Intro

    isee3 is the old ICE satellite that used AO as the down link back in the 80's when it flew by the comet giacobini-zinner.
It is now coming back toward earth and it is   transmitting two tones.

The setup:

sband wide (dual linears, Tsys about 35K) was used  to look at the isee3 tones.

• The transponder frequencies are:
• Transponder B:  2217.5 MHz (lcp)
• Transponder A: 2270.4 MHz (rcp)
  
• I setup sband wide (dual pol linears) to cover two bands (each centered at 1 of the transponder frequencies)
• no doppler tracking was done during the data acquisition.
• The data acquisition system is locked to the hydrogen maser.
  
• The jpl horizons ephemeris was used to track the satellite (and to compute the doppler shift offline).
  
• The mock spectrometer recorded the data with:
• 12 bit i,q voltage sampling with a 160Khz bandwidth about each frequency
• 300 or 320 seconds of data was taken.
• a 10 Mhz offset was added to the analog mixer and subtracted from the digital mixer to get rid of DC.

Processing the data:

• 09apr14:
•  spectra were computed using 128K long transforms
• 160Khz/128K --> 1.2207 Hz resolution. .819 seconds/spectra
• 366 spectra covered the 300 seconds of data.
• I then looked at portions of the spectra about the received signal
• 09apr14, 14ap14.
• 2 or 3 blocks of 16 million points (about 100 seconds) were processed.
• For each block
• doppler correct the voltages using the ephemeris (no time dilation was done)
• Compute the 16million point   spectra.. giving about 10 millihz resolution
• this .01/2217e6 or 1 part in 4.5e-12 .. so this is probably pushing the accuracy
  
• the transmitted tone
• maybe the ephemeris

specs

• transponder A
• tx: 2270.4 Mhz
• rx: 2090.66
• ratio: 240/221
• oscillator fundamental: 9.46 Mhz
• transponder B
• tx: 2217.5 Mhz
• rx: 2041.9479
• lo for mock mixing of uplink signal
• 2041.9479 - 335.01 = 1706.9378
  

coherent echoes

    Scans when we received echoes from coh , ranging commands

rise,transit,set times

Timeline: (top)

• Note: references to 64bps mode were actually 16bps until cmd sequence corrected on 10jun14
  
• 140409 - record satellites tones 
• 1 320 sec scan
  
• 140414 - record satellites tones
• 2 320 sec scans
  
• 140518 - record satellites tones
• 2 320 sec scans
• 140519 - record satellites tones
• 2 320 sec scans
• 140521 - record satellites tones
• 1 320 sec scans
• then did 600 sec cross patterns moving +/- .1deg in az,za
• saw signal passing through sidelobes at edges..
  
• signal still very weak
  
• 140522 - record satellites tones
• 6 320 sec scans
• 1st scan using normal ephemeris tracking.
• signal strength 15 so 20 sigmas
• used hp spectrum analyzer with 4 hz resolution to look at a 1khz band about the expected tone frequency
• moved telescope in ra,dec offsets till source got stronger.
• went from not visible to 52 db above the noise floor
• ended up with an offsets of:
• -.47deg   in ra
• .08 deg in dec
• This resulted in a J2000  Position of: 07:31:58.9, +22:01:09.3
  
  
• the beam width at 2212Mhz is 128 arcseconds
• So prior to this we were 13.4 beams away from the satellite
• this says we have a really sensitive telescope (or that we have really band sidelobes!).
• 140523 - receive, transmit, receive
  
• tracking tone got a new position offset:
• -.48 deg in ra
• .08 deg in dec
• ra dec j2000:07:34:09.5, 21:56:02.5 J2000 .. this was around 19:15  utc
  
• 320 sec scan tracking satellite
• setup to transmit
• Used sat receive transponder: 2041.9479 Mhz
• Computed transmit signal at 16:00 ast to be 2041.9281 Mhz
• took expected receive freq for 2217.5 transponder=: 2217.519731
• took difference and subtracted this from sat rest freq of 2041.9479 (since xmited signal blue shifted).
• trkinit sbtx model, closed shutters,  moved turret, hooked up modulated signal  into waveguide switch of sb transmitter
  
• transmitted for about 1 minute. then stopped
• if/lo fiber optic transmitter upstairs was saturating.
  
• setup if/lo for sbtx (sbtxmode).. this adds more switch isolation for the transmitted signal to get to the fiber xmiter.
  
