notes.txt

Important numbers
-----------------

LWA
N 79 deg 24.898'
W 090 deg 44.921'

MegaMIST
N 79 deg 24.895'
W 090 deg 44.879'

pol00 = EW
pol11 = NS

Radials: 8 total, 2.9m length

Nivek's recommendation is ~4.5 bits at the ADC, which roughly
corresponds to -3 dBm at ADC input

---------------------------------------------

All times below are local MARS time = CDT = UTC-5


20 July 2019
------------

21:30
System powered on, baseband_autorun will start at some point later
after power off

~16:00
A few 4-bit baseband runs capturing shortwave on LWA

7:30
Radio sked as usual.  Power off overnight.  Turned off system around 7:50.


19 July 2019
------------

21:30
System powered on, baseband_autorun set to start in about an hour.
Not sure if power will still be on then.

7:30
Radio sked as usual.  Power off overnight.  Meteor scattering event last night!


18 July 2019
------------

* 2-bit baseband from Nivek: "I just worked back from the autos of
  pol00 trying to undo it to and got the number 280, which was then
  tweaked so that the 2nd bit was flicked ~40% of the time."

10:00-16:00
Hiked to site #2 for extended RFI survey.  Sharky on NS, LoWavz on EW.
12:40 - first run on first battery, ethernet cable plugged in at the beginning
13:44 - second run on second battery, ended roughly 14:40

7:30
Turned off SNAP.  DAQ wasn't running, must have crashed early in the
night.  Powered on briefly again around 7:55 to confirm that
baseband_autorun.py works (it does).


17 July 2019
------------

Power off overnight

7:05
Copied data from overnight LWA run, turned off SNAP.  Power turned on
at base.


16 July 2019
------------

* New marion-tools pulled, changed <I to >I in spectrum number reading
* New quad_input_poco_gbe_2019-07-15_2121.fpg pulled
* Change file size and drive safety factor for baseband_autorun on RPi

Full power off at night

19:20
Powered SNAP on, LWA is plugged in as normal without extra length of
coax.  Power is still on, and generator is going.

6:55
Automated baseband tests overnight with uneven LWA.  Stopped
baseband_autorun.py, copied data, powered system off.  Note: 4 files
1563224097.raw
1563224233.raw
1563224321.raw
1563225075.raw
written to MARS2 before we ran into safety factor -- manually copied
these to MARS1.


15 July 2019
------------

20:55
Turned system on to let Trimble settle down

20:30
Added extendo section to LWA NS for phase shift tests overnight

------------------------------
Notes from Nivek:

spectrum number is: it a count of the spectrum coming out of the
fft. So depending on the number of spectra in one packet (should be
5), the number should increase by 5 every time. If it's not then,
there's packet loss.

to fixed the packet loss, i increased the buffer size. To do this, you
need to edit the /etc/sysctl.conf and to the bottom of the file change
net.core.rmem_default and net.core.rmem_max to 536900000, where
536900000 is the buffer size in bytes.

try running udp_stats.sh that's in the old_daq folder to check if
there's packet loss. If there is the last number in the output should
be changing

MIST should have 12% lower scale facor than LWA
------------------------------

6:50
Plugged in ethernet cable to stop DAQ, pull data. Power back on,
turned system off.


14 July 2019
------------

22:00
Full power off overnight

20:00
Started dump_spectra.py, LWA plugged in.  SNAP powered on about 30min prior.

7:30
Stopped DAQ and powered off

6:13
Small inverter at the kitchen turned off, turned back on around 7:30am


13 July 2019
------------

Full power off for the night

15:20
Power back on

15:00
Full power off while the humans are away.  Baseband tests:
- 15:04-15:06 One bit with extended frequency range (884-1376 = 54-84 MHz)
- 15:12-15:14 Four bit with channels 884-1004, shift at 2^27
- 15:17-15:20 Four bit with channels 884-1004, shift at 2^28

14:17
Powered SNAP system on to give Trimble some time to settle

14:00
Disconnected MegaMIST, connected LWA NS with extra ~2m coax section

8:00
Powering off SNAP to pull data

7:30
HF on for radio sked

6:20
New small inverter + solar system has been installed in the kitchen as
of yesterday.  Small inverter turned off briefly at 6:20, will be
turned on again around radio sked.

