I’m looking for help from DIYers and 3D printing enthusiasts in the Yarbo community to design a custom mount for a GPS amplifier setup that will replace the High RTK antenna mount on the Yarbo core module.
What I’m Trying to Build:
A replacement mount that can securely house the ArduSimple AS-ACC-RFINLINEAMP-L1256-2SMA-00 GPS inline amplifier, with the following design goals:
Internal Diameter: Wide enough to fully enclose the amplifier (~0.75 inches / 19mm diameter, ~2.6 inches / 66mm length)
Mounting Location: Replaces the stock High RTK antenna mount on Yarbo
Cable Management: Allows cables to pass through cleanly and weather-seal if possible
Antenna Integration: The mount should allow the RTK antenna to:
Screw directly onto the amplifier, or
Be connected via a short SMA jumper cable
Why This Is Needed:
I’ve been experiencing unreliable GPS due to signal loss and/or multipath issues. This amplifier should help, but I want to integrate it cleanly into the system without dangling cables or exposed hardware.
Call for Help:
If anyone has:
Experience with CAD design
Access to a 3D printer
Ideas for weatherproofing or modular designs
…I’d love to collaborate! Even early concept sketches or prototype designs would be incredibly helpful.
Feel free to reply here or DM me if you’re interested.
I can help you design and print something. I have a couple ideas. One would be to redesign the entire mount and the other would be to create a slip mount that slides over the existing RTK mount. I’ll send you a DM.
That seems small enough that it could easily tuck inside using just a short jumper to connect to the antenna, have you looked under the cover yet? Somewhere in the forum someone made a photo study of replacing the cable, may be able to tell from that.
Then secondly, since you have it installed already, do you have then data that shows a difference in signal or performance using that amplifier? It’s not cheap, but at around $160 I’d probably invest in it if I knew it would eliminate the fairly regular waiting for GPS signal issue I see here.
this likely won’t help what you are wanting it to help. This is for extending the cable length. This isn’t for increasing the GPS sensitivity. For that, you need a new antenna. However, multipath issues definitely can’t be fixed by a more sensitive antenna.
From the site selling it:
“An inline amplifier is required when the antenna cable length attenuates the GNSS signal to a level below the receiver’s recommended minimum input level. With ArduSimple’s low loss antenna cables and antennas, this happens after 15-25m of cable between the antenna and the GNSS receiver.”
I would strongly encourage testing of this to see if there is any improvements before spending any time designing an enclosure for this.
Good point(s)! I mostly wanted to see if the high RTK mounts could be useful for anything besides extra pieces! I’m still not quite sure what the root cause is for the signal disconnects. Is there a way to determine if it’s a multipath issue vs line-of-sight issue? I was almost positive it was multipath until “El Jefe” (my mowbot’s name) lost signal in an open area that normally has no trouble except there was overcast during this series of disconnects. Has anyone had any luck with with a different antenna gain or shape?
There are some playing with $150+ triband helical antennas (similar to what Yarbo is using). I am sure those will provide a good signal, but probably not enough to justify the costs. Especially if they get broken off by a branch or something else.
As for how to determine the multi path issues, the best way is to look at your Heading DOP in the app. AKA HDOP. You typically want to see this number below 1. Anything over 1 is more problematic. You typically don’t want to exceed 1.3 and 2.0 is unusable. Now this will fluctuate a lot. That’s normal. As long as you are not consistently high, it’s ok.
As for the loss of signal, Yarbo reports GPS errors for a few different things. There is a data path component as well to RTK. If you have no connection to the internet, it will “lose GPS” while it fails over to local mode. If HaLow is down or the signal is greater than -83 it will lose a data path and report loss of GPS. This is where cellular backup is great, because it will fail over to cellular to get to the internet and then your data center and your rover can communicate the correction data with each other VIA Yarbo’s servers. This is the default mode. If you look under the Data Center in diagnostics. This is connection mode 2. Connection mode 3 is local HaLow mode and only works over HaLow or WiFi. Although, if you’ve setup WiFi on your rover, I would recommend removing it, as this will definitely cause communications issues.
You also need a decent number of L2 satellites. The minimum is 4, but it has to be the same exact 4 that the Data Center is seeing for them to be used. So the recommended minimum is an L2 of 10.
The only line of sight issue is the GPS/RTK antennas seeing the sky. The rover and Data Center do not need to see each other. Just see the same satellites.
If you could post screenshots of your diagnostics when you are having the issue, we could see if we see anything obvious. This might just be an issue a lot of have been seeing and that is probably software related. Also, post some pics of you DC install location and the rover where it’s stopped.
@bryan.wheeler - To add more work to your plate, a brief article distilling your GPS expertise for folks to learn from and help diagnose issues would be excellent. I’m sure Yarbo would be willing to post it on their site as a Support KB article, yes? You will get a cookie - I promise.
I fell for the cookie trick last time. Fool me once shame on you. Fool me twice, shame on me. I do like cookies though.
There are some good KB articles on the support site. Here are some.
This article is for the 2023 Yarbo. It is still relevant, but replace RTK Base with Data Center and LoRa with HaLow. Also, do not follow the WiFi recommendation here as there is a communications bug with using both HaLow and WiFi simultaneously. Make sure and delete any associated WiFi SSID’s from the core.
Just to add to what Bryan is saying I did some research around similar issues I was having and if it’s truly multipath related issues then a ground plate is more the typical GPS recommended solution. I was able to make something and it really did very little to help though. The antenna being used in the Yarbo is a decent helical antenna for what it is. Known software issues aside I think it’s much more likely that it’s unable to see the same set of satellites between the rover and the data center due to obstructions and so unable to calculate the RTK error. Each satellite may have a different carrier phase correction so just seeing a bunch of them doesn’t necessarily help. That’s why given enough time it often resolves with various satellites coming into view (although it is certainly frustrating to lose the mowing time). It would be a nice add I think to have some additional information in the debug screen around this (e.g., # shared satellites seen by rover and base).
I read a post that you suggested about turning off the WiFi so I turned that off a few days ago. Here are some screen shots and clips from today where we had overcast but no rain. There were so many dead spots that I thought the amplifiers made it worse. But when I took them off it was just as bad with the regular high mount RTK antennas
DC is on a mezzanine level deck. The 2nd story partially blocks the eastern view of the sky. Anything not blocked by the roof will get blocked by the big tree in the SE corner.
. Lots of dead spots in previously good reception areas.
This is a screen recording when El Jefe is stuck next to the house with an entire view of open (cloudy) sky. There is no LOS to the base station here but both devices have clear but different views of the sky. I’ll try to mount the base station on the roof next weekend so the mowbot can see the same satellites.
What does your DC installation location look like? Your L2 satellite numbers are a bit low. This could be why as your DC and rover aren’t seeing the same satellites to get the minimum.
It’s hard to tell from the satellite view but the DC might not be high enough to see satellites that might be blocked by the house. Looks like the rover is on the side of the house. Just a theory.
Fool me once shame on you. Fool me twice, shame on me. 
I do like cookies though.
