Can we build these stations cheaper? faster? better?

Stations
1

Hi fellow Motus installers. We know these stations work but can we improve? Here are my questions: 1. Should we forget about installing lightning protection but instead simply ground the mast?
2. Is 1.75" OD EMT the best mast material for the money? It is $28.00 for 10’ and easier to work with than water pipe.

  1. Is LMR400 cable overkill for cable VHF and UHF?

  2. How much detection range will we loose if we install 3 element VHF antennas instead of the 9 element?

  3. Is there a proven SDR other than the Funcube for VHF that is cheaper and holds up as well?

  4. There seems to be a trend to only install UHF antennas to avoid SDRs and big antennas. Others say that not installing VHF is not in the spirit of Motus. However, if we all got on board and settled on 1 frequency, we could probably install 1/3 more stations for the money. Am I alone here? Are the UHF tags small enough for smaller animals?

Robert Rosenberg
Kansas Motus

2

Thanks, Robert, for the questions. I’ll put my 2 cents here.

  1. Lightning protection is an industry standard that we would be good to abide by in most circumstances, especially if we are co-locating on other’s properties… and often it’s a requirement for colocation. I’d recommend against deleting this step.
  2. EMT has thinner walls than the schedule 40 water pipe, but can suffice in some cases where you have fewer antennas on the mast. EMT will also weaken over time with vibration and constant (wind) pressure fromm one direction.
  3. LMR400 might often be overkill if our cable lengths are relatively short. Refer to any signal loss by cable type table for details here. (with LMR400, at 100 foot, we lose ~30% of signal for 166MHz, and ~45% of signal for 434MHz)
  4. Bob Morton developed a table / worksheet calculating assumed detection range for several sizes of antennas, I’d be happy to share this when I can find it.
  5. As of now, the FunCube ProPlus is the only available SDR that has a 16bit analog to digital converter, which makes it the most sensitive SDR for low strength signals like what we deal with in Motus. Other SDRs work fine, but are far less sensitive, and would potentially result in fewer tag detections.
  6. I’d say stick with both frequencies as a courtesy to others and to maximize the usefulness of the whole network. But, there are others who share the same thoughts as you here.

Matt

3

For what it’s worth as someone who tracks many bats in North America (mostly Ontario) and some in Central America, my only option is digitally coded tags that are 0.2-0.3 g in size. Well…I could go bigger for a few species of bats. As far as I know, Lotek is the only provider of these on the 166.380 VHF frequency. I don’t have another option.

When I setup towers, I usually use 3x 4-element yagi and on most I add the 2x 434 kHz antennae for those on other projects using CTT tags, per guidance from Birds Canada.

In theory, an SRX1200 (Lotek) or Ares (SigmaEight) receiver could work as a receiver station just like they were being used handheld because they can read and log coded tags. However, they would use much more power and be a lot more expensive than the funcube, SG options.

Not sure if that info is of value for this discussion.

Derek Morningstar, M.Sc., PWS

Myotistar

myotistar@gmail.com

www.myotistar.ca

(226) GOT-BATS

4

2 cents from the peanut gallery:

I use “KMR-400” cable, which is well-priced and well made (Amazon.com). It’s not identical to LMR-400 but mostly interchangeable. For 25ft+ runs this is quite a bit better than cheaper thinner stuff. But in the end, I believe the biggest cost ends up being tools, connectors, and training/mistakes. Each cable type has its own connectors and procedure for attaching them. Unless you do all of them all the time you are very likely to mess one up if you haven’t done one in a while, at least that’s what happens to me. So if you have any long runs it’s better to use LMR-400/KMR-400 consistently, have one bag of extra connectors, one crimping tool, one sheet detailing the process, one strip gauge, etc.

WRT frequencies, I believe it is quite important to the health of the Motus network to have multiple vendor options and multiple technology options.

Thorsten

5

I will second Thorsten and Matt have said about the cable. I do a lot of station installs and still frequently have an issue due to switching connectors and not correctly modifying how I prep the ends of the coax cable. A recent example cost us a. Extra 6 hours as we had to bring down a pop-tower to redo the top connectors.

I am finding it harder to support the point about continuing to put up dual frequency stations as the number of birds tagged with 166.378 MHz dwindles. More than half of equipment and labor, and most of the risk of putting up a station come from installing antennas for 166. For now we need to do it to support the broader Motus network.

Patrick Lorch
plorchgm@gmail.com

6

My thoughts on this:

  1. Should we forget about installing lightning protection but instead simply ground the mast?
    No- as Matt stated, it is an important safety component- at a minimum protective of equipment and may also save lives.

  2. Is 1.75" OD EMT the best mast material for the money? It is $28.00 for 10’ and easier to work with than water pipe.

I find the EMT to be too flexible unless well guyed and yes the flexing will weaken the metal over time.

