Creating circular polarization.

• Use an x,y coord system with y pointing north,x pointing east, z up.
  
• radial or tangential dipole
• Which 1/2 dipole of a pair the center conductor feeds:
• tangential: rightmost (looking from center of dish.. this is plus x)
• radial: 1/2 dipole pointing toward the center of the dish. (this is -y)
  
• the 120  deg angle between the towers
• any phase differences in the cables going from the transmitter to the dipoles.
• The 6 phase generators (1 for each dipole) generate a phase delay
• if  dipole 1 has  0 phase
• dipole 2 has 30 degrees phase
• Then the dipole 2 waveform is 30 degrees behind the dipole 1 waveform.
• The original phase we start with is relative to the output port  (coupler) of a transmitter.
• We then calibrate the path through the transmitter and back through the monitor cable by:
• Use the rhode and schartz (or field fox) to measure the phase of the output signal
• pick 1 dipole as the reference. Set this to 0 phase.
• Put the correct phase (set at the output of the tx) for each dipole in turn.
• Adjust the phase output by the synth until the measured phase matches the theoretical phase.
• This removes the phase through the tx and monitor cable.

 The tables below summarizes the measurements for the 5 and 8 mhz systems

Setup to measure the phase

• set fieldfox to vector voltmeter mode
• a/b  with reference in A
• for 5 mhz tx 2 is used as the reference
• for 8 mhz tx 5 is used as the reference
• these were randomly picked long ago.
• The calibration is done for:
• using the older ao system
• using the nrl timing generator
• For both of these we need to do the phasing for rcp,lpc, (linear?) at both 5 and 8 mhz.
  
• Older ao system
  
• set freq:
  
• 5mhz we used 5.125Mhz
• set ampl; -0.8 dbm  (this then goes through a splitter to the 2 sets..??
• bring up the hx tx to at least 65KW.
  
• You need to adjust the power before starting since we disconnect the power power monitoring cables to do the measurements.
• from front panel, disconnect tx ref (hf2 for 5mhz) from multiplexor and plug into ref A of FF.
• cycle through each hf tx unplugging from mux and plugging into ff B
• using ao gui, adjust phase till it matches the values in the table below.
• req phase using tx 2 as reference  for 5mhz. be sure and used the most recenter tables.
• log the phase used to get the requested phase.
• generate an ascii file with these values (see hfsynth files on hf computer0. label file with pol,power, and date.

5 MHz phase

    There are two sets of calibration data for 5MHz:

• up to 31oct18. before dana's box was modified to have dual frequency inputs
• after 31oct18. after dana's box was modified to have dual frequency inputs
  

    Phase correction Notes:

• (1*). equation cos(w*t + phase)  has a positive phase leading  wt.
  
• When the signal generator outputs a phase, it lags wt
• to correct for this we must use the negative lead phase for the lag phase output.
  
• (2*):180 Degree offsets if the cable lengths is not the same number of half wavelengths as the reference cable.
• 5Mhz cable lambdas: 6.5,6.5,6.5,7.0,8.0,8.0.. tx2 is the reference, so the last 3 tx have a 180 phase difference in the path length.
  
• (3*):180 degree offset for radial dipole because we feed the dipole pointing inward
  
• (4*). this is the theoretical phase to add to the waveform  .. eg: cos(wt + phaseDeg*pi/180.)
• In this column,  positive phase means it  comes before the reference.

5  MHz calibration after 31oct18

5 MHz calibration 02nov11 to 31oct18

8 Mhz phase (10 feet down from 5 MHz mesh position)

• tx 5 is used as the reference.
  
• tx 3,4 have 1/2 wavelength length differences from the tx 5 reference.
  
• 12oct16 calibration measurement 1st done.
•  The tx power was about 50 KW each tx
  
• we also generated a file for 15 feet down. It has the same phases as 10feet down, but we increased the drives as high as we could go.
• 31oct18: recalibrated after dana's box was modified for 2 freq input.
• There are now two inputs. each driving 3 dipoles.
• This was added to do the 2 freq experiment.
  