• tranmitted for about 5 minutes
• moved to receive
• setup sbw , trkinit sbw, moved turret, opened shutters
  
• 320 sec scan receiving.. looking for sidebands.
  
• none seen.
• Notes: we used the receive ephemeris for xmit.
  
• i checked the difference in the receive position. it varied by 2 arcseconds over two minutes (about 1 round trip time) so the transmit receive positions are essentially the same (within 2asecs).
• 140526 - tracking (but didn't have much time..)
  
• tried peaking up on sat.
• offsets from ephemeris file:  ra: -.47,dec:.08 deg current (around 21:00 utc)
  
• tried calibration scan +/-3beams. only cut source in za strip (3beams off) so peaking up not very good
• Updated the 2d fitting routine.. got a good position fit.
• no timedomain data recorded
  
• 140527 - tracking
• spider scan on satellite
  
• 100 sec recieve with mocks: filename 200?
  
• transmit
• 100 sec receive: filenum 300
  
• 320 sec receive. filenum 400

• operation	duration secs	filenum	Notes
  spider scan		x101.1	a few beams out good fit
  receive	100	200
  xmit
  receive	100	300
  receive	320	400

  
  
• 140528
• The table below shows the sequence of events
• The received data is time stamped to the station clock.
• 160khz bw time domain data is recorded in both polarizations at both frequencies.
  

• operation	duration secs	filenum	Notes
  spider scan		x101.1	corfile
  receive	120	000
  Xmit to 2041.9479 transponder			linked with 2217.5 tone
  xmit
  receive	97	.100
  xmit
  update position from spider scan			about 12 asecs
  receive	28	200
  xmit
  receive	120	300
  Xmit to other transponder			linked with 2270.4 tone sweeping frequency
  receive	211	400
  xmit			regular command sequence
  receive	300	500	modulation received
  receive	300	600
  receive	300	700
  receive	300	800
  receive	300	900

• 140529
• spider scan
• first spider scan had incorrect ra,dec center
• 2nd spider scan ok
• map ra dec center:074718.5  212251.1
• ra,Dec Current offsets used to get center: -.5576,.1018 deg
• ra,Dec error measured:-0.0160,0.0045 deg (add this to map center)
• New ra,dec offsets:-0.5416    0.0973 deg.
• new ra,dec J2000 position: 07:47:14.7,21:23:07.2 @ 140529, 18:53:23 utc
• 
  

• operation	duration secs	filenum	Notes
  spider scan	242	x101.1	start on scan 414900007
  receive	75	0000
  xmit 1			turn telemetry on transponder b (2217.5)
  receive	37	200	didn't turn on . (file 100 was off src..skip)
  xmit 2			turn telemetry on transpB. sweep through freq
  receive	320	300	telemetry did no turn on
  receive	320	400
  updated pointing			tone got 5db stronger
  xmit  2			single request for b telemetry on
  receive	9	500	aborted.. still slewing
  receiver	120	600	no modulation on B
  			I wasn't commanding the telescope to track the source after entering xmit mode.. we were just sitting! So xmits 1,2,3 were not to the satellilte.
  xmit			short sequence to turn on telemetry on transponder B
  receive	320	700	transponder B telemetry is now on.
  receive	320	800
  receive	320	900
  receive	304	1000
  xmit			transponder B from 1kilo bit to 64 bit rate
  receive		1100	transponder B did not change
  receive		1200
  xmit			transponder B to 64bit rate. sweep the frequency
  receive	320	1300	no change transponder B
  receive	16	1400
  xmit			transponder A to coherent mode
  receive	104	1500	no change A
  xmit			retry transponder A to coherent mode
  receive	102	1600	no change A
  xmit			Transponder A to coherent mode. used correct parity
  receive	196	1700	not change A
  xmit			try B to 64bit rate, A to coherent
  receive	320	1800	no change a, b

• 140530
• spider scans
• did multiple patterns
• updated the map centers twice
• receive: two scans of 320 seconds
• 140601
• 2 spider scans
• 140603
• 7 spider scans
• 320 second receive (filenum 00000)
• 140604:
• receiver warm. tsys=150K
  
• 3 spider scans
• 320 sec receive (filenum 00000)
• 140605:
• used S17 ephemeris (with j2000 coord).
  
• 320 sec receive (filenum 0000)
• 3 spider scans. polA was jumping a little.
  
• 140606:
  
• xmit,recieve . used s17 ephemeris, J2000 coordinates, no offsets.
  
