling between two frequencies at 1MHz? I would want the signal to be phase continuous.

Ideas?

Thanks

Rich

 

[biff44]

get an adf4350 on the eval board, and capacitively couple a digital signal onto the Vtune pin.

 

[richas]

get an adf4350 on the eval board, and capacitively couple a digital signal onto the Vtune pin.

That is interesting. I am not sure in what way applying a digital signal (square wave) to the Vtune pine would affect the output. This pin is the input to the VCO and determines the output frequency. I suppose capacitively coupling the digital signal to that pin would change the frequency but how would you control by how much?

Thanks

Rich

 

[tony_lth]

Biff44 is right. We had done it for many years. But I only designed thr RF circuits, I don't know the details. The method can be used for 4FSK.

 

[FvM]

I suppose capacitively coupling the digital signal to that pin would change the frequency but how would you control by how much?

It's the usual method for modulation frequencies beyond the PLL loop bandwidth. The FM swing can be calculated according to the VCO sensitivity and the capacitive divider ratio.

 

[biff44]

Well, obviously you need to tailor the exact digital swing with a voltage divider circuit so that you get the exact frequency deviation you desire. You probably want something like a 0.3 V peak to peak AC square wave riding on top of whatever DC voltage the PLL wants to lock the VCO to.

you also want to use a low phase detector frequency (i.e. a high divisor ratio N) to keep the PLL operating in a quasi linear region. And a narrow control loop bandwidth.

 

[richas]

Thank you very much for the help. I will order the adf4350 eval board.

I am thinking about the digital signal interface. Perhaps I will feed it to a comparator fed by a precision voltage source, then to an R-divider, then to a decoupling cap..............


Thanks again.

Rich