Maximum detection range of Lotek nanotags

user975 · May 31, 2024, 11:32am

Hello everyone.

We have a project that will involve setting up Motus stations for the detection of Nathusius’s pipistrelle bats. Due to the small mass of these bats, in terms of tags, we are limited to Lotek nanotags. In particular, the tags that I’ve been testing are the NTQB2-2-M nanotags (0.32g, 150.1MHz, burst interval 7.1 seconds).

The Motus stations we plan to install will need a detection range of at least 2km (and preferably more) to get full coverage of the area we want to monitor. I’ve been trying to optimise a testing setup to get as large a detection range as possible. I’ll go into a lot of detail here; you can skip ahead to the bold part below for a summary of the questions I have. The testing setup is as follows:

A 5-element (Laird Y1505) or 6-element (Sirio WY 140-6N ) Yagi antenna. The Laird Y1505 is gamma matched as per the instruction manual. I’ve mostly been using one antenna at a time for measuring. The antenna is attached to a telescopic metal pole, which in turn is attached to a fire escape. This gives the antenna a height of ~9m above the ground, and ~2m above the metal railing of the fire escape.
The nanotag is attached to the end of a metal pole, which is attached to a bicycle, putting the tag ~2.5m above the ground. The bicycle is then cycled up to the dike, where I can continue cycling on top of the dike for a few kilometres while keeping line of sight to the antenna.

Antenna output has been analysed using a SensorGnome, a SensorStation, and a Lotek SRX1200. For the SensorGnome/Station, the SDR used was a FunCube dongle pro+. I’ve also looked at raw radio signal from the FunCube using SDRsharp.

During an initial test, the tag could be detected in the raw signal up to a distance of ~1300m, where the signal is barely above the noise level. Of course, this doesn’t mean that any software would also detect the tag, so we tested it with the SRX1200. With the antenna gain set to 71 (as high as it’d go without pelting us with a flurry of false detections) we were able to detect the tag up to a range of ~1100m. We didn’t have the SensorStation at this point, so we unfortunately couldn’t test it yet with these same conditions.

The SensorStation arrived last week, and I’ve been busy trying to at least replicate the 1100m result with it… but to no avail. Over the last few days, my testing has become extremely inconsistent, and I simply cannot figure out why. Using the same setup as before, I’ve gotten a detection range of ~700m with both the SRX1200 and the SensorStation… once. With that same setup, and the pole holding the tag attached to a sign at ~600m, two out of three times the tag was simply undetectable. I checked the raw radio signal as well, and it didn’t show up there either, so the problem must’ve been the tag itself. Other times, I could clearly still see the radio signal (quite a bit stronger than the 1300m case earlier), and the software still wouldn’t detect the tag.

I’ve been trying to tinker with the SensorStation settings in the deployment.txt file, but it seems like the important ones are completely undocumented (regarding a lot of SDR gain/filter settings).

Overall, I’ve got the following questions as a result of the tests I’ve done so far:

*** What is the maximum detection range other people have achieved with Lotek nanotags and any detection software? I’d really like to know what range I can expect under the most optimal of circumstances. Assuming line of sight between tag and receiver is always achieved, does tag/receiver height affect the detection range at all?**

*** What are the allowed values for lna_gain, rf_filter, mixer_gain, if_filter, and if_gain in deployment.txt? Is it ever necessary to change these parameters? **

*** * What value should we use for tuner_gain? The default is 40.2 - does this imply a 40.2 dB amplification? Can we use higher values in environments where we expect less radio noise? How high can the tuner gain realistically go?**

TvE · May 31, 2024, 4:29pm

You may want to consider modifying your attachment so you do not place metal next to the tag where it can influence the antenna. The recommended set-up is to use a non-conductive pole (e.g. wood, PVC, fiberglass) and to attach the tag to something that simulates the animal body. The reason is that the antenna needs a counterpoise to radiate effectively. The two “popular” methods are a piece of hotdog or a small plastic flask/bottle filled with a sponge and saline solution. Both should be sized/filled to weigh about the same as the target animal. Then attach the tag to that with the antenna pointing in the expected direction, for example, horizontally to simulate a bird in flight.

