Links
230111: tests with box as delivered
230328: tests after osvaldo worked on making the box rfi tight.
summary

Intro

    The 12meter wideband receiver has a dewar monitoring box that will be located close to the antenna. The rfi emitted from the box was tested  in the screen room prior to mounting in on the 12m.

• 11jan23 : box as delivered by hamdi was measured in the screen room.
  
• It showed lots of rfi when the  box was closed.
• mainly coming from holes in the side of the box  and the power cord
• 28mar23: Osvaldo worked to make the box rfi tight
• feed through connectors used,
• wide band power line filter installed
• ethernet fiber optic transceiver switched from 1Gbit to 100Mbit.
  

---

230111: initial testing in the screen room.

    The box was located in the screen room for testing. We looked at the box as delivered by hamdi.

• The lid to box gasket is rubber rather than metallic
• the holes for the cable coming out of the box are open.
  

Setup:

• box was measured first with the lid open, and then with the lid closed.
• this let us first see which freq would have the strongest rfi.
• The photo shows the setup with the lid open (.jpeg)
• box is open
• the vacuum gauge is also connected. It is outside the box.
• The door was  closed and the lights were off when the data was taken.
  
• antenna:
• 1-10 ghz log periodic
• 26 db gain amp at antenna output (800MHz to 21GHz)
• cable to field fox spectrum analyzer (located outside of the screen room).
• Field Fox setup
• spectrum analyzer mode.
• 300 MHz span, 30Khz rbw, 10001 pnt, pos  detector, preamp on, attn=0
• step cfr 2400 MHz to 9900 MHz (26 steps of 300MHz each)
• average 10 sweeps. record avg, pkHold,minHold
• datataking
• directory: : x101/rfiscrn/ffbwrcv.dat
• ffscrn cmd_ffwbrcv.dat
• to take data
• output:
  
• /share/phildat/wbrcv/scrn/
• dwm_opena_230111_nnnnn.dat    nn 0 ->25   lid open
• dwm_closea_230111_nnnnn.dat    nn 0->25 lid closed
• each file: 32bit float binary.
• 10001 points
• 1 freq reg (in MHz
• 3 traces (avg,pk,minhold)

Looking at the data

    The data for each measurement (lid open, lid closed) was first read in.
The  26 spectra of  10001 points,  300MHz span were concatenated into 1 large spectra.
The 10 sweep average was used for the plots.

    The first  plots show the spectra  (.ps) (.pdf)

• Top:lid open
• Vertical scale dbm input at field fox (after 26 db external amp).
• black:spectra
• red:  spectra smoothed with a 1001 length median filter (30MHz).
  
• this is used for the the noise floor of the measurement.
• Top: lid open
• bottom:lid closed.
• the step in the spectra at 7.5 GHz is probably the band step of the field fox.
• the noise floor is being set by the field fox noise floor.

    The 2nd plots show the spectra divided by them smoothed noise floor (.ps) (.pdf)

• Page 1:
• The spectra were divided by the median filtered version. The plots then show db above the noise floor.
• the red + are all points 6db above the noise floor when the lid is open.
• top: lid is open
• bottom: lid is closed
• There are still many points at least 5db above the noise floor after the lid is closed.
• the red +  still show the points that were 6db above the noise floor when the lid was open.
• Page 2: blowup in freq:
• Top: lid open
• Bottom: lid closed.
• the red + now show points that are 10db above the noise floor when the lid is closed.
• they are plotted in higher freq resolution on the following page.
• Page 3: Freq blowup of points 10db above the noise floor when the lid is closed.
• There are 7 frames displayed
• Each frame is +/- 100Khz about the center peak
• The freq (in GHz)  of the center peak is printed in red.
• You can blowup the rbw around these points when searching how these come out of the box.

SUMMARY:

• There are many points 10db above the field fox noise floor when the lid is closed.
• We need to see how these are getting out and then improve the shielding.

Resolution:

•  First set of probing showed that the stronger rfi is (2.500070Ghz) is coming out of the holes in the side of the box from the ethernet switch.
• Turning the switch off made them go away.
  

processing: x101/230111/dwmon_rfi.pro

---

28mar23: after Osvaldo worked on the box

    The same tests were done on 28mar23 after Osvaldo made changes to the huffman box.

•  hole in side of the box converted to shielded connectors
• conductive lining between top and box installed to give better contact
• wide band power line filter installed.
• eithernet fiber transciever switched from 1 Gbit to 100Mbit
• this made the biggest difference.

Setup:

    The setup and acquisition was identical to the 11jan23 measurements

• x101/rfiscrn/  holds the config files:
• ffscrn to run program
• cmd_ffwbrcv.dat for field fox config
  
• the data was written to /share/phildat/wbrcv/scrn/dewarmon_230328/

Looking at the data:

    The first plots show the averaged spectra with the box open and closed (.ps)  (.pdf)

• Top: average spectra box open
• black data
• red: data smoothed to 30 Mhz to use as noise baseline.
• Some values are  > 20db above the noise floor.
  
• Bottom: average spectra with the box closed
• nothing shows up above the noise floor.

    The 2nd plots show the spectral density normalized to the noise floor of the field fox (.ps)  (.pdf)

• Top: box open
• Bottom: box closed
• the red + are values that were >= 6db above the noise floor when the box was open

processing: x101/230328/rfidwmon/dwmon_rfi.pro

---

Summary:

• The box as delivered by hamdi  had lots of rfi with the box closed.
  
• After Osvaldo worked on the box, with the box closed there is no rfi sitting above the field fox noise floor (-100dbm in 30Khz rbw)
• Even with Osvaldo's changes, we could still see some rfi spikes with the box closed when the 1Gbit fo xciever was used. When we switched to 100Mbits, these final spikes fell below the noise floor.