  1. There seems to be a trend to only install UHF antennas to avoid SDRs and big antennas. Others say that not installing VHF is not in the spirit of Motus. However, if we all got on board and settled on 1 frequency, we could probably install 1/3 more stations for the money. Am I alone here? Are the UHF tags small enough for smaller animals?

I am of two minds about this and am working to understand the real world trade-offs as well. While I also like the idea of maintaining compatibility with previous analog 166 technology and not be tied to one frequency, it does present a number of problems that I could do without. To mention a few- power consumption is increased which means you need bigger batteries than you otherwise would, the antennas are large and therefore increase safety concerns for installers and limit installation locations and configurations, the frequency is noisy to begin with and no noise filtering happens at the sensorstation, so all that noise adds to data charges, dongles are relatively pricey and hard to come by, and as you point out, it adds substantial costs.

I am in California. I have 17 stations that have come online over the past 3 years. I need to comb through the data but I think I have had only a handful of detections on 166 compared to hundreds on 434. The digital technology of 434 means that most researchers will continue to trend this direction.

I think a good path forward would be for installers to add a single 166 omni if they can to maintain this frequency a bit longer. But I would like to hear from folks using 166 tags.

7

This is good feedback. We see fewer Lotek tags than we used to on the stations I monitor. I forget that they are the only option for some smaller target species. On larger projects the cost of installing dual mode stations adds up quickly. I guess we could do as you have done and just put up an omni or a smaller number of 166 antennas. In the end, this all bears further discussion. I want to support other projects, we just want to be careful to periodically evaluate the need explicitly.

8

My thoughts:

  1. Should we forget about installing lightning protection but instead simply ground the mast?

No- lightning protection is an important safety component and we should keep using it. At a minimum it might save a sensorstation but it might even save a life. I agree with what others have said on this. I also get better SWR readings when there is good grounding through the lightning arrester.

  1. Is 1.75" OD EMT the best mast material for the money? It is $28.00 for 10’ and easier to work with than water pipe.

I think EMT is too flexible unless it is thoroughly guyed. That flexing weakens the metal and leads to failure.

  1. Is LMR400 cable overkill for cable VHF and UHF?

No strong opinions- I have only ever used LMR400.

  1. How much detection range will we loose if we install 3 element VHF antennas instead of the 9 element?

I am not sure. We got nearly a mile out of 3 element 166 antennas.

  1. Is there a proven SDR other than the Funcube for VHF that is cheaper and holds up as well?

Matt covered it well.

  1. There seems to be a trend to only install UHF antennas to avoid SDRs and big antennas. Others say that not installing VHF is not in the spirit of Motus. However, if we all got on board and settled on 1 frequency, we could probably install 1/3 more stations for the money. Am I alone here? Are the UHF tags small enough for smaller animals?

This is a really important question and discussion that needs to be had. I am thinking through this as well. In principle, I like the idea of continuing to support 166 so we have a diversity of frequencies to work with.

But there are notable problems that continuing this compatibility poses. To name a few: Power consumption is higher, perhaps nearly double, so that means larger more expensive batteries and solar panels. The large awkward antennas add lots of time and risk to installations and constrains the options for locations, configurations and materials (relevant to your question #2). Dongles are expensive, unreliable and hard to come by. The frequency is inherently more noisy and with no sensorstation filtering, all that noise has to be relayed on. This impacts data charges.

So should we continue to shoulder these problems? I am in California. Granted I struggled to obtain dongles for a long time, so a more careful analysis is warranted, but I have only detected a few individuals tagged with 166 compared to hundreds with 434. The digital 434 tags are so much more reliable and with less noise problems that I think researchers will continue to trend away from 166.

Perhaps a good path forward is to use a single 166 omni when possible to maintain some level of compatibility. This would mitigate many of the problems at least in part. But, to me, the spirit of Motus is shared contributions resulting in shared benefit. Dropping 166 would not be contrary to this philosophy- especially if it means we can get more coverage and a stronger, more reliable network without it.

For the 166 tag users out there: what are the risks of trending away from this frequency that we might not be thinking about?

9

For reference, there were still more Lotek tag than CTT registered to Motus in 2023:

CTT 4721

Lotek 6093

While there is an uptick in usage for CTT, and while FSK technology is probably preferable for encoding in the absence of other constraints (e.g. animal size), Lotek tags remain the most used for the time being. I am sure the ratio varies geographically. For 2024 so far, the registration numbers are about equal (about 250 each), but we’re still early.

Cheers

Denis

image003.png

10

Or if we really want to be disruptive, we could just ask Apple to make AirTags a bit smaller. Then we can ‘find my bird/bat’ without ever having to install a tower again (except where there is no cell coverage). Just sayin…

Derek Morningstar, M.Sc., PWS

Myotistar

myotistar@gmail.com

www.myotistar.ca

(226) GOT-BATS

11

As others have pointed out, 166 tags are currently the only option for small birds and other small animals. Unless/until there are smaller UHF tags, installing stations without VHF will limit the usefulness of Motus for an entire set of users.

