Who has experience with AD8333 to demodulate a square wave?

[bouyang]

I am wondering if anyone has used AD8333 to demodulate a square wave input (Not the LO, but the RF input).

 

[thankz]

how do you demodulate a square wave?

 

[bouyang]

I have a CW square wave, I am hoping to extract its phase and amplitude related to a reference signal (supplied tot he LO port).

 

[berkeman]

I am wondering if anyone has used AD8333 to demodulate a square wave input (Not the LO, but the RF input).

how do you demodulate a square wave?

I have a CW square wave, I am hoping to extract its phase and amplitude related to a reference signal (supplied tot he LO port).

First, you're not likely to get a square wave from an RF source. What is your source, and what is the frequency?

And as pointed out by thankz, a square wave by definition has no variation in phase. I suppose you could amplitude modulate it, but then all you need is an envelope detector to decode it -- no LO is needed.

A typical AD8333 application circuit...

http://circuits.datasheetdir.com/315/AD8333-circuits.jpg

 

[jim hardy]

Square waves?

I've AND'ed them after shifting to logic level voltage.
Resulting duty cycle is in proportion to phase and lends itself to averaging.

But no, i never used that device.

 

[bouyang]

The source is a photo detector output of a square wave modulated light source at 500KHz. What I hope to do is to measure the phase shift/delay between this output and a reference signal.

If I supply a reference signal (4 x frequency) to LO.

Mixing of the RF input: sqrt(2*pi*w+ph)=sin(2*pi*w+phi)-1/3**sin(2*pi*3*w+3*phi)+1/5**sin(2*pi*5*w+5*phi) and
and the reference: sqrt(2*pi*w)=sin(2*pi*w)-1/3**sin(2*pi*3*w)+1/5**sin(2*pi*5*w)

after a lowpass filter the DC component will be: cos(phi)+1/9*cost(3*phi)+1/25*cos(5*phi) = cos (THETA)

I was hoping the demodulator output can give me the ability to measure THETA and then derive phi based on the above relation.

Do you think this is valid?

Thanks again for all your help!

 

[bouyang]

Square waves?

I've AND'ed them after shifting to logic level voltage.
Resulting duty cycle is in proportion to phase and lends itself to averaging.

But no, i never used that device.

Jim

Thanks. That's what I original thought. But my RF input is very low...

 

[Baluncore]

I would feed the small noisy received square wave, (RX), through a wideband limiting amplifier first. Then I would phase lock an oscillator to that limited signal. The PLL could have a long time constant since only phase shift is required. A Voltage Controlled Crystal Oscillator, (VCXO), might be used. Measuring the phase difference between the TX and PLL(RX) would be significantly more reliable than processing noisy signals.

Design will depend on the bandwidth of the RX signal. Are you designing a laser range-finder ?

 

[bouyang]

Yes, I am essentially trying to build a laser range finder... Thanks you for the advice! In my case the RX signal will be 500KHz.

 

[bouyang]

BTW,

I would feed the small noisy received square wave, (RX), through a wideband limiting amplifier first. Then I would phase lock an oscillator to that limited signal. The PLL could have a long time constant since only phase shift is required. A Voltage Controlled Crystal Oscillator, (VCXO), might be used. Measuring the phase difference between the TX and PLL(RX) would be significantly more reliable than processing noisy signals.

Design will depend on the bandwidth of the RX signal. Are you designing a laser range-finder ?

Also, I would greatly appreciate if you can recommend a wideband limiting amplifier? Thanks again.

 

[Baluncore]

I would first consider something like the AD8306.
See this link for; AN-691 "Operation of RF Detector Products at Low Frequency"
http://www.analog.com/media/en/technical-documentation/application-notes/AN-691.pdf
It lists;
AD8302 Special Purpose
AD8306 Demodulating Logarithmic Amplifier
AD8307 Demodulating Logarithmic Amplifier
AD8309 Demodulating Logarithmic Amplifier
AD8310 Demodulating Logarithmic Amplifier
AD8361 RF rms-to-dc Converter
AD8362 Exponential Logarithmic Amplifier

Some limiting amplifier ICs only generate the logarithmic RSSI signal, but you need output(s) from the limiting amp.
The latest WB limiting amps are used for xGHz optic fibre receive data amplifiers.

Many FM receiver chips have a limiting amplifier with RSSI and signal output(s). For example; the nxp SA604
http://www.nxp.com/documents/data_sheet/SA604A.pdf

 

[bouyang]

I would first consider something like the AD8306.
See this link for; AN-691 "Operation of RF Detector Products at Low Frequency"
http://www.analog.com/media/en/technical-documentation/application-notes/AN-691.pdf
It lists;
AD8302 Special Purpose
AD8306 Demodulating Logarithmic Amplifier
AD8307 Demodulating Logarithmic Amplifier
AD8309 Demodulating Logarithmic Amplifier
AD8310 Demodulating Logarithmic Amplifier
AD8361 RF rms-to-dc Converter
AD8362 Exponential Logarithmic Amplifier

Some limiting amplifier ICs only generate the logarithmic RSSI signal, but you need output(s) from the limiting amp.
The latest WB limiting amps are used for xGHz optic fibre receive data amplifiers.

Many FM receiver chips have a limiting amplifier with RSSI and signal output(s). For example; the nxp SA604
http://www.nxp.com/documents/data_sheet/SA604A.pdf

Thank you very much for the help!