Question for @bryan.wheeler and @Ken -
What range do you get from Halow, before it fails over to 4G?
I’m barely getting 70 feet out of it. From 40 to 70 feet, the constant flopping between Halow and 4G is not useful, lol.
I’ve seen 4G come up in the network circle in the App today, and I’ve got cell off. No clue how that’s happening, but it points to the issues I’ve been having today with GPS, networking, whatever, that I mentioned in the PPP call. A mess. You are seeing oddness as well?
HaLow should be hundreds of feet at the very least - I always forget. What, some 32 acres or something? 70 feet no way. Something’s wrong.
I rebooted everything (WiFi Extender, rover, DC), and it’s slightly better anyway. HaLow is currently showing in that network circle in the App, and I have GPS. And tomorrow…?
My phone has been acting weird with WiFi for the past week or so. My Internet provider (Fidium) may have updated their equipment or something, and my ancient Note 9 may not be coping.
2 Likes
I’m having HaLow issues myself. I used to have no issues but it is getting progressively worse for me. I have run entire plans on cellular lately. I am working with support on it. It’s amazing how well it flips back and forth from HaLow to cellular. I lose the app during the transition, but it comes right back. The rover never feels it, it just keeps working.
Have you tried checking your HaLow antennas and making sure they are tight on both the DC and the Rover? The rover has a rubber vibration washer that sometimes will cause a poor connection if the antenna isn’t tightened enough. You can try removing the washer and see if you signal improves. Look in the diagnostics and anything below -80 it will be on HaLow. At -80 it starts transitioning and at -83 it flips entirely to 4G. -90 is off the scale unusable.
Depending on where your DC is mounted, there are other factors that can impact the signal too. If the DC is ground mounted for example and the rover goes down a hill effectively below grade of the DC, well the signal will not penetrate earth. If the DC is up high, it has a better chance of having “line of sight” (doesn’t need it to work, just easier for the visuals) and the signal being above the ground.
My current setup I used to have -10 on my HaLow signal strength while docked. The DC is 4FT away from the rover. I am currently anywhere from -28 to -41 on HaLow signal strength. This is enough to make it unusable in a good majority of my yard.
Try putting your rover as close as possible to your DC and see what your signal strength is. It will be higher if the DC is roof mounted. I would expect -40 to -50’s. It shouldn’t vary all that much though. It’s supposed to have an effective theoretical range of up to 32 acres and I’ve seen many user tests/PPP tests where this is true. But, structures and ground, etc can impact the signal levels and decrease the maximum effective range some. For most properties, this shouldn’t be an issue. Especially since the mower is only rated for 6 Acres but theoretically can do up to 12 if you only cut once a week.
Was that the long answer you were looking for? 
4 Likes
It’s a start, lol. I very much appreciate the specifics!
DC is mounted on the rails of a raised deck, probably 8 feet above ground.
At 25 feet away, signal reports at -75 ~ -78.
Directly underneath (8 feet below), -73 ~ -75
Reseated whips on core and DC, no change
Pulled rubber gasket from core, 20db improvement
It’s now -55 directly under the DC
…which is still wrong, lol.
Taking the core around the corner of my house, 4G immediately. Not even 40 feet. Straight LOS away from the DC isn’t much better. Manual control via the app is pretty much unusable if Halow is a thing, but I’m glad it isn’t just me.
Since this is new behavior to you, me, and Ken,
AND remembering that there was a DC update a short time ago,
I wonder if someone at Yarbo had reacted to some EU rules that the FCC doesn’t have, here. I recall that “someone else” discovered they were violating EU rules last year, and boy was that a trainwreck for their RTKs. Even moreso on the forums, since the US market had full power RTKs that work. This year, their new ones all come with a cellular RTK service. Amazing!
Since you’re working with support, and if someone at Yarbo DID accidentally drop the DC’s output power, or they restricted the transmit monopolization because of EU rules, they’ll hopefully roll those changes right back with a hotfix tomorrow, because EU. rules. do. not. apply. here, and those changes have broken their product 
3 Likes
I have suspected for a while that it’s a bug with the Transmit Power Control AKA signal-strength-based power adjustment feature. I am not sure if the DC even does this or not, but it seems like it was enabled at some point and is flaky. I replaced all my HaLow antennas and actually got back to -14 for around a day. Then it went back to -28.
It was weird too, because I removed the original antenna (started at -36 before doing anything) and it always goes to -70 with no antenna on the dock 4ft away. Expected. I removed the rubber washer and put the antenna back and it was now -58. Seemed like a failing antenna, right? Got my spare out, swapped it in and kept hitting refresh. It stayed at -58 then dropped to -45 and then -36 and kept dropping to -12 but eventually settled out around -14. These in between numbers aren’t exact, but you get the jist.