Those are pretty cool screen shots! I have not been able to capture the signals with SDRsharp, they go by too fast for me. The only app that has worked for me is SDRangel, which has an “interesting” UI… But if SDRsharp works for you all the better!

I have been primarily working with RTL-SDRs, see the thread on RTL-SDR vs. FUNcube, but I may be able to help a little. On Sensorgnome V2 the file that controls some of the SDR detail settings is /etc/sensorgnome/acquisition.json, on SSv3 it lives at /data/sg_files/deployment.txt. In that file there are “plans” for rtl-sdr, funcubePro, and funcubeProPlus. Be sure to edit the plan that pertains to the SDR you’re using (e.g., "devType": "funcubeProPlus"). After editing, you need to restart the main process, on SSv3 that would be sudo systemctl restart sensorgnome and on SGv2 that would be sudo systemctl restart sg-control.

In terms of changing the devParams, I have only experience tweaking the ones for RTL-SDR. You mention the tuner_gain: that this setting is for RTL-SDR and not FCD. The FCD plan has gains for various stages and not being intimately familiar with FCDs I can’t comment on the default settings. The filter settings control the frequency band filters that the FCD can switch in/out and these are extremely unlikely to be wrong.

In terms of detection sensitivity: a Lotek tag emits bursts of 4 short (2.5ms) RF pulses with a spacing of up to ~100ms which encodes part of the tag ID, and these bursts are repeated every few seconds (7.1 in your case). Each pulse needs to rise at least 6dB above the surrounding noise (that’s the minsnr setting), where surrounding is ~20kHz in frequency and some ms in time (I don’t know the value). You will see the raw pulses in the Web UI (the SGv2 UI shows SNR, which I find very helpful in this context). In order to show a tag detection in the Web UI two things need to happen: you need to have uploaded the tag definition to the SS/SG and it needs to see at least two pulse bursts with the correct burst interval (it accepts 2x the interval as well).

When you compare distance results with other people remember that most probably use a 9-element Maple yagi, which has more gain (but also a much narrower beam). The height above ground does affect the range due to what’s called the fresnel zone (Fresnel zone - Wikipedia).

This is what comes to my mind at the moment… I hope others will chime in!
Welcome to the forum!

user975 · June 3, 2024, 12:09pm

Hi Thorsten, thanks for the reply!

On the topic of what to attach the tag to:

We’ve heard before that it’s best to attach a tag to a dead animal of whatever species you’re interested in. The biomass would increase the signal strength emitted by the tag+animal combination compared to just the tag by itself. For want of dead bats, we did try a few other “biomass” objects, such as a damp sponge, and we’ve tried wrapping the tag in leaves as well. In a “from the office” test with the antenna on my desk and the tag attached to the wall (2m or so between the two), the leaves in particular seemed to improve the signal a lot. For longer distances, however, we didn’t get any improvements by putting the leaf-wrapped tag on a pole. If the tag needs a ground to function properly, I’d personally have expected the pole-bicycle setup to have acted as one. I have noted that pinching the tag between my fingers does significantly increase the signal strength.

The metal pole is a bit baffling to me as well. You’d expect it to mess with the signal emitted by the tag and to cause all sorts of interferences, but so far, it’s given us the best results. I’ve tried attaching the tag to both a wooden pole and a plastic tube, and in both cases, the signal only got worse.

In all cases, I’ve had the tag attached in a vertical configuration, and the yagi’s polarisation plane has been vertical. I’ve read that this maximises the signal strength. From a physics perspective it doesn’t make the most sense to me for the yagi elements to be perpendicular to the electric field from the tag - how would this field induce standing waves in the elements - so it makes sense to test out the horizontal configuration as well.

