Phase detection in microwave radio

[DanTheMan]

I have a microwave radio link where the transmitter and receiver are linked together by a common clock.

The transmitter upconverts a baseband low-frequency i/q signal to a higher frequency and this is sent out to the receiver which down-converts it to the same i/q frequency.

With the transmitter and receiver spaced distance d0 apart there is a particular amplitude attenuation and phase (say amp_0, phi_0), whereas if the transmitter and receiver spaced distance d1 there is another amplitude phase response (say amp_1, phi_1). The amplitude difference is easy to calculate; however how could one calculate the phase difference, phi_1 - phi_0?

The reason for doing this would be to be able to compute/verify the spacing change between the transmit and the receive through the measurement of the phase difference...

Any thoughts?

Thank you!

 

[biff44]

I think you would have to further explain what you mean by "linked together by a common clock". Also, what sort of distances are you talking about? 1 meter, 2 km?

In general you can not resolve distance with a one way phase measurement. You can step the frequency of the carrier, and impart different different phases and calculate distance using the group delay equation T=Δphase/Δf. The old HP link analyzers did the same measurement with modultion sidebands.

If you do in fact have a clock somehow shared between the two, you could measure phase, but would have to take out the phase shift of the clock phase moving with distance (the clock has to propagate to the receiver too, and if that distance changes, then its phase is a moving target too).

On top of this all, you have to use a frequency low enough to not form phase ambiguities, or to resolve them with an alogrithm.

Rich

 

[DanTheMan]

Each of the transmitter and receiver have a VCO+PLL for up- and down-conversion to IF, respectively...in this case we're able to synchronize them with a common clock. The distances discussed here are in the range of 1 - 5 m.

We're not trying to get absolute phase measurement, but rather a difference in phase between two locations of the transmit/receive pair. Assume that at time 0 the transmitter and receiver are at distance d0 and there is a phase measurement of phi_0, then say transmitter is kept fixed and receiver is moved to distance d1. The phase measurement should now be phi_1 (which could be 2*n*pi + delta_phi away from phi_0). Using this relative phase measurement of phi_1 - phi_0 we should be able to verify the distance (allowing for the 2*n*pi complete wavelengths that may be in the measurement).

Does this make sense? Thanks.

 

[biff44]

Sure.

So there is a common xtal oscillator clock feeding both PLL's, thru two flexible high quality cables?

The problems are that the cable carrying the clock will vary in phase length as it is bent. You can buy a very expensive cable if it is in your budget (like they use in network analyzers). Or you could use a fiberoptic cable to distribute the clock.

You have to make sure the pll's stay on and stay locked with no phase slips. As an example, if you have a 10 MHz clock and a 1000 MHz carrier frequency, you have 100 possible phase steps that the PLL can be randomly locked too. Make sure the power level into the PLL divider chain has plenty of rf power to minimize the chance of a missed count.

Other than that, you just have to figure out what method you want to compute the phase with.

Rich

 

[DanTheMan]

What about in the case where the PLL/VCO components on each of the Transmit/Receive are used only to drive the LO? This should be equivalent, correct? I.e. if we have a source at 100 MHz that gets up-converted (in the transmit) and downconverted (in the receive) to 3.65 GHz (and 3.65 GHz is transmitted through the air) and the two LOs that drive the mixers for up-down-convert are locked to a common clock will this still hold? My understanding is that it should ...

 

[biff44]

I have to admit that I am not quite following you. Can you sketch up a system block diagram showing the fixed and moving parts?