So based on that, I suspect power output is being throttled automatically and has an issue. And it’s not just you. I’ve had several users reach out to me and one PPP as well.
The European theory is interesting too.
2 Likes
If you’re familiar with the EU thing, you can skip this reply.
Otherwise,
“That other company” got nailed for being overpower, and for monopolizing the channels.
The final post I saw mentioned that they’d gotten the power dialed back, but were making xmits over one second in length across the various freqs, each of which is still WAY to long for EU.
The US version just blasts the airwaves as loud as it wants to, nonstop. And it’s fine, I get almost a mile range from mine.
If you get into a discussion about this with support, things to look for besides output power is burst length and time domain. EU prohibits being keyed up for more than some fraction of a second, per second, so that’d be the other aspect that someone would mess with. And again, none of that is appropriate for the US market.
Did you ever get confirmation on what freqs are going through those halow whips?
I’m half tempted to just make some full-waves and let the core drag the wire around, lol. And there’s no reason whatsoever, at least in my application, for a quarterwave or halfwave on the DC. Assuming it isn’t some 100m band, lol
3 Likes
HaLow uses the 900mhz frequency bands. Like the old cordless phones.
2 Likes
For some fun entertainment,
a) Cruise around and watch halow strength in diags
b) shut down core and DC. Swap the whips from core and DC.
c) Cruise around and watch halow strength in diags
For those reading, remember to shut things off before removing an antenna. Xmitters don’t like it. At all. Like, not even a little bit.
For me, it’s looking like tolerance of the connector vs the tolerance of the whip. Might be the whip for the DC, might be for the core, but it smells like one of them is bottoming out before the post has a good seat. The rubber-seal thing you had me remove is compatible with that idea.
I’ll measure them all after dinner.
2 Likes
Names of things:
Doesn’t matter if it’s the male or female. Turns out, they’re the same basic shape but one does not have a pin, and the other is upside down.
The female on the core, obviously, does not have a pin. The base is the top of the hex nut. The “plastic” is that gold socket in the center. The base dictates how far down you can screw the whip.
Top to base - 4.16mm
Top to plastic - 2.03mm
Plastic to base - 2.13mm
Core’s whip has a pin, plastic, and a top (shown above).
Top to plastic - 4.22mm
Top to tip of pin - 1.92mm
Pin length - 2.30mm
That gives us about .21mm of pin insertion with floppy antenna on the core.
Datacenter halow
DC female does not have a pin. Base is a soft rubber seal. “Plastic” is the little gold socket in the center. The rubber seal dictates how far down you can screw the whip.
Top to base - 4.72mm
Top to plastic - 2.03mm
Plastic to base - 2.69mm
DC whip has a pin, plastic, and a top, per the diagram.
Top to plastic - 3.58mm
Top to pin - 1.20mm
Pin length - 2.38mm
Insert length is 1.49mm
Looks like the Top-to-plastic on the Core Whip, and Top-to-Tip-of-pin is what’s costing us that 1mm on the core. The plastic on the DC whip is much more shallow. It core’s pin DOES connect, but could connect more better. If I were to guess, the 1.5mm hex nut holding the core’s female might be related, since that is the “base” in this setup. It is probably eating about half a mm of our pin. For yucks, I’ll remove that nut, later, and see if signal improves.
I am, however, probably wrong. Completely. It’s confusing enough that I tossed it into CAD to verify the math, lol. And I’m still probably wrong.
One way or the other, we’ll find out in our next installment!
1 Like
Stay tuned kids! Same Yarbo time….Same Yarbo Channel!
@Ken bring beer 
1 Like
@bryan.wheeler and @Ken
Success!
-69db at about 250feet through steel BBQ grill, a house, and a maple tree.
It used to be -75ish directly under the DC. Now it’s -23.
Yay!
Now we just need a crap-ton of low profile hex nuts, lol.
1 Like
Let it bake in a while. Mine is intermittent. Usually only lasts a day or two.
(Parenthetically…)
No problem with HaLow (-42 with rover underneath the DC, could be better I suppose, but doesn’t seem to be a problem)…but my rover gave up on GPS all day yesterday while I was away and remotely running it. Different topic I will be looking for on FB to see if anyone else has hit, 'cuz I’m to the point of thinking this ain’t me because reasons…until it is…ugh.
Ayup, had TWO of those beers yesterday, @bryan.wheeler, after being in the sun for a while watching the GF’s daughter and her classmates graduate into full-fledged RNs. So awesome – the beer (and the GF’s daughter’s huge accomplishment, of course, yeah, that, too).
2 Likes