On the topic of tuner_gain: that implies I’ve been trying to test out a setting that changes exactly nothing. That definitely explains why I haven’t gotten much out of it. I’ll have to look into the FCD settings, then.

TvE · June 3, 2024, 5:26pm

polarization - Which way is a yagi polarized? - Amateur Radio Stack Exchange looks reasonable to me. The yagis in your two photos are horizontally polarized, that’s not great if you run your tests with vertically oriented tag antennas. In the end you should match the orientation of the antennas on the flying bats, although I’m not sure how much it really matters.

This is what I would have expected. The metal pole is a decent counterpoise. Swap it for wood/PVC and you loose that. The goal here is to get something realistic, not to get the best.

user975 · June 4, 2024, 10:30am

Assuming the metal pole to be a better counterpoise than a bat in flight, this creates quite the problem for the project! I attached the tag to a small container filled with salted water and bits of sponge, with the container being attached to a wooden pole. As a reference measurement, I also used one tag attached to the metal pole, as I’ve used before. At the same distance from the antenna (maybe ~20m horizontally, with a ~7m distance vertically) the metal pole got signal strengths of -19.5dB to -20.5dB with noise levels of -47dB to -49dB, whereas the sponge container got -31dB to -33dB, with noise levels of ~-50.5dB. The detection range was abysmal, going maybe as far as 200m. Frustratingly, I also again wasn’t able to reproduce 600m detections with the metal pole, even though we’ve gotten those detections before (using, as far as I can tell, as close to the same setup as I can manage).

If getting detection ranges of ~600m with a metal pole as counterpoise is already inconsistent, then surely detecting tags attached to actual bats will only be worse. This frustratingly clashes with another report, where ranges over 5km were achieved with a tag glued to a dead songbird (and over 2km with a 5-element yagi). That’s why I’m really looking for more range measurements from other parties, to see if everyone’s able to get >2km detection ranges with these nanotags.

user975 · August 2, 2024, 2:52pm

The saga continues…

We’ve been able to get a detection range of ~700m with a 5-element Laird and a SensorStation & FCD combo. The antenna is at ~4m and is not surrounded by any structures, just a dike and water. The tag was taped to a wooden pole, and that wooden pole was taped to a longer metal pole for additional height.

Since other tests were (still) giving inconsistent and plainly bad results, we’ve bought an antenna analyser. The antenna as it was when we got 700m had a VSWR of 1.7. We’ve adjusted the gamma match that comes with Laird antennas to where the VSWR reads ~1.18. With this adjustment, using the exact same setup as earlier, we got… 200m.

The antenna can’t be the cause of issues at this point, so I’m looking in other directions. Any information about the following would be great:

We’re using Lotek nanotags (NTQB2-2-M, 150.1MHz, BI 7s) that we activate for every measurement, and also deactivate when we’re done. We’ve had these particular tags for ~6 months, and they’ve probably been (de)activated over 30 times. Does Lotek nanotag performance in general drop over time, and does the (de)activation accelerate this drop?
We always use the Live Known Tags window with a local tag database (downloaded from our Motus project) to perform tests. We put a tag at certain distances from the antenna, and write down the signal and noise strength (and also which IF gain was used). One thing we’ve noticed rather consistently is that noise levels also go down when the tag is further away from the antenna, likely due to some of the tag’s signal being registered as noise. Is there any possibility at all that this way of testing can negatively impact detection range? Do other people who test their detection range also do it this way?

TvE · August 2, 2024, 9:55pm

Hi Lynn! Sorry to read about your frustrating experience!
My reaction upon reading your post is that you need to go back to square one and check every piece. You seem to have multiple tags, can you verify that they are all operating and on the correct frequency? You can use an FCD or RTL-SDR with spectrum analyzer software to ensure they’re on 150.1Mhz and if you place them away from near-field (across the room) you should also be able to check that they all have about the same power.