Full power off overnight


12 July 2019
------------

Night...
Added more 3dB attenuators to MegaMIST so that the full front end
chain is ZFL500LN + 3dB + ZFL500LN + 3dB + bias tee + 3dB.
** NOTE: one 3dB att was on the RF+DC side of the bias tee **

19:35
MegaMIST connected again.  Front end is bias tee + 2 x ZFL500LN + 3dB
att, piped into SNAP NS input.

19:26
4-bit baseband on uneven LWA, power on at base

19:00
Power off for radio sked, took 1-bit baseband for uneven LWA during quiet period.

~14:50
Ended 1-bit baseband dump.  "config12.ini"

~14:15
Reconnected LWA with extra ~2m coax section on NS leg.  Taking 1-bit
baseband and dump_spectra.

11:40-13:00
RFI surveying at 2 sites (#1 and #8).  LoWavz on EW input (pol00),
sharky on NS input (pol11).

11:00
Disconnected MegaMIST, reconnected LWA NS with additional ~2m coax
section to introduce phase shift.  Attempting dump_spectra and 1-bit
baseband.

~7:00
Power back on.  Turned off system to pull data.


11 July 2019
------------

~23:00
Full power off overnight.  Big inverter was "broken" because the
switches are completely non-obvious.

----------------------------------
Baseband notes from Nivek:

From the previous plots of mist/lwa configuration it looked like
fringes started to occur ~60Mhz. So maybe try using a frequency range
of 60 to 67 Mhz which are channels 982 to 1102. You should also plot
the histograms to check if you are using all of the four bits and
shift up accordingly.

You will need to write the integer number of the shift, so for eg if I
want to shift up 4 bits: 2^4=16 and the parameter channel_coeffs in
the config.ini would look like 982:1102:16

from lab tests with ~4.3 bits being used, a shift of 2^24 worked
best. Maybe 2^25 or 2^26 might work for you.

For 54-61 MHz, channel range is 884:1004

Frequency/channel mapping is freq(Mhz)/125.00)*2048, but...
for one bit we can use 492 channels, two bit is 246 channels and four bits is 120 channels
----------------------------------

~21:00
Added amplifier to MegaMIST to boost signal level, front end has bias
tee + 2 x ZFL-500LN.  Signal level is a little too high, added a
couple attenuators.

~19:00
MegaMIST up and running again with bias tee + 1 x ZFL-500LN on the
front end, piped into SNAP NS input.  Various baseband tests
afterwards with power off at the base (Taj driving), but lots and lots
of helicopter and ATV traffic.  Data will be a mix of good and shit.

15:00 and onward -- various baseband tests with different
configurations and power off at the base.  Taj is driving, so he has
all the notes.

13:08-13:10
config4.ini

13:05-13:07
config3.ini

13:02-13:04
config2.ini

12:58-13:00
config1.ini

12:55-12:57
config0.ini

12:54
Powered back on to do some quick baseband tests, LWA plugged into both channels

11:39
Powered off SNAP box to charge batteries

~11:28
Killed baseband DAQ.  ~7.1G in ~12 minutes running.

~11:20
Power was off briefly, might be useful to check histogram levels

11:16
First baseband test started, writing 54-61 MHz at 4 bit.
Inadvertently killed dump_spectra because we didn't stop/restart it
while running config_fpga.

9:50
Stopped DAQ, restarted with latest version but with digital gain
setting lines commented out in albatrosdigitizer.py.  Taking a few
spectra to see if mirroring bug is gone.  Confirmed dump_spectra.py is
fine.

7:15
Power back on


10 July 2019
------------

22:37
Full power off

15:09
Ended survey, packed up stuff but left system cabled and running from
the helicopter

14:52
Laptop plugged in for data checks

14:48
Surveying third site (#7).

14:25
Ended survey, packed up stuff but left system cabled and running from
the helicopter

~14:05
Laptop plugged in for data checks

14:02
Surveying second site (#9).

13:45
Ended survey, packed up stuff but left system cabled and running from
the helicopter

~13:30
Plugged in ethernet cable to check on data for a few minutes

13:22
Surveying first site (#4).  LoWavz on NS input (pol11), sharky on EW input (pol00).

~11:00
Generator on to charge batteries

10:39
Power back on

10:32
Full power off to test sharky + bias tee + 2 x ZFL500LN

10:18
Adding LNA to sharky so that full front end chain is bias tee + 2 x
ZFL500LN.  Fully connected at 10:23.

