radar antenna characteristics on rotating

[nauman.anwar]

if the radar antenna is rotating, how does the phase of received PRIs change during one scan

 

[HMS1021]

The antenna typically sets on top of a rotary joint. As the joint rotates around a E field probe 1/4 lambda from the end short there is very little change to the geometry therefore the phase is constant (<1 degree) over a 360 degree rotation. Ideally the phase change from one scan to the next at a given antenna angle is due to the target changing its range from the antenna. This is the basis of Doppler radar. Realistically due to ground effects, aspect angle changes of the target, the exact pointing angle of the antenna at the time of the pulse etc there is a distribution of angles even from stationary targets. To offset these effects the antenna beamwidth and scan rate are set to ensure multiple pulses on target per scan, >16 pulses (think points in an FFT) or more is a good start.

 

[nauman.anwar]

thankyou very much HMS1021
your reply helped me alot
but when i compare a rotating radar to a circular array of sensors, as there is a phase shift between each sensor, similarly a rotating radar can have phase shift for different PRIs in one scan.
what do you say

 

[HMS1021]

If you are trying to build a interferometric radar with a range resolution of a fraction of a Tx wavelength then the Tx and Rx are likely mounted directly on a narrow beam antenna that is pointed or beam steered. Only processed signals and power are sent across a rotary joint or slip rings.

For something like a Doppler weather radar the relative phase change Tx to Rx from pulse to pulse over a limited number of pulses is what is important. We are looking for a frequency shift at baseband based on some limited number of contiguous pulse samples. We do not care about absolute phase.
As viewed from your antenna what is the target's max angular velocity?
How much does the angle change across PRI * 16 or 32? (likely number of FFT points)
Would you expect very much transmission phase change (Tx to Rx) through the circular array with this angle change?


Note: Its is group delay that you need to watch.

 

[nauman.anwar]

thankyou

i am simulating a monostatic pulse doppler radar.
the system consists of a rotating antenna
targets velocity is 330*2 m/s
range = 300 km , which means angular velocity is 0.0022 rad/sec
PRF = 400 HZ, and antenna rotation = 3.5 RPM, beam width = 20 degrees, so i receive 381 pulses per 20 degree in one scan

so during the 20 degree scan, will the phase of tx and rx PRI change due to antenna rotation?

 

[FvM]

You are asking about variation of effective path length. As with the circular array, a certain path length modulation can be expected. But can you answer the question without analyzing the antenna design details?

 

[nauman.anwar]

Mr FvM
according to you what type of antenna design details are reqd?

 

[nauman.anwar]

how much frequency shift is significant in a band of 5 MHz

 

[biff44]

sounds like you have a very bad rotary joint

 

[HMS1021]

Sorry for the delay, I just returned from a week of radar testing in Colorado.

"how much frequency shift is significant in a band of 5 MHz" Which 5MHz are you referring to?

"so during the 20 degree scan, will the phase of tx and rx PRI change due to antenna rotation?" This leads to two questions:
1.) Yes, it will change, the question is: how much. If there are 381 pulse across the 20 degrees then a group of 32 pulses will only span 8.4% of the 20 degree FOV or 1.7 degrees.
2.) You write "when i compare a rotating radar to a circular array of sensors " implying non rotating but then state "the rotation is 3.5RPM". I'm confused if this is a fixed array with a 20 degree FOV for each antenna, or is it rotating, or both. Another option is it is a phased area with many 20 degree elements but a narrow beam is formed in the far field.

Across 1.7 degrees I would not expect group delay to be an issue even at the beam edges assuming the 20 degrees is defined by the 3dB points. The assumption here is the antenna's RF bandwidth is much wider than the Tx bandwidth. If you are operating at the beam's edge AND the edge of its RF bandwidth then the phase shift and group delay, could be large, in fact very large. As suggested by FvM without a HSFF model or equivalent, or antenna near field range measurements, no one can guess the amount of phase variation.

 

[nauman.anwar]

thanks for reply. i was waiting for you. how was your testing?

my antenna is rotating and 381 pulses are received in 20 degree scan. each tx and rx pulse has time during which antenna rotates 0.05 degrees.

you talked about group delay, but how is group delay important?

 

[nauman.anwar]

how can i model an EM signal transmitted in a particular direction in MATLAB.