• signal strength was 40db above the noise floor in 4.7 hz rbw using spectrum analyzer (same as previous days)
  
•  we tried putting TrA in coherent mode. We ended up turning off the modulation, but the frequency never changed (see tonedrift)
  
• 
  

• operation	duration secs	filenum	notes
  receive	43	00000	before any xmits
  xmit			req: pressure transducers on, Trb to 64bit/sec..
  receive	128	00100	rslt:no change
  xmit			req: telemetry TrA off,put in coherent mode
  receive	82	00200	rslt:: no change
  xmit			req: B to 64 bits/sec mode
  receive	82	00300	rslt:TrB now 64 bits/sec
  xmit			req:Transponder A to coherent mode
  receive	210	00400	rslt: TrA telemetry turned offl. Frequency of tone unchanged
  xmit			req:TrA 500Khz ranging tone with modulation?
  receive	157	00500	rslt: no change
  xmit			req:TrA 500Khz ranging tone.. no modulation ?
  receive	87	00600	rslt: no change
  xmit			req: TrB to 512 bits/sec rate
  receive	7	00700	slt: no change. stayed at 64bits/sec (we may not have waited long enough for the return)
  xmit			req:TrB to 512 bit/sec rate. use narrow sweep
  receive	7	00800	rslt: no change. stayed at 64 bits/sec  (we may not have waited long enough for the return)
  xmit			req:TrB to 512 bits/sec rate. use wide sweep.
  receive	19	00900	rslt: TrB switched to 512 bits/sec rate
  xmit			req:TrB to 64 bits/sec
  receive	47	01000	rslt: no change
  xmit			req:TrB to 64 bits/sec. wide sweep
  receive	36	01100	rslt: no change
  xmit			req:TrB to 64 bits/sec. wide sweep
  receive	36	01200	rslt: no change

  
  
• 140608:
• switched to using 750narrow band ifsetup
• using s17 ephemeris
  
• 3 position patterns
• 320 secs receive. filenum 00000
  
• 640 secs receive: filenum 100
• 140609:
• xmit,receive
• switch bps of TrB
• put trA in coherent mode. send ranging tones
  
• using s21 ephemeris

• operation	duration secs	filenum	notes
  xmt			TrB to 64bps
  rcv	45	00000	no change
  xmt			TrB to 64bps doppler correct +10Khz offset
  rcv	45	00100	TrB switched to 64bps
  xmt			TrB to 2Kbps doppler +5khz offset, send cmd once
  rcv	67	00200	no change
  xmt			TrB to 2Kbps doppler +  10khz offset, send cmd once
  rcv		00300	no change
  xmt			TrB to 2Kbps doppler + 10khz offset, send cmd 3 times
  rcv		00400	no change
  xmt			TrB to 512bps doppler + 5khz offset, send cmd 1 time
  rcv		00500	TrB switched to 512bps
  xmt			TrB to 64 bps doppler + 5khz offset,send cmd 1 time
  rcv		00600	TrB switched to 64bps
  xmt			TrB to 2Kbps doppler + 5khzoffset, send cmd 1 time
  rcv		00700	no change
  xmt			TrA to coherent mode TrB to 2Kbps doppler + 5khz offset, send cmd 1 time
  rcv		00800	TrA .. check TrB switched to 2Kbps
  xmt			cmd TrA to sweep by 125 Khz
  rcv		00900	see plots (10.)
  xmt			TrA phase modulated ranging tones(500,50,5 hz)
  rcv		01000	see plots(11)
  xmt			TrA phase modulated ranging tone (5khz) doppler + 5khz offset
  rcv		01100	no change
  xmt			TrB to 64bps doppler + 5khz offset
  rcv		01200	TrB switched to 64bps

• 140610
• xmit, rcv using ephm s24
• see data table 10jun14 for details
• 140613
• rcv and pointing patterns S24
• 3 320 second rcv
• 3 pointing patterns
• 5 320 sec rcv
• last rcv let sat drift out of beam for 320 sec. look at far out sidelobes (no images made of this scan)
• 140615
• xmit,rcv ephm s27
• tried recording the xmitted signal.. but i got the 10Mhz offset backwards on the lo so the recorded data was out of band.
• about 25 xmit rcv cycles. see data table 15jun14 for details
• 140620
• xmit, rcvm ephm s27
• recorded the tx signal
• got coherent mode reply
• 140623
• 3 pointing patterns.
• 140624
• xmit, rcv
• used ephm S32.
• 140625
• xmit,rcv
• used ephm s32
  
• 140627
• used ephm S32.
  