• we split the siggen signal. 1 sig was sent to tx 5 (the reference). The other sigwas sent to each other tx 1 at a time to measure the phase difference.
• The phase was measured with 80KW going to each tx.
  

8 MHz calibration (after 31oct18)

8 MHz calibration  (oct16 - 31oct18)

Creating linear polarization.

15may19:

• The values in the first table below were used in the 2015 8 MHz linear experiment
• They give a linear polarization. If each dipole has unit amplitude,  the output vector is length 3.
• The 2nd table table has a set of values that gives an 8MHz  output vector of length 3.65
  

    Frank and herb wanted to transmit linear polarization  using 8mhz.. On 17mar17 we measured the phases needed to transmit linear polarization.

The 8 mhz measurements used:

• 8 MHz Phases used  for 2015 experiment (giving relative amplitude of 3.0)

  transmitter	1	2	3	4	5	6
  rad or tang	r,y	t,x	r,y	t,x	r,y	t,x
  tower	2a	2b	4a	4b	6a	6b
  initial phase (deg)	0	0	240	120	120	-120
  lead to lag for siggen	0	0	-480	-240	-240	240
  lambda/2 heliax correction	0	0	180	180	0	0
  radial fed to inward pointing dipole	180	0	180	0	180	0
  req Phase at output	180	0	120	60	60	120
  req phase using tx 5 as reference	120	300	60	0	0	60
  phase used.	111	248	29.3	335	ref	24
  output phase measured	120.1	-60.8	60.1	1.8	ref	59.8

  
  

processing: x101/170317/franklin.pro, x101/Y15/150325/hfphase.pro
                   redid computations more legibly 21may19 .. x101/hf/plotphase/linphase.pro

Phasing the nrl siggen

    The nrl siggen: when you put in 90 degrees, that goes in as wt+phi.. (a lead.. rather than a lag of the older system)

• 07jun19 ..measured the phases using fieldfox vvm..
• But we  use A/B ..  with the reference in A.. This gave -90 deg when the other tx had a +90 deg phase shift..
• We should have used B/A .. with the reference in A. this will measure leads.
• I fixed the numbers we measured by reversing the sign of the phase used.. so the table should be correct.
  

NRL 5MHz RCP phase calibration (10jun19 with tx bug fixed)

transmitter	1	2	3	4	5	6
rad or tangential	t,x	r,y	t,x	r,y	t,x	r,y
tower	1a	1b	3a	3b	5a	5b
initial phase (as lead) (deg)	0	-90	-240	-330	-120	-210
initial phase  (as lead) (deg) (1*)	0	-90	-240	-330	-120	-210
lambda/2 heliax correction (2*)	0	0	0	180	180	180
radial fed to inward pointing dipole (3*)	0	180	0	180	0	180
req Phase at output	0	90	-240	-330	60	-210
req phase using tx 2 as reference	270	0	30	300	330	60
phase used	258	0	13	303	292	57
output phase measured	270.7	0	29.5	299.9	329.3	60.1
output phase measured with old system using A/B..	90	0	330	60	30	300

 

NRL 5MHz RCP phase calibration (used 10jun19 with bug)

transmitter	1	2	3	4	5	6
rad or tangential	t,x	r,y	t,x	r,y	t,x	r,y
tower	1a	1b	3a	3b	5a	5b
initial phase (as lead) (deg)	0	-90	-240	-330	-120	-210
initial phase  (as lead) (deg) (1*)	0	-90	-240	-330	-120	-210
lambda/2 heliax correction (2*)	0	0	0	180	180	180
radial fed to inward pointing dipole (3*)	0	180	0	180	0	180
req Phase at output	0	90	-240	-330	120	-210
req phase using tx 2 as reference	270	0	30	300	30	60
phase used	258	0	13	303	348	57
output phase measured	270.7	0	29.5	299.9	30	60.1

NRL 5MHz LCP phase calibration (after 31oct18)
these stising_the_dipoles.html ll have the lags.. we need to redo it..