I would then go to the antenna and repeat the measurements by plugging the FCD (connected to cable & antenna) into the laptop, leaving as much of everything the same as it was when connected to the sensorstation. With the tag in proximity you should get a clear detection (yagis have multiple lobes, so behind, below, at 45 degrees in front tend to be pretty good, check out the charts for your antenna). If you have another cable, it may be worth swapping out and comparing. Also, you probably have some small “rubber duck” antenna you can use in the lab for the first tests, you could connect that to the cable instead of the yagi and compare at a “across the room” distance to your lab measurements.

Then hook up to the sensorstation and watch the signal and noise data and see whether it makes sense vs. what you saw with the laptop. I would then place the tag in front of the antenna about 10m away and measure signal. Then double the distance and expect a 6dB drop. If that works out, double the distance again and look for a 6dB drop. Etc.

BTW, you did double-check that your SS tunes the FCD to 150.1Mhz?

My assumption is that if you follow the above steps at some point you will hit a “this doesn’t make sense” point where what you observe is out of line with expectations.

Some thoughts:

if you have a wifi hotspot on your sensorstation then a directional antenna may allow you to stay connected to the SS with your phoen so you can see the readings if you don’t have a 2nd person to work with (alfa has a small panel that is pretty decent and not expensive, if your wifi dongle has a external connector)
if you get a rPi you can use the latest software that has better information and prints SNR clearly

I hope this helps and that you can figure this out!

TvE · August 5, 2024, 3:23am

Since I gave so much “do this and that” advice I thought I’d try some of it myself

So I grabbed an rPi Zero-2W, and FCD, and a small 169Mhz antenna, which is broad enough to work reasonably at 166.38. Then artistically strapped that to a LiPO pack:

Since I was doing this alone I decided to do it the other way around from what I suggested and I placed the tag somewhere and then walked around with the Sensorgnome and my phone talking to the SG’s hotspot.

Here goes the tag, taped to a little bottle with a sponge holding saline solution, about 34g total weight:

I first had the SG laying on the rock about 1m from the tag and saw:

SNR: 25dB, sig/noise:-21/-46dB

I then ~~walked~~ climbed about 10m away, held the SG up above my head and looked at my phone and recorded:

SNR: 8dB, sig/noise:-28/-37dB

I didn’t like that SNR but notice how the issue was more the noise level rising.
Then at 30m, 40m and at 70m:

SNR: 7dB sig/noise:-30/-38dB
SNR: 9dB sig/noise:-28/-38dB
SNR: 6-7dB sig/noise:-33/-41dB

At 70m the tag was not always recognized. This is what it looked like from 70m, the tag is on top of that big rock up there:

What did I learn:

with a little rubber duck antenna the range is pretty small
the noise floor seems to be as much an issue as the signal strength
as a set-up, this works pretty well, just need to shade the phone a bit so I can read it
if I had a yagi I could hand hold that would work too

I then tried to get a couple of long range detections using the 9-element Yagis I have on the tower. The first test was at a distance of 455m nicely in-line with one of the yagis. The result was:

SNR: 9-10dB sig/noise:-44/-55dB

This is the view and the tag:

I then tried a longer distance, about 5km and that didn’t work. It’s about 30 degrees off from the other Yagi at my station. I will reorient the antenna a bit and do some more tests (then I can also test some intermediate distances, which is difficult here in the mountains…).

So my test got further than your 200m but 5km a bit off-axis was a no-go.
To be continued…

user975 · August 5, 2024, 6:52am

Hi Thorsten,

Thanks for the helpful suggestions and your own findings! I’ll check all the individual components to make sure I’m not missing something, then I’ll have another go at getting good detection ranges (testing will have to continue until range improves). If I find something noteworthy, I’ll post it here again.

user975 · August 5, 2024, 2:33pm

I’ve done more testing today, and everything was behaving as you’d expect it to: both tags were on the correct frequency, with similar signal strength. Readings from SDRSharp were consistent with SensorStation detections in terms of SNR, and I got logical values for SNR close to the antenna at different angles with respect to the antenna’s boom. I attached the two tags to a plastic container with some damp sponge, for a total mass of 30.6g.