• xmit, rcv looking for  optimum freq offsets,modulation index using dummy command
• 140701
• used S32
• xmit, rcv
  
• 140702
• used S32
• used Trasponder B
• successfully spun up the satellite.
  
• 140708
• tried S33 (with dsn doppler). signal 20 db low. pointing off by about 200 asecs
• used S32 for runs
• tx,rx to satellite (mainly tx).
  
• 140709
• used S32
• xmit to satellite
• did 3 pointing patterns
• 140715
• used S32
• did 3 pointing patterns.
• 140819
• did 3 pointing patterns
• s41 ephemeris was off by about 22.5 amin, so did not use
• spent some time finding ra,dec offsets from s41.
• ended up using ra 230639.3, dec 130840.4 J.
• 140918
• Used s45 ephem, but did not see the signal.. what we did
• Start 23:30 ast 17sep14. started near transit
• ephem vel of 2.4km/sec gave offset of +18Khz
• used 50khz span about cfr + 18khz on spectrum analyzer
• sbw beam size is about 119 asec (fwhm).
• did steps of .05 deg in ra,dec 
• dec .05deg= 1.64 beams
• ra .05deg step (at dec=13)=.1.59 beams
  
• 2217.5: tracked ephem. saw nothing. switched to 2270.4
  
• 2270.4
• we covered a  1 degree grid (+/-.5 degrees from the ephemeris center position)
• inside of .35 degrees the grid step was .05 deg (1.6 beam)
• outside of .35 degrees we used a .1 deg step (3.2 beams).
  
• we saw no signal
• outside of .25 degrees offset we switched to use two spectrum analyzers so we had both frequencies displayed
• What could have gone wrong:
• both transponders are off
  
• 2270.4 is off and the offset is less than .25 degrees from ephemeris
• the ephemeris is off.
• The signal is too weak for us to see?
• on 27jun14 the satellite was at a similar distance.
  
• We saw the modulated carrier at 45 dbn above the noise floor in 1.2 hz
• we probably had 200 hz rbw. this would make the signal 10db closer to the noise floor. If we are in a null of the antenna patterns we might have another 10db less. This would still give us 25 db above the noise floor. So if we pointed at it we should have seen it.
• When we had a 45 db signal, the .1 deg stepping would easily see the signal, at only 25 db, the course stepping might have made it a bit harder.. our 1st sidelobe is 15 db down...
  
• After the satellite set, we continued with cband calibration runs. The pointing looked normal (errors < 10asecs)
• Morehead state did see the signal on 16sep14. it was weak in a 9.2 hz rbw.   they have a 20 meter dish so their beam probably about  .34 deg
• So my guess is that the signal is probably .3 degrees or more from the ephm. For the larger offsets we probably needed to do finer steps.
  

Measured positions:  (top)

What exactly is being measured....

• the rms pointing error from the model and checks is about 6 arcseconds
• the pointing model error residuals in az and za are not completely random.
  
• There are some az,za that can have 8 to 10 arcsec errors that are repeatable.
  
• The model errors at low za  (<4 deg)  tend to be larger:
• there are not as many input points to the model around this area
• the azimuth has to move a lot faster to track sources here.
• For the errors in the satellite position, i guessed that the pointing error might be 10 arcseconds (6seconds rms plus 4 arcsecs from fit, other junk)
  
• For some azimuth and za's this is probably too low.
• The following days had some difficulties:
• 22may14: the satellite position was obtained by peaking up the ra,dec offsets while tracking the source. No pattern was made
• 26,27. The ra,dec offsets used for the pattern we a few beams away. Some of the strips did not cut close to the mainbeam.. The fit was probably not as accurate as when all 4 strips cut through the main mbeam.
  

• this used data through 03jun14.
  
• the 2nd order fit will show outliers over a short period of time.
  
• Page 1: measured ra,dec positions measured vs day of month in may.
• daynumber is utc.  > 31 are days in june.
• the red line is a 2nd order poly fit to the positions (using no weighting).
• Page 2: measured - fit positions. units are great circle arcseconds.
  
• top: black is ra difference, red is dec difference
• bottom:i added the errors in quadrature.
• Page 3: az,za positions for the center of each pattern .
• Page 4,5: blowup showing data from 30may14
• the satellite was transitting around za=3 for points 3,4,5
• 
  
• The gradual increase/decrease in the errors occur because of the 2nd order fit.
• within 1 day, the spread in positions probably gives an idea of the position accuracy.
• The reported errors are relative to the current pointing model. If the model does a poor job of fitting parts of the dish, then these errors would need to be added to the reported positions.
• I should probably track  a source with a similar ra,dec  and subtract these errors from the reported satellite positions (unfortunately i've got a lot of other junk to do...).

