I see that big installations of infrasound sensors tend to use various sorts of arrays of pipes (or hose) to reduce pickup of wind noise and other unwanted sounds. Is there any benefit to doing this with the RBOOM? and how would you do it? (has anyone done it?)
Revision (found the instructions!) https://manual.raspberryshake.org/boom.html#extension-tubing
But still interested in the discussion …
Ken
Hi Ken,
There are several ways to do this, from simply extending the tubing away from noise sources, to quite elaborate solutions like @jochen.richert’s infrasound array seen here: http://wetter.richert.ch/infrasound/infra-details.html
The benefit is that you get less anthropogenic or wind noise and cleaner atmospheric pressure fluctuations. It’s similar to installing a Shake away from a house in order to isolate it from people walking around, for example. Perhaps someone from the community can comment more fully than me.
Ian
hmmm - not much discussion. It seems to me that you need to limit any pipe array (“rosette”) to 3.4 m radius in order to avoid a resonance near 50 Hz. I have seen some larger ones in online papers, but they were looking at lower frequencies. I would guess that PVC pipe would itself generate noise when stimulated by wind or rain so that something heavier would be better (as shown in your link). Either that or protect the array from wind and rain (not easy - unless perhaps buried, but then it would pick up seismic activity).
Ken:
Buenas tardes.
You probably saw this already but we provide some recommendations and links to additional resources here.
branden
Yes, thanks, this is the same as the link in my revised text of May 30. One question I have: I intend to mount the RB in a diecast metal box for RF shielding. After some thought, it occurs to me that I really should put the filter tube inside the box to avoid extending the effective length of the filter tube and changing the time constant. So the box will need a vent hole - which is undesirable for outdoor use…
On the other hand, if I double the volume of the tube between the RB and the capillary orifice by placing the filter outside the box, all it would do is double the time constant to ~2 seconds (in the case of the 1 second filter). I would be happy with 2 seconds.
What is the consequence of telling setup that I have a 1 second filter when, in fact, it is a 2 second filter?
Ken:
Buenas tardes.
Please send me a photo of all filter parts when they arrive and I will annotate it to explain things a bit more.
Yours, Branden
It occurred to me that since there is some tubing after the orifice I can just connect that to a port going outside the box. Photo below shows the box and the bulkhead feedthroughs that I will use. I just need to make sure I can curl up the tubes inside the box without kinking them. The non-filter port will go to the pipe array and the filter will just have a bug screen on it. This metal box will go inside another (plastic) box for rain-proofing.
Both ports should be joined/ go to the same pipe array.
branden
aha!
Thanks for that.
So the little black gizmo in the instruction photo is a sort of T fitting or manifold… I guess the connection is required for the high pass-filter to properly cancel signals below cutoff.
Yes
Both tubes (the red and blank; or yellow and blank) should see the same atmosphere. The filter is entirely contained within the colored tube. Once the colored and blank tubes are plugged ino the same T fitting, the tube that runs from the end of the “T” to the outside can be any length.
branden
Looking great!
To get the correct polarity, switch the tubes so that the banded tube is on top.
And I would keep that Ethernet cable away from the blue board.
branden
That’s a good looking enclosure.
It is a standard Hammond 1590-series diecast box with gasketed lid. The rectangular object to the right of the RPI is an 5/10 kV ethernet isolator. Ferrite cores are placed on the ethernet cable both inside and outside the box. The DC feed (at 24 volts) has RF filtering, as well.
I used this approach with my Shake and it withstands the RF environment up here quite well. (The “home run” cable from the unit is fiber optic using Ubiquiti F-POE).
I thought you would like to see the effect of putting cover on the box while the Boom was running on my workbench.
Ah yes, from the manual:
“WARNING: The Raspberry Shake board, if left uncovered, will produce some long-period wander. To ensure the highest quality signal, always operate the Raspberry Shake inside of the enclosure provided by Raspberry Shake or your own.”
In practice, covering the board reduced noise across a broad spectrum.
branden
My experience seems to indicate that electrical shielding would help - as much as people like to look at the raspberry pi in its plexiglass case, it really needs to be fully enclosed in metal.
It seems that the RBOOM makes a pretty reasonable seismometer - responding to acceleration rather than velocity - so if you have it on a table top or other resilient surface, it’s going to pick up a lot of signal that is seismic in nature rather than infrasonic. By seismic I mean man-made vibrations or “cultural noise”. My background noise level went down when I took it off the work bench and put it on the concrete floor.
I wonder if it is more sensitive in one axis than the other? My tests only involved moving it up and down (z-axis). If it is less sensitive in the X or Y directions, it might help to mount the whole thing sideways.
ok - Since no one has taken me up on my offer I decided to shake my RBOOM up/down, front/back and side-to-side. In one direction, it seems, the unit is relatively immune to acceleration effects.
Assuming that most ground vibrations will be in the Z direction (up and down) then the preferred orientation of the RBOOM is with the X-axis vertical. By X-axis I mean the axis in line with the nozzles or barbs on the differential pressure transducer. This is not the way the standard RBOOM is configured, but not hard to do.