Going from 100m to 200m, I initially got signal/noise values of -41 to -44 / -60 to -61 @100m and -55 to -56 / -65 to -66 @200m. Not a great SNR at those distances, but a ~6dB drop in SNR for doubling the distance makes sense.

However, when doing some more tests for different tag orientations, I noticed that at one point I seemed to introduce ~8dB of noise by slightly changing the cable connection to the FCD. I was able to get rid of this noise, and reintroduce it, by doing nothing more than fiddling with the coax that goes into the FCD. For the test setup, we’ve just got the SS and other components in a steel box, without much concern for strain relief.

So possibly the coax cable connection has been the culprit with regard to the inconsistencies we’ve experienced. That’d make sense, since every time you build the setup, the connection is going to be different if you don’t pay attention to possible tensions in the cable. I think we’ve tried in the past to make sure the cables weren’t being strained, but we didn’t notice any difference at that point.

With the cable as in the picture (we’ll have to think of something going forward, so it’s a temporary solution for now) I got:

-35dB to -36dB signal with -64dB to -65dB noise @100m
-43dB to -44dB signal with -67dB to -68dB noise @200m

Which is a 6dB drop in SNR for a doubling of distance. Tomorrow we’ll test how far this gets us - if it keeps dropping by 6dB every time, we should be able to get to ~1600m.

With regards to possible coax straining problems: has anyone else experienced anything of the sort, i.e. a ~8dB increase in noise just from the coax being in a certain configuration? I know coax cable can be rather sensitive to strain, but I wasn’t expecting an effect to be quite as strong.

TvE · August 5, 2024, 3:57pm

Glad you’re finding something! Your cable is broken. This is not about moving the coax and it being sensitive. Most likely the connector crimp has failed, possibly the center conductor is broken. RF signals are wild and don’t mind jumping a tiny crack … with attenuation depending on the crack dimensions. If you re-crimp the connector or replace the cable I expect you’ll be fine.

WRT the 6dB: you should loose 6dB when doubling distance, but that’s the theory, measurements vary… Your measurements show a bit more than 6dB going from 100m to 200m, which can be due to a variety of factors. SNR (as the name implies) also depends on noise, which shouldn’t change for a given fixed installation but does change with location and antenna orientation.

For example, at 1km the radius of the fresnel zone is 13m so antenna height and animal flying altitude start to matter. Fresnel Zone Calculator - everything RF

TvE · August 8, 2024, 12:36am

I’ve continued doing some tests. First I reoriented one of the 9-el yagis on my station to point more into the terrain a few km away.

Then I placed my test tag on a boulder on that ridge coming down from the right in the photo. The view back:

The distance is 570m and the FCD is barely picking up the tag with an SNR just above 6dB. Sometimes it only sees 3 out of the 4 pulses:

PLS p3 00:16:22: 1.891kHz snr:6.3dB (-45.0/-52.2dB) 21.8ms
PLS p3 00:16:22: 1.893kHz snr:7.2dB (-44.8/-52.7dB) 19.5ms
PLS p3 00:16:22: 1.897kHz snr:7.0dB (-51.8/-59.6dB) 0.2ms
PLS p3 00:16:22: 1.896kHz snr:6.2dB (-45.5/-52.6dB) 24.5ms
TAG L3 00:16:22: TestTags#1.1@166.38:25.1 1.892kHz snr:7.5dB (-46/-53.5dB)
PLS p3 00:16:47: 1.904kHz snr:7.0dB (-44.6/-52.4dB) 25032.1ms
PLS p3 00:16:47: 1.892kHz snr:6.7dB (-45.0/-52.5dB) 22.0ms
PLS p3 00:16:47: 1.89kHz snr:6.3dB (-45.0/-52.2dB) 19.7ms
PLS p3 00:17:38: 1.898kHz snr:6.1dB (-45.4/-52.4dB) 50154.8ms
PLS p3 00:17:38: 1.887kHz snr:6.8dB (-44.8/-52.4dB) 22.0ms
PLS p3 00:17:38: 1.891kHz snr:6.5dB (-44.9/-52.3dB) 44.0ms
PLS p3 00:18:03: 1.889kHz snr:6.4dB (-45.1/-52.4dB) 25032.3ms
PLS p3 00:18:03: 1.894kHz snr:6.6dB (-45.0/-52.5dB) 22.0ms
PLS p3 00:18:03: 1.888kHz snr:6.1dB (-45.4/-52.4dB) 19.5ms
PLS p3 00:18:03: 1.885kHz snr:6.1dB (-45.3/-52.3dB) 24.5ms
PLS p3 00:18:28: 1.896kHz snr:6.1dB (-45.4/-52.5dB) 25032.3ms
PLS p3 00:18:28: 1.892kHz snr:6.7dB (-51.7/-59.3dB) 22.0ms
PLS p3 00:18:28: 1.882kHz snr:6.4dB (-45.1/-52.4dB) 19.5ms

I notice that the noise floor is pretty high at -52dB. You had around -60dB, others have reported below 70dB. Now, the antenna is pointed at Santa Barbara, so that may be part of the issue.

I then changed the FCD for an RTL-SDR with LNA and there I’m getting better results by a whopping 6dB, so it looks like I could perhaps double the distance!

PLS p3 00:13:01: 1.507kHz snr:13.6dB (-24.7/-38.5dB) 25032.5ms
PLS p3 00:13:02: 1.506kHz snr:13.5dB (-24.6/-38.3dB) 22.0ms
PLS p3 00:13:02: 1.503kHz snr:13.6dB (-24.0/-37.8dB) 19.5ms
PLS p3 00:13:02: 1.512kHz snr:12.8dB (-25.1/-38.1dB) 24.5ms
TAG L3 00:13:01: TestTags#1.1@166.38:25.1 1.507kHz snr:13.6dB (-24.6/-38.2dB)
PLS p3 00:13:27: 1.496kHz snr:12.2dB (-25.0/-37.5dB) 25032.3ms
PLS p3 00:13:27: 1.491kHz snr:12.6dB (-24.6/-37.4dB) 22.0ms
PLS p3 00:13:27: 1.498kHz snr:12.6dB (-24.4/-37.1dB) 19.5ms
PLS p3 00:13:27: 1.489kHz snr:11.6dB (-25.0/-36.8dB) 24.5ms
TAG L3 00:13:27: TestTags#1.1@166.38:25.1 1.493kHz snr:12.5dB (-24.7/-37.2dB)

In any case, if the theory holds then I can double the distance with the RTL-SDR and I should get to 1.1-1.2km. Not exactly far.
Now if the noise floor with the FCD was -70dB that would give me another 18dB, i.e. 8-folding the distance, which would be ~4-5km.
Beyond that I don’t see how reliable reception is possible, however, it’s possible that every now and then a burst gets through.

user975 · August 8, 2024, 7:01am

Hi Thorsten,

Tuesday I did some more testing, still with the same coax cable (which you suspect might be broken). After taping the coax cable to the back of the steel box, the signal got quite a bit better. There’s also a 149.6 - 150.6 MHz bandpass filter between the antenna and the FCD now.

This way, we got:

-34 to -36 dB signal with -67 to -68 dB noise at 100m
-45 to -46 dB signal with -73 to -75 dB noise at 200m
-67.5 to -68.5 dB signal with -75.5 to -76 dB noise at 800m
Occasional singular live pulses at 900m, but no tag detections at that distance.