• The arc second error is a guestimate. It is great circle.
• Peaking up using manual offsets:
• ra,dec offsets added to maximize the received tone
• not real accurate.. probably 30 to 60 arcseconds.
• Peaking up using spider scans
• a spider scan pattern uses. 2 crosses rotated by 45 degrees from each other.
• each leg of each cross is +/- 3 beams
• The system is driven through each leg of the cross while data is recorded at a 1 hz rate
• This is the normal method used to generate the telescope pointing model
  
• 2d gauss fits were then done to the frequency channel containing the tone.

• 140522
• used manual offsets to peak up on source.
• Got 52 db above the noise floor (in 4hz) so we were probably pretty close (maybe 30 to 60 asecs)
• 140526
• spider scan pattern used. 2 crosses rotated by 45 degrees from each other.
• Source was a few beams out.
• the error was probably 15 asecs
• 140527
• spider scan pattern
• source was about a beam out
• error probably 15 asecs
• 140528
• spider scan
• source was within 12 asecs of initial position
• error about 10 asecs
• 140529
• 1 spider scan
• 140530
• 11 spider scans
• 140601
  
• 2 spider scans
• 140603
• 7 spider scans
• 140604
• 3 spider scans
• 140605
  
• 3 spider scans. polA jumping a little
  

09apr14- looking at the transmitted tones.  (top)

• 300 seconds of data was taken. this allowed for 2 spectra of 16 million points.
  

Looking at the data 2217.5 MHz tone:

• dynamic spectrum of the signal (.gif)
• The horizontal axis is Hz Offset from 2217.522480 MHz. This turned out to be the center of the returned signal
• the vertical axis is seconds after 20140409 19:49:57 (UTC).
• the drift to lower frequency is from the doppler. It decreased by about 60 Hz over the 300 seconds.
• Line plots of the individual spectra (.ps) (.pdf)
• page 1: over plot the 366 spectra
  
• the horizontal scale is Hz offset from 2217.522480 MHz)
• I've shown +/- 100 Hz about the center of the return
  
• the vertical scale is in sigmas
  
• Top: polA,
• Bottom: polB
• Page 2: over plot spectra after doppler correction.
• I used the horizons ephemeris to compute the velocity and then doppler shift for each spectra.
• The doppler vel ranged from -4.0035 to -3.995 km/sec
• A negative velocity is coming toward us, so the carrier was shifted to higher frequency
• i computed and then  removed the doppler shift from the x axis of each spectra
• Top: over plot the spectra after doppler correction.
• I've averaged the pols.
  
• The vertical scale is arbitrary power units
• After doppler correction the spread over the 300 secs is a few Hz
• Individual spectra did not resolve the carrier in 1.2 Hz resolution
• After doppler correction, the carrier is offset -7102 Hz from 2217.5 MHz.
  

Looking at the data 2270.4 MHz tone:

• dynamic spectrum of the signal (.gif)
• The horizontal axis is Hz Offset from 2270.419746 MHz. This turned out to be the center of the returned signal
• the vertical axis is seconds after 20140409 19:49:57 (UTC).
• the drift to lower frequency is from the doppler. It decreased by about 60 Hz over the 300 seconds.
• Line plots of the individual spectra (.ps) (.pdf)
• page 1: over plot the 366 spectra
  
• the horizontal scale is Hz offset from 2270.419746 MHz)
• I've shown +/- 100 Hz about the center of the return
  
• the vertical scale is in sigmas
  
• Top: polA,
• Bottom: polB
• Page 2: over plot spectra after doppler correction.
• I used the horizons ephemeris to compute the velocity and then doppler shift for each spectra.
• The doppler vel ranged from -4.0035 to -3.995 km/sec
• A negative velocity is coming toward us, so the carrier was shifted to higher frequency
• i computed and then  removed the doppler shift from the x axis of each spectra
• Top: over plot the spectra after doppler correction.
• I've averaged the pols.
  
• The vertical scale is arbitrary power units
• After doppler correction the spread over the 300 secs is a few Hz
• Individual spectra did not resolve the carrier in 1.2 Hz resolution
• each minor tick mark on the horizontal axis is 5 Hz.
  