12

It’s also worth mentioning and keeping in mind that there has been a lot of work and effort put into encouraging bat researchers to use Motus. As of now, the Lotek tags are the best option for bats. As we continue to build out the network, removing the 166.380MHz frequency will limit bat research. I would discourage a single omni for 166, the best we could expect with an omni might be 1 to 2km of detection range, greatly limiting the network capabilities. I think we’d be better served with effort put into finding (or developing) an additional dongle that works as well as the Funcubes, and limiting the number of antennas we put on the mast to begin with… good questions might be: is 3 antennas as good as 4? can we get away with 2 antennas per frequency? etc. reducing the number of antennas but still supporting both frequencies will save cost and reduce weight.

13

Hey everyone,
Sorry to double post- When my first response didn’t show up, I rewrote it.

Anyway, regarding the apple airtag- as far as I can tell the Blumorpho from CTT is basically the same principle.

Levi

14

Thanks Denis. Having real numbers helps justify keeping the 166 antennas going up. As I said, for now it seems justified. It also seems justified to look at the relative numbers of tags at each frequency within flyways.

Patrick Lorch
plorchgm@gmail.com

15

I like Matt’s idea of looking more critically at antenna number. It should be possible to use real data from Motus to look at this question. An analysis that asks what do we lose if we drop detections from 1 or 2 antennas with the least detections in an array might get at this question. A similar analysis of antenna height might be worth a look. This kind of data driven redesign effort might pay off.

Patrick Lorch

plorchgm@gmail.com

16

All:

Currently Lotek makes the lightest tags that can be detected on the majority of Motus stations- so for anyone wanting to tag species that require a tag lighter than 0.35g, that's the choice. In terms of frequencies, there is definitely validity in multi-mode stations- that was the rationale behind us creating the SensorStation with so many potential inputs: eventually a new frequency would come along and provide an advantage over the others, for specific applications; or a new feature would be requested and we wouldn't want Motus collaborators feeling like they had to wholesale change over their station. The shorthand for this was "future proofing".

For instance, we have a client in Peru who has a LoRa network for communicating when traps are triggered, so we were able to adapt one of their LoRa radios to one of the 5 pads typically reserved for the 434MHz radios. In other cases researchers required six (6) or more 434MHz antennas for doing localization ala the ARTS (Automated Radio Telemetry System) setup, so with the plethora of USB ports these could be added via dongles.

Now we've launched a new line of tags on a new frequency, 2.4GHz, which, with a simple plugin adapter, can be added to existing SensorStations- and (I can hear you holding your breath from here) will require additional antennas. We installed the first in Mexico last week, using four 2.4GHz panel antennas which each have a footprint of 10cm x 10cm x 8.5cm- think of really small cubes. These four antennas can pick up tags up to two or more KM (possibly farther, we've only tested out to 2KM so far), but each provides a 90 degree pickup pattern, so with 4 we get full 360 degree coverage around the station. These tags weigh only 0.06g, but are solar-only, so no good for bats yet (the battery-only version is coming out in the next couple of months and weighs ~0.13g and beeps for 46 days at a 3 sec interval), or nocturnally migrating birds (although the hybrid version si coming out mid-2024 and will weigh ~0.15-0.2g) where nocturnal detection is important- but it makes sense to think about adding this frequency to all Motus towers eventually, as this will allow for a massive expansion in species which can be tracked- from insects (the tags are already being deployed on Monarch butterflies and Asian hornets), small birds (we deployed 11 on 5 species of hummingbird last week in Mexico, the smallest being 3.1g and in all cases the full tag and harness was < 3% of the bird's mass), and eventually even the smallest bats.

So while I absolutely understand the real costs of multiple frequencies (I manage my own Motus stations in Cape May, NJ), it makes sense to think about how we can best support all frequencies with the smallest footprint. For instance, when 13 Gulf Coast stations were upgraded to SensorStations a few years ago, there wasn't enough funding to install a full compliment of 434MHz yagi antennas- so only a single omni was added at the time. Those omnis detected a bunch of birds moving between across the gulf in both spring and fall! Since then they've been incrementally upgraded with additional yagis as funding allows.

Just my 3 cents.

Cheers,

David

David A. La Puma, PhD (He/Him) • Vice President, Global Market Development
celltracktech.com • (609) 889-0305 x104 • Book​​​ Time With Me

“If we did all of the things we are capable of, we would astound ourselves.” - Thomas Edison

17

Both Omni directional and yagi antenna are fairly straightforward to build
and can be done so at a fraction of the cost of buying online. There is no
easy solution to safely mounting two 9 element antenna (especially the
larger ones for 166.38) up high. An array of ground mounted omnidirectional
is what I am opting for, possibly with one or two strategically placed yagi
on a natural high elevation unobstructed point.

