[time-nuts] St Veran gravity red-shift misson - GNSS thread

[Magnus Danielson]

Dear all,

So far I have described mostly the power side of things. I thought I 
would cover more of the GNSS side of things, as I think there might be 
interest in that.

Besides continuous operation, measuring the clock is an important part 
of the mission. This has it's own set of challenges.

The normal way of operating for ORB is to use common view, and in 
particular the PPP tool of Pascale Defraigne, which is one of the most 
distringuished experts in the field. She has been a great supporter for 
this.

So, already from my home outside Stockholm Sweden, I had prepared to 
operate my Septentrio Mosaic-T. However, I did not have time to mount an 
antenna on the roof of my car, so that only happend at my summerhouse 
600 km and 4 degrees south. On the car I mounted a Leica AT504 choke 
ring antenna with a radome. I was a bit worried about the mounting 
holding up for the trip, by now 4300 km later and at speeds up to 167 
km/h, it just sits there untouched. Since I was not used to the 
Mosaic-T, I was unable to figure out how to enable logging, so Bruno had 
to do that for me later.

A particular problem I had is that the RxTools software run on Windows 
and Linux, and naturally I only got my Mac with me and I did not get it 
to run under Wine. I managed to do a few things using the web-page and 
commands in command window. Ah well.

I knew that Bruno would bring a Septentrio PolaRx4TR, so I had installed 
a GPS Source antenna-splitter.

I brought two Leica antennas, one for which the radome was not mounted. 
Futher I had brought a Sokkan SK702 pin-wheel antenna. A varity of 
cables and adaptors was also brought. Turns out that this was needed.

As we transfered car in Saint Veran village, I realized that it would be 
too much work to transfer the antenna on my roof, but since I had the 
same antenna as spare we used that. Similarly, the pin-wheel was better 
adapted to record the climb from Saint-Veran village to the Observatoir 
de Saint Veran. That I had a suitable cable available resolved that 
problem. Bruno held the pin-wheel antenna out the window, doing his best 
Hamlet "To be or not to be" interpretation, but without the text. As we 
did the climb, we where honoured to see plenty of marmots. They are very 
unafraid, so they move away a few meters.

As we climbed up to the observatory, we transfered the clock. Then we 
took the other choke ring and placed on the roof. The room suggested for 
us was ideal, as there where pre-existing holes in the wall for cables. 
Bruno had brought a new temperature stable version of the LMR-400, and 
we had to extend it just a little to reach from the ideal antenna 
location over to the clock assembly. Using vulcanizing tape and black 
isolation tape, I secured the connectors. We only had a little rain. 
There can be quite a bit of wind, but we saw no trace of the antenna 
movement, and I had made only a poor attempt at securing the antenna, 
but it turns out the efforts done was sufficient this time. For next 
time I will improve on that.

On our way down, we just did the same thing in reversal, with the 
exception that we moved the clock over to the observatory van the 
evening before, just to avoid stress in the morning. We kept the choke 
ring and just dropped the antenna cable down so it would be reconnected 
in the van, along with "house power". That location exposed it to more 
temperature variations, which is a risk, but that risk is much lower 
than the risk of stressing the other things.

So, both the recievers was getting GNSS signals. The primary goal was to 
have them running redundantly in parallel and log onto their internal 
disks. This keeps their dependence to external things to a minimum.

Both receivers god a 10 MHz reference from the cesium. An auxillary PPS 
also provided.

This should be enough to measure the frequency of the cesium throughout 
the experience.

We got early results, showing the 1 ms steps I talked about. Resetting 
the PPS on the cesium resolved that issue, and does not change the 
frequency of the 10 MHz, as also explicitly stated in the manual.

A curious thing is that my Mosaic-T lost the ability to provide PNT 
solution, yet being able to log track all signals as normal. I have yet 
to resolve that, but we concluded it could log things and we can 
post-process the data.

In addition I pulled the PPS out of the PolaRx4TR to a TADR TICC, also 
fed a 10 MHz from the cesium to the TICC. I had intended to time-stamp 
both the PPS from receiver and clock, but my TICC did not time-stamp 
many of the PPS pulses from the cesium ever, and none while on the 
mountain. Because of power failure issues, I pulled the TICCs out in 
Grenoble and re-installed later in Saint Veran, but the output was not 
useful, essentially looking at delay noise and such a small slope that 
no real frequency can be given. I have therefore chosen to ignore the 
data. I will see if I god more data as we decended.

Because we failed to get a good clock charcterization before going up, 
we do that now afterwards. The cesium now sits in the wellcontroled 
environment on ORB and receives signal from their antenna in parallel 
with their clock system. We want a long reference trace to draw 
conclusions. Also, key personal is on holliday. This is why you do not 
get any nice plots right now. I had intended to have the TICC trace fill 
that purpose as a preliminary analyis. It would have even cooler to do 
realtime display, but I simply did not have the time.

I should add that the TADR TICC has phenomenal performance per W power. 
This is very relevant for setups like these.

Data is recorded. Data will be analyzed. More lessons to be learned.

We learned immensly about how difficult it is to do this. I could forsee 
many obsticles, but not all.

Once home, I aim to setup my time-lab properly, and to be honest, part 
of this setup is more or less what I need to setup in the lab.

Cheers,
Magnus