• After doppler correction, the carrier is offset -10541.5 Hz from 2270.4 MHz.

Doppler correcting the voltages:

• Do the following for the data set for each tone
• input and doppler correct 100 secs of voltage.
• interpolate the ephemeris velocities (spaced by 1 minute) to the time of each sample (160Khz bandwidth)
  
• compute the power spectrum.
• There were 2 100secs of data for each tone.

• Page 1: 2217.5 Mhz tone.
• black is the first 100 second spectra. red is the second 100 second spectra.
• The vertical scale is sigmas on the noise of the power spectrum.
• The central peak is -7102.21 Hz below 2217.4 Mhz:. The horizontal scale hz offset from 2217.49289779 Mhz
  
• Top: Pol A spectra
• Middle: PolB spectra
• Bottom: average the two 100 second spectra and the two pols.
• The blue dashed lines are .311 hz from the center of the plot
• Page 2: 2270.4 Mhz tone
• Top:polA
• middle: polB
• bottom: polA + polB and average the100 sec spectra.
• the spectrum moved by about .12 hz between the 2 100 spectra.
• I don't know if this is a problem with my doppler correction, or the transmitter is moving around.
  

09apr14 Summary:

• Data acquisition started at 20140409  19:49:57 (UTC)
• After doppler correction the spread over the 300 secs is a few Hz
• Individual spectra did not resolve the carrier in 1.2 Hz resolution
• sensitivity:
• bw*tau for the single spectra is 1, so 1sigma = Tsys which is about 30-35Kelvin.
  
• 2217.5 carrier;
• frequency offset -7102.209 Hz below 2217.5 MHz
• about 50 sigma in .82 secs per pol
• 2270.4 carrier
• frequency offset: -10541.55 Hz from 2270.4 MHz
• about 40 sigma in .82 secs per pol
• doppler correcting and doing a 100 second spectra showed sidelobes at about .31 hz.
  
• The 2270.4 xmiter moved by .12 hz between the two 100 second spectra. I don't know if this is real or a bug in my doppler correction (but the  same code on the 2217.5 Mhz tone had no shift).

14apr14: looking at the tones.  (top)

• two blocks of data of 320 seconds were taken
• blk 1 started at 20:07:46 utc
• blk2  started at 20:59:44 utc

• The 3 spectra of each block are colored black,red, and green
• The vertical scale is sigmas on the noise.
• For each page:
• top: polA. 3 spectra overplotted
  
• middle : polB. 3 spectra over plotted
• bottom: average of the 6 spectra (3 for each pol times 2 pols).
• Page 1: 1st block starting at 20:07:46 utc. 2217.5 Mhz tone
  
• The return is centered -7101.68 hz from 2217.5 Mhz
  
• Page 2: 1st block starting at 20:07:46 utc.  2270.4 Mhz tone
• the return is centered -10526.28 hz from 2270.4 Mhz
  
• Page 3: 2nd block starting at  20:59:44 utc. 2217.5 Mhz tone
  
• The return is centered -7101.94 hz from 2217.5 Mhz
  
• Page 4: 2nd block starting at  20:59:44 utc.  2270.4 Mhz tone
• the return is centered -10525.81 hz from 2270.4 Mhz

27may14: how fast is the ra,dec position changing in 1 run. (top)

• top: ra,dec change vs hour of day
• black: rachange ,red dec change. Both arcminutes
  
• the rachange is little circle (not multiplied by cos(dec).
  
• I've removed the mean ra, mean dec from the plots
• bottom: i added the ra delta (great circle) and the dec offset in quadrature. (sqrt ((radelt*cosdec)^2 + decdelt^2)
  
• The satellite is moving .98 arcminutes (great circle) per hour
• This is about 1 arcsecond per minute of time.
  

• we can have the telescope track the ephemeris file during the pattern
• or we can have the telescope track a fixed ra,dec that includes an instantaneous offset we measure
• In this case the 5 minute calibration scan will have the fixed ra,dec smear by about 5 arc seconds from the start.

09jun14 - satellite responds to coherent mode request

• The each spectra is .7 seconds. +/- 2.5 khz is displayed
• the carrier started moving about 38 seconds into the receive scan

• At 20:14:12 UTC we started receiving the signal.
• 13 seconds of data was used starting 36 seconds into the receive cycle.
  