Detaching the coax from the back of the wall can increase noise by ~8dB or so, but firmly taping it to the wall seems to consistently optimise signal strength for the time being. We’ll have to take down the mast and replace the coax to see if it’s truly broken. There are likely more optimisations that can be done in terms of cable/electronics management, so we’ll be trying some other things soon as well. Suggestions are welcome!

TvE · August 8, 2024, 7:43am

Problem with the filter is that it adds noise and thereby reduces range. Anything in the signal path adds noise. If you don’t have an intermod problem then the filter only hurts. The device I use is an LNA plus filter, the LNA has a lower noise figure than the rtl-sdr front-end and the gain compensates for the losses and noise added by the filter. It’s not without downsides either but it can actually reduce noise.

Your coax looks pretty thin… Something else you could perhaps try is to get a 3m USB cable and attach the FCD directly to the antenna? I don’t know what the distance is, but might be worth a try.

Overall it seems our results are compatible. You’re getting a lower noise floor and lower signal, but I’m using a higher gain antenna and I’m pointed right at town. I’m planning to point the antenna in a different direction tomorrow and see what that does.

William_D · August 10, 2024, 5:34pm

I have a question about tag deployment. when attached to an animal in flight is the tags antenna parallel to the animal’s flight path or perpendicular. I am asking because wouldn’t the orientation of the xmit and rcv antennas have an impact on the signal reception.

user975 · August 12, 2024, 6:57am

Hi William,

I’m not personally involved in tagging animals, but from what I’ve seen, generally you’d attach the tag such that it is parallel to the animal’s flight path. For our study we’re interested in bats, and there it’d look something like this:

bat_tag

(Image taken from https://archived.sensorgnome.org/Field_Season_Notes/)

For my most recent tests, I’ve attached the tags to the sponge container horizontally, and I’ve oriented it such that the tags are perpendicular to the antenna’s boom (and thus parallel to its elements). For me, this configuration has resulted in the best signal strengths. This would also imply that a fly-by is easier to detect than an animal approaching the antenna head-on.

Below an image of how I’ve had the tags attached to the sponge container for the last few tests. I’ve had one pointing left and the other right so that symmetry won’t affect signal strengths.

user975 · September 2, 2024, 1:25pm

Update on the measurements:

I’ve taken everything out of the steel box, except the 230V AC → 12V DC adapter, as it is likely a noise source that affects the shielding in the coax cables. I’ve also removed the bandpass filter, as it did not seem to consistently reduce noise levels. For these tests, the SensorStation was inside a small Pelican case. A 6-element Sirio antenna was used, with a VSWR of ~1.12. There are ferrite clamps on various locations along the coax cable, as well as at the point where the coax cable connects to the antenna.

I was able to get noise levels of ~-80.5dB by closing the door to the steel box. Noise would be ~4dB higher with the door opened. With the door closed, the noise levels were quite consistent. A few averages of signal/noise at distances:

100m: signal -35dB, noise -65dB
400m: signal -56.5dB, noise -80dB
800m: signal -70dB, noise -80.5dB
1000m: signal -73.5dB, noise -80.5dB
1100m: no tag detections. Live pulses were detected at -74dB with -80.5dB noise.

1000-1200m was the range I expected for a setup with our antenna at ~4m and the tag on top of the dike. It likely would have ended up as at least 1100m, but the signal dropped a bit faster at the 200m to 400m range than in some previous measurements. We ended up with 2-3dB lower signals, but also ~5dB lower noise, yielding a net improvement in detection range.

From all the testing I’ve done so far, the culprit in my case definitely seems to have been the power supply. Moving it in certain ways can easily swing the noise by up to 10dB in some cases. Completely isolating it inside the steel box has given me the best results so far. Luckily, for our actual motus stations we won’t have to deal with an adapter so close to the coax cables, but this noise source can definitely mess with testing setups and waste a lot of time if you’re not aware of it!