transmitter	1	2	3	4	5	6
rad or tangential	t,x	r,y	t,x	r,y	t,x	r,y
tower	1a	1b	3a	3b	5a	5b
initial phase (as lead) (deg)	0	90	240	330	120	210
initial phase  (as lead) (deg) (1*)	0	90	240	330	120	210
lambda/2 heliax correction (2*)	0	0	0	180	180	180
radial fed to inward pointing dipole (3*)	0	180	0	180	0	180
req Phase at output	0	-90	240	330	-120	210
req phase using tx 2 as reference	90	0	330	60	330	300
phase used.	259	0	45	298	339.8	41
output phase measured	90.3	0	300.5	60.4	330.2	300.2

• lcp tx3 phase used is wrong .. off by 30deg.
  

nrl 8 MHz calibration ()

For the calibration use vmm with B/A

Using the hfplotsignal  idl program

    I wrote a short idl program to plot the output Efield of the 6 dipoles given various phase offsets.

How the program works:

• It uses Ei=Aix*cos(wt + phi_ix)  + Aiy*cos(wt + phi_iy)   where i goes 1..3 .. for the 3 dipole pairs (1 at each tower).
• Ei is computed for each tower using the amp and phases specified by the caller
• The routine then rotates each E vector  to the angle for the specific tower.
• Rather than using the physical location of a tower, you just use the angle to move the 2nd,3rd towers from the first..eg 0,120,240 (CW) or  -[0,120,-240] CCW.
• For some reason the rotation of the final vector is opposite from what is known from looking at the sky. when the vector rotates CCW this is the phase offsets for lcp, nor rcp.
  

• location: x101/hf/plotphase/hfplotsignal.pro
• some code to drive it in hfphase.pro, linphase.pro
  
• The call is:
  
• hfplotsignal,xPh,yph,rotFrq=rotFrq,thStp=thStp,step=step,posAndDeg=posAndDeg,amp=amp,$
•    xamp=xamp,yamp=yamp
• Arguments:
• xPh[3]  x phase in degrees for the tangential components
• yPh[3] y phase in degrees for the radial components
• rotFrq :  Rotation frequency for the E field vector (for display) .def = .1 Hz
• thStp   : degree step for rotation (def = 10 deg). will recompute  vector every 10 degrees.
• step     : if set then run in single step mode. stop after plotting each vector.
• posAngDeg[3]: What it does:
• The Efield for each tower is computed using the x,y  phase offsets.
  
• The Efield is then rotated into the position of the tower using posAngDeg[3].
• A CW  rotation is positive.
• Normally the angle is measured relative to the tower that has the 0 deg phase reference.
  
• eg:  5Mhz uses towers 1(0deg),3(120 deg),5 (240 deg).
  
• If the x,y entries are towers 1,3,5
• posAngDeg=-[0,120,240].. -120 deg to go from t1 to t3, -240 deg to go T1 to T5
• amp[3]: amplitude for E field applied to both tangential (x) and radial (y) dipoles.
• def: [1.,1.,1.]
• xamp[3]: amplitude to use for x (tangential dipoles). def: amp[3]
  
• lets you see how things degrade (on axis) when amplitudes don't match
• yamp[3]: amplitude for E field applied to  y (radial) dipoles. def: amp[3]

    Running the program will generate the Efield of each tower, and a circle of radius 3 (def for amp=[1,1,1]) with a white vector sweeping around at 10sec/rotation.
The plotted symbols are:

• yellow circle of radius 3 (for def efield amp of [1,1,1]
• + red  symbols. Efield for first tower
  
• diamond ,green symbols: E  field from 2nd tower
• triangle : blue symbols: E field from the 3rd tower
  
• There is 1 symbol a the origin, and at 2nd symbol for the end of the vector
• white line: sum of the 3 E field vectors.
• If no line appears, the 3 vectors are summing to 0.
• The rotation  direction :
• CCW  right circular
• CW: left circular

Running the program:

• cd /share/megs/phil/x101/hf/plotphase
• idl71
• @phil
• @geninit
• look at hfphase.pro as an example of how to call it.
•  Once it starts rotating:
• space bar will stop
• hit again will restart
• hitting q will return to the idl prompt.
• exit to leave idl.

• rcpRot: direction vector rotates
• rcpAct: direction transmitted on the sky..
  

home_~phil