• Each spectra was 12.8 millisecs of data
• The data was doppler corrected before the spectra were computed (using S24 ephemeris)
• The carrier started at 2270.390697 Mhz (red line)
• around 38 seconds the carrier jumped to 2270.403588 Mhz
• at 42 seconds, a low level modulation started. this lasted for 1.6 seconds
• at 43.7 seconds 7.5 Khz sidebands appeared on the offset carrier. They lasted for 2.4 seconds
• at 46.2 seconds, the sidebands stopped and the low level modulation appeared again. This lasted for 1.7seconds
• at 47.9 seconds the offset carrier jumps back to the original carrier frequency

• top: average spectra from different parts of the receive cycle
• The vertical scale is in units of Tsys
  
• black: before the carrier jumped
• red: carrier has jumped, before modulation starts
• green: shifted carrier with modulation (7.5khz offset) on.
• Bottom: The total power from polA and polB transponder A.
• The total power was computed for the voltage sample  during the 13 seconds of the above images.
• the receiver is native linear polarization
  
• Black in polA, red is polB
• I've scaled the veritical axis to Tsys. We can assume that Tsys for polA and polB is close.
• The power has a 1.28 hz periodicity
• polA and polB are 180 degrees out of phase
• this period is 4 times the rotation period of the satellite
• This may be from the 4 sets of antennas rotating into the AO direction
• This could occur if the off axis polarization did not remain circular.
• need to check this on the data from the other transponder (but i think i've see it there).
  

Summary:

• The satellite responded to the coherent mode request
• the received signal power oscillates in each pol at 4 times the rotation frequecy of the satellite
• Pola,polB oscillation is 180 degrees out of phase.
  

15jun14 sat responds to cmd sequence:  modOff, cohOn, rngOn

• modulation off TrA
• coherent mode on TrA
• ranging mode on TrA
• the transmitted frequency was nominal + doppler + 11.125 Khz offset

• The each spectra is .7 seconds. +/- 2.5 khz is displayed
• the carrier started moving about 39 seconds into the receive scan

• At 19:54:34 UTC we started receiving the signal.
• 13 seconds of data was used starting 36 seconds into the receive cycle.
  
• Each spectra was 12.8 millisecs of data
• The data was doppler corrected before the spectra were computed (using S27 ephemeris)
• The carrier started at 2270.3908884 Mhz (red line)
• around 38 seconds the carrier jumped to 2270.408959 Mhz
• it was unchanged for 4 seconds (this is the time between cmds)
• (I remeasured these freq to about 1hz resolution.)
  
• at 42 seconds, a low level modulation started. this lasted for 1.6 seconds
• There are 4 seconds between commands
• a command is 2.4 seconds long
• the carrier then drops for 1.6 seconds. this is the low level modulation period.
• a command lasts for 2.6 seconds. The
  
• at 44 seconds the offset carrier jumps back to the original carrier frequency

• the horizontal dashed lines are my best guess at which data packet the satellite is processing.
• The s/c is locking onto the cmd for the 2.4 seconds of the cmd bits.
• During the 1.6 secs when the carrier drops out, the satellite carrier starts to wander.

• Assume that we have used the current ephemeris file to doppler correct the data.
• Since the ephemeris is not perfect, there will be a freq offset
• ephErrL = ephemeris error at 2090.66 Mhz
• ephErrH = ephemeris error at 2270.4 Mhz
• The ratio of these two H/L = 2270.4/2090.66 = 1.08597 = Fratio
  
• The normal tone we receiver is 2270.4 + oscErr +  ephErrH
• The freq transmitted to the sat for ranging cmd was:
• 2090.66 + doppler cor + 11.125 Khz
• The freq the satellite rcvd (after locking) was:
  
• satRx=2090.66 + 11.125Khz + ephErrL
  
• The satellite then echoed this back at 2270.4 mhz.
  
• We can assume that there is no oscillator error since it locked to the AO carrier.
•  satTx= 2270.4 + 11.125Khz  + ephErrL (assume the conversion is done by mixing instead of multiplying)
  
• AoRx= satTx + ephErrH= 2270.4 + 11.125khz +  ephErrH + ephErrL = 2270.4 + 11.125Khz + ephErrH(1+1/fratio)
  
• AoRx=2270.408959D
  
• ephErrH=(AoRx - 2270.4d - 11.125d-3)/(1.+ 1/fratio)
  