~10:00
Helicopter landed and took off, there will be lots of shit in the data

~8:32
Power off briefly for radio comms, probably hammered with HF during
this time

8:23
Sharky plugged into SNAP NS input, bias tee + ZFL500 at front end

Full power off overnight


9 July 2019
-----------

19:00
Swapped batteries in metal box, haven't started DAQ yet.  Power off at
base for radio sked, taking measurements with Low Wavz antenna +
handheld spectrum analyzer.

Afternoon...sometime...measurements with Low Wavz antenna + handheld
spectrum analyzer while power is on, generator going

14:29
Power off Low Wavz antenna, LWA NS plugged back in, power back on at
station

14:20-14:24
Power still off, adding another ZFL500 (low noise version) to Low Wavz
at front end

14:14
Full power off, Low Wavz still piped into SNAP box at NS input

13:58
Unplugged LWA NS at SNAP input, plugged in Low Wavz antenna with bias
tee + preamp at front end, preamp switched on.  Power is on at the
base, and generator is going.

10:48
Switched power switches back on, also started generator for the building

10:42
Turned all building power switches off for a few minutes to take quiet
data.

10:33
Lost power in the building, decided to pull the box and remove
attenuators.  2nd stage chains are now 2 x ZFL500 + 2 x 3dB att.
Reinstalled box, config_fpga reports 3.5 bits at ADCs.

7:30
Usual radio sked, followed by power on


8 July 2019
-----------

22:45
Full power off

21:35
Restarted system with extra ZFL-500 amplifier + 3 x 3dB attenuators in
each chain.  LWA antenna plugged into both inputs.  Running older DAQ
version without digital gain mods.

20:07
Spectra from quiet period look completely messed up -- seem to be
horizontal mirror images around the clock line.  Reverted back to
previous DAQ version, LWA antenna still plugged into both channels.

20:00
Power back on

19:53
Unplugged bias tee + 50 ohm terminators, plugged LWA antenna back in
to both channels, power off briefly at 19:54.

19:30
Powered box back on, EW chain has extra ZFL-500 amplifier + 5 x 3dB
attenuators.  Using latest DAQ with digital gain set at 1.  Plugged in
LWA inputs briefly, then plugged in bias tee + 50 ohm terminators
(around 19:38).

18:00
Tried a bunch of DAQ shit for the past hour and got grumpy.  Powering
off now to squeeze amplifiers into box.

#--------nivek's new daq changes------
add a completely untest code for setting the adc gain

1, 1.25, 2, 2.5, 4, 5, 8, 10, 12.5, 16, 20, 25, 32, 50

any one of these numbers can be used. I'm assuming the adc multiplies
the incoming signal by this number, but I need to confirm this. The
adc a default natural gain of one. There's a parameter in the
config.ini file where you can set this.
#--------

16:30
Unplugged LWA and low wavz antenna, attached bias tee + 3dB att to
both SNAP inputs.  Data acquired with old DAQ between roughly 16:33
and 16:35.

10:35
Generator on for charging batteries

10:16
Power back on

10:08
Full power off test with Low-Wavz connected

9:41
Low-Wavz antenna plugged into SNAP NS input (with bias tee).  Low-Wavz
is physically located at MIST/MegaMIST location, about 17m east of
LWA, and preamp is connected.  Tied to one of the MegaMIST posts,
pointing up at the sky.

7:30
All power off (including inverter bypass) until radio sked, then power
back on


7 July 2019
-----------

21:58
Power off.  MegaMIST is installed (larger MIST roughly tuned for 70
MHz), cabled into SNAP NS input.

19:06
Power back on.  Nongaussianities are still in the data.

19:00
Power off for radio sked, this time including the bonus AC input
generator switch.  DC switch is still on though.

18:35
System powered back on

18:22
Powering off system to swap batteries

17:30
Replaced LWA NS extendo arms with triangular wire structures, also
about 3m long each.