Sensorgnome receiver is also fairly cheap and easy to build

In my opinion, it makes far more sense to build stations with the rtl-sdr
or similarly inexpensive dongles instead of the funcube, and then do
everything else you can to preserve signal strength- get a nanovna to make
sure your antenna has an swr close to 1 and use short cable runs with
lmr-400 using minimal connections.

Get a proper lmr400 prep tool and some nice ratcheting crimpers for putting
on connections and you will find that thicker coax cable is not so bad to
work with when you have the right tools. It's actually kind of fun. And
knowing you can easily make your ideal combination of connections like N to
sma in a matter of minutes makes working and prototyping quite nice.

As for new tech, LoRa seems extremely promising but still a bit heavy.
There are an abundance of LoRa towers across the world, but some sort of
access would have to be arranged and purchased for a small fee. Anything to
prevent researchers and underfunded orgs from having to worry about station
infrastructure and maintenance is probably a good thing. Similarly, if
cell/gsm based tags keep getting lighter, I am sure that everyone would
prefer to get extremely accurate location data. I have not seen one under
1k

IMO the biggest hindrance to MOTUS studies is the cost of tags, and as
mentioned several times before me, the weight of more advanced tags that
operate at higher frequency than lotek prevents work with smaller species.
A cheap open source tag design would be nice

18

Hi all,

I’ve been enjoying the discussion of potential improvements to Motus stations. This might be a slightly iconoclastic opinion, but I think variation in station design, from tower structure, to antenna number, to height, to lightning-safety are always going to be variable place to place and dependent on budget, logistics, permitting and research question. I think we have to embrace that fact as a community as opposed to trying to homogenize a beautifully chaotic and collaborative network. The one exception to this, in my opinion, is dual-mode stations. None of us can predict the ebbs and flows in popularity/utility of 434 vs 166 and I think it would be a terrible waste to not make an effort to build every single new station as dual-mode. Overall – By all means, lets iterate and improve where possible, but I really don’t think that’s where we should be putting our energy at this moment, see below.

If you’ve got time and energy to devote to Motus, I think it’s network maintenance and resource sharing that are the huge, screaming priorities. Large areas of our shared network are super tenuous, both physically and financially, and big big numbers of those lovely dots on the Motus map are non-functional and/or nearly unsupervised. If we are going to innovate in the Motus space, let’s do it there!

Ciao

Andy

1 Like
19

Andy, I like your iconoclastic opinion and your suggestion to focus energy on improving the reliability of the existing network :-). Since I work on the technical end, the Sensorgnome V2 software specifically, I have a couple of suggestions on that front. The V2 software is designed to leverage an internet-connection (it also works without) and there is a long list of big reliability improvements that come with that. In particular, these stations are monitored in near real-time which allows problems to be spotted before the season is over and the SDcards retrieved. In many cases troubleshooting can happen remotely and fixes or workarounds can be deployed promptly. The monitoring and remote fixing capabilities are not yet where I’d like them to be but every time I get a problem report I iterate and improve things. If more Motus stations were to deploy the new software with connectivity this iteration would happen faster. At this point in the US it costs less than ~$150 to get cellular modem and 1 year of connectivity and given the benefits it pays for itself in short order.

Baxter, you mention LoRa, which I like a lot. It could increase the detection distance and reduce antenna requirements significantly. The thing that doesn’t convince me is the power budget, because long range LoRa transmissions are very slow and thereby consume a lot of power. I haven’t done a careful analysis but in order for it to pen out one would have to consider the entire system, including increasing the TX interval in light of having a greater detection distance.

I’ve also wondered about an open source tag design and I know I can make the technical aspects happen. Recall that the CTT tags were originally designed and made at Cornell. The issue about “cheap” is far more tricky, however. I casually listen in on the CTT slack and there is a non-stop stream of issues around tags. There are multiple types of antenna wires with varying lengths and insulation materials. Which to pick depends on the species and habitats (antennas getting pulled out by the birds or getting entangled in vegetation, etc). There are also multiple types of harnesses and fasteners/glue. Again, which to use depends on the specifics of each project. The vendors provide all these different tag accessories and they provide support. This costs money. So yeah, I could get open source tags manufactured at 1/10th the cost of a tag you currently buy but all this other stuff doesn’t magically come for free…

Thorsten (Sensorgnome V2 software dev)

20

About IoT- tags: There are small IoT tags running on the SigFox network; they works great, but <1.28 gram without GPS becomes very challenging. See Wild et al 2023: https://animalbiotelemetry.biomedcentral.com/articles/10.1186/s40317-023-00326-1

SigFox is a commercially available network in Europe that shows great maps with coverage, but is not that great in some regions…

René