• Ephemeris error:
• at 2270.4 Mhz = -1.128 khz
• at 2090.66      =  -1.038 khz
  
• velErr= .149; km/sec 
• The frequency wants to be smaller. so the ephemeris negative velocity (toward the observer) is too negative by .15 km/sec
• We can then use the ephemeris error at this time to compute the oscillator error
• Sat transmits telemetry data at:
• satTxH= 2270.4 + oscerr
• aoRxH = satTxH + ephErrH = 2270.4 + oscerr + ephErrH
  
• aoRxH=2270.390884
• oscErr=aoRxH - 2270.4  - ephErrH = -7.988 Khz
  

Summary:

• the satellite responded to the command sequence by jumping in freq by 18.13 Khz
• The telemetry data stated that all 3 commands were accepted.
• After this we sent various frequency sweeps, ranging tones. We saw no change in the carrier
• there could have been short jumps at the beginning of the scan that we could have missed since we didn't always catch the start of the return echo.
• this is similar to what occurred on 09jun14
• The time before the low level modulation turn on was about 4 seconds. This is similar to the 09jun14 time.
• No sidebands occurred about the offset carrier (unlike 09jun14).
• I wonder if the 1.6 seconds of low level modulation is when the sat rcv is trying to demodulate the rcved signal , before trying to remodulate  the echoed signal?
  
• We have only gotten carrier responses during the commands: telemetry off, coherent mode on, ranging on.
  
• It looks like the satellite is seeing the 3rd command as the coherent signal to lock onto,.
• It tried locking, when this carrier goes away (at the end of the command sequence), it stop tracking.
• The question is whether the satellite remains in coherent mode/ ranging mode after this sequence
• You could test this by sending another command (later) with a similar 11.125 khz offset. If the return jumps, then you are still in coherent mode..

• The satellite is interpreting commands as tones (probably in coherent mode)The low level modulation that is seen,  lasts for the same time that the 4 second command drops the carrier (1.6 seconds).
• The computed ephemeris/ oscillator errors at 140615: 195514 utc
  
• ephmErr .149 km/sec  (ephem velocity too negative) 
  
• this is the error after applying ephemeris S27.
  
• oscillator error: -7.988 khz
• this is the error for 2270.4Mhz.. not sure how the sat generates the two frequencies.
  
• I re measured the received frequencies using a 1.2 hz resolution fft. so the freq resolution is close to 1 hz.
  

the data  (top)

    The table below has the processed data for each scan of each day .
Each scan contains:

• 320 seconds of data
• Recording setup:
  
• 12bit A/Ds complex sample the input at 160Mhz.
• the input signal has been offset by 10.x Mhz to move the tones signal away from baseband 60Hz problems
• digitally mix to put the tones + .01mhz at the center of the band
• digitally decimate to 160 khz
• record polA,polB 16 bit i,q data for each transponder.

• Column 1: plots showing 390 .8 sec spectra
• freq  resolution=.1.2 hz.
• page 1 overplots each spectra. this shows the doppler drift
• page 2: overplot the spectra after shiftting each spectra by it's doppler shift.
• Column 2:dynamic spectra:
• polA and b are averaged by .8sec spectra
• dynamic spectra are made for the scan
• The images are scaled so the lut covers +/- 3 (or 5) sigmas.
  

Notes

• 140522: updated pointing before  run 2.
  
• added: ra offset  -.47 degrees, dec offset .08deg
• Our beam is 2 amin or .033 degrees. so prior to this we were about 15 beams from the satellite.
• the signal strength in .82 secs went from 50sigmas  to 52db sigmas (158000 sigmas) in .82 seconds.
  
• plots are db above 1 sigma
• images have 60 hz harmonics. they are mixing with the tone in our IF.
• the baseband sampled cfr is shifted about 10.01 Mhz from the tone so DC from baseband won't show up in the plots.
• the 2270.4 transponder has a weak signal about 100 hz from the tone.
  
• It could be some other signal mixing in our IF mixing with the tone (sbn post amp chassis birdies?)
• 140523: the dynamic spectra are +/- 310 hz.
• I searched over +/-80Khz but saw no (strong DIf we use a fixed ra,dec for a ) sidebands).
  

Summary  (top)

• ra,dec ephemeris position moves by about 1 arcs per minute of time.
• The received transponder frequencies (or the 260 Mhz receiver values) can be found using compfreq.pro
  

• there are sidebands of .31 hz on the 2217.5 tone and .34 hz on the 2270.4 Mhz tone
• are these two antennas that are mounted at different angles relative to the spin axis of the satellite?
• the 2270.4 tone seems to be moving a lot more than the 2217.5 Mhz tone.l
  

 home_~phil