16:10
Power back on

15:35
Full power off at the base (weather station left running though)

12:56
Reconnected LWA cables at SNAP inputs

12:40
Attached bias tee + 50 ohm terminator at SNAP inputs

12:27
Disconnected LWA cables and attached bias tee + 3dB att at SNAP inputs

11:52
Disconnected MIST at SNAP NS input, reconnected LWA NS

11:00
3m extensions tied on to LWA NS petals.  Horizontal distances from
central post to north and south tent pegs are 287cm and 282cm,
respectively

>>> LWA/MIST baseline length is 17 meters

9:35
Stopping DAQ temporarily, changing RPi system time to UTC because I'm
defeated.  Local time is UTC minus 5 hours.

7:30
Power back on early morning (maybe ~6:30?), followed by power off for
radio sked at 7:30, then power back on

*** NOTE: during last night's power off, AC input generator switch was also off ***
*** Spectra from last night have noticeably more gaussian noise than before ***


6 July 2019
-----------

22:31
Full power off late at night

21:40
Plugged in ethernet cable for ~1min to copy data off RPi

21:35
MIST antenna is piped into SNAP NS input.  MIST antenna is oriented
EW, placed east of LWA antenna by ~15m (to be measured more precisely
if we see some sweet fringe action).

21:30
Disconnecting LWA NS to pipe in MIST instead

19:00
Power off for radio sked

18:35
Restarted dump_spectra with DAQ updates, including new firmware

18:14
Stopped DAQ to try baseband

16:24
Reconnected LWA antenna

16:11
Power back on, 3dB attenuator still attached to SNAP inputs

15:58-15:59
LWA antenna still disconnected, attached bias tee + 3 dB attenuator to SNAP inputs

15:50-15:52
Disconnected LWA antenna from SNAP box, attached bias tee + 50 ohm terminator to SNAP inputs

15:44
Full power shutdown, system is observing the sky

14:45
Reconnected LWA antenna

14:08
LWA antenna still disconnected, attached bias tee + 50 ohm terminator to SNAP inputs

13:05
Disconnected LWA antenna from SNAP box, attached bias tee + 3 dB attenuator to SNAP inputs

Full power shutdown from previous night, turned back on around 7am for
the humans.


5 July 2019
-----------

General notes from earlier:
Weather station broadcasts at 902-928 MHz --> should be aliased to 27-53 MHz
Handheld radios broadcast at 168 MHz and 171 MHz (channels 1 and 2, respectively)

Late night: full power shutdown at camp

20:00?
Generator ran out of gas somewhere in the middle of dinner.  System
still taking data, power on as normal.

19:15
Generator on to charge batteries

19:11
HF off, followed by power back on

19:00
Radio sked, power off

15:40
Turned system off to charge batteries

13:12
Keyed radio for funs:
- Full strength
- Antenna removed
- Antenna reattached with 3dB att
- Antenna reattached with 9dB att

~12:48
Chris buzzed the electric lighter at the kitchen

11:58
Laptops closed, phones in airplane mode

11:38
All batteries out of weather station, box closed again.  Raul was
tootling around the MIST antenna, opening and closing the box
(containing spectrum analyzer and his computer) around 11:53 for a few
times.

11:35
Pulled batteries out of weather station, opened box briefly to check on DAQ

11:15
Main red switch to the inverter turned off

11:05
"Load circuit 12V DC" turned off

10:55
"PV MPPT output one" turned off

10:45
"PV array input one" turned off

10:40
"Load circuit 110V AC" turned off

10:36
"Normal" power inverter turned off.  All things turned off will remain
off for the rest of the tests unless otherwise noted.

10:34
Opened box to transfer all data and check on DAQ, still running fine

10:26
Honda generator turned on briefly for testing purposes

10:02-10:04
Keyed handheld radio three times

7:30
Radio sked, data look clean again in that window


4 July 2019
-----------

20:50
Opened box and copied data off the RPi, left DAQ running

19:07
Power back on, HF switched off after radio sked.  Also, HCC realized
that she had been playing with the electric lighter / RFI source.

18:55
Power is off, HF coming on

18:30
Spooling done, humans away from the antenna

17:55
Starting to work on spooling out ground radials

17:25
Powered system back on with both pols connected

17:15:
Powered off system briefly to connect NS polarization.  Note: Raul is
mucking with MIST antenna stuff at the same time, so there are
computers being powered on/off etc.

~17:00:
LWA antenna powered on, only EW polarization connected at the front
end.  Took data for a few minutes to verify life.  Swamped by low
frequency junk.  Looks like the channel mapping is
pol00 = EW
pol11 = NS