Devices similar to gotenna

Interesting, but also intimidating to those of us who don’t develop software.

The idea of open source can be very appealing to those with such skills, but leaves the other 95% of humankind rather baffled. I say this not to deprecate the idea of open source, but to point out how it’s benefits mainly accrue to the few who do have such skills. Open source makes the inner workings of such projects more usable for the few, but tends to make the outcome more limited for the general public.

A good example of how this happens is the first question I have: Is this FCC-approved? Digging around in the extensive documentation briefly did not turn up a ready answer.

I suspect that with 34 bands across such a wide spectrum (699 Mhz to 2.140 Ghz) that large parts of this bandplan are significantly restricted according to which country it operates in. This would tend to make it more useful for hams, but could lead to others operating on bandwidth that is not authorized. This likely legally works for selling kits to roll you own, but may not allow the sale of consumer devices without such functionality restricted according to legal parameters extant in each jurisdiction - which is what goTenna is able to do.

I share your concern, and the lack of plug-and-play software is why I was reticent regarding the and CATAN projects. The neat thing about this is the open source software already works for end-users. If one isn’t a developer, it’s there. If one wants to extend or improve it, or if the company folds, that potential exists.

Yes. Here is the page listing the various board certifications. The relevant boards there are the LoPy and FiPy.

The end-user PyGo devices (similar to GTM) are also FCC, CE, RCM, IC, and LoRa Alliance certified [PDF]

The way I mentioned this above likely added confusion. That 34 bands comment only applies to the cellular/LTE chip on the board, not the other 4 radios (WiFi, Bluetooth, LoRa, or Sigfox) on the FiPy. In the US, it works on Verizon initially, and potentially other providers as they roll out
LTE-M CAT M1/NB1 services (cellular data for IoT devices).

The PyGo devices are… certified for FCC, CE, RCM and IC… cellular network certified to work out of the box [source]


Not that GTM doesn’t have a lot of advantages, but PyMesh currently has an edge on this point. GTM is not FCC approved to sell the boards, hence the hacks required to add external antennas. PyMesh dev boards with antenna connectors (LoPy, FiPy, etc.) and built-devices (PyGo 1 & 2) are both FCC approved.

It would be awesome if goTenna released a dev board. I’ve seen others make that comment in other forum threads.


Thanks for locating that FCC info. It’s instructive.

In the 900 MHz ISM band, the devices are limited by the FCC to under 0.1 W. In other bands, they are allowed up to ~ 0.2 W. That’s starkly less power than the 1 W GTM, suggesting that the flexibility on other points, including ones attractive to open source projects, by the FCC is possible, but only at the loss of capability.

I could be wrong, but I doubt this is mere happenstance. I think the FCC is being pretty cautious about the power available to users, one of the big factors I suspect the FCC is concerned holds the potential to cause interference to other ISM band users unless it is closely controlled.

I suspect that even leaving other factors aside, it’s FCC regulation which is the primary factor limiting goTenna from releasing code that is open source. It might be possible to release development boards for the GTM, but I’d bet if this were allowed their output would be more in line with the Py hardware than it would be with the current GTMs.

1 Like

LoRa uses a proprietary chirp spread-spectrum technology specifically designed to squeeze out every last bit of range while mainaining low power consumption. I don’t know enough about the modulation goTenna uses to make a theoretical comparison. I do know there’s more under the hood of digital RF connections than power alone, so I’m deferring to the tests and experiences of others, including yourself.

From what I’ve read in the forums so far, the longest ground-ground GTM connection is around 50 mi. “The guy with the Swiss accent” on YouTube successfully connected 2 LoRa radios at 201 km (125 mi).

That isn’t to say goTenna and/or LoRa can’t go farther by changing variables. Either way, I’m evaluating options that will handle roughly up to 11 miles (with elevation) between repeaters, so both are sufficient in theory. What would be a significant difference, as you point out with FCC power restrictions, is the LoRa connections are being made with less power. I read somewhere in the forums that the GTM devices have been measured in the 140-200mA range for TX, compared with LoRa at 20mA. Even if I don’t understand the secret engineering sauce either are using, that’s a significant measured power difference which has real implications, particularly in sizing battery and charging systems for base/repeater nodes.

My suspicion is that GTM will handle more connections in a mesh, and that choices surrounding the RF variables were made to support more connections while sacrificing some power efficiency. PyMesh tops out at 200 mesh connections, which if I understand correctly, is partially a limitation from the use of LoRa. That 200 can be increased with gateways, but it’s worth considering.

For my use-case, power consumption is more of an issue than 200+ local connections. But I’m aware that being off-grid 2° below the Arctic circle with no other wireless options makes me an edge case.

If I could more easily attach an external antenna to a GTM via a connector, I would already have ordered some. Alas, it seems the FCC rules which apply to goTenna devices makes that a dicey enough proposition that it’s not (yet?) an option. That’s not a deal-breaker, it just makes my decision harder.

1 Like

Oh, that and the thing where if the GTM loses power it doesn’t resume its prior state. Soldering on an antenna connector is way simpler than building another system to reset the device into a repeater node. That one miiiight be a deal-breaker for me. Maybe.

1 Like

The nominal wattage is often cited as 1 watt for North America. In Europe that’s reduced to a half-watt. The testing docs linked here measured around 0.8 watt (in Section 4.3.3).

IIRC, the modulation schemes used are different between the devices, which could account for a difference in long range propagation. But I tend to look at reliability at normal ranges as a better measure for what difference power makes.

Ah. Thanks.

I think I was remembering someone’s numbers in the forums who was measuring the USB draw during TX on their own device. That range matches up with your numbers.

140mA x 5V = .7W
200mA × 5V = 1.0W

1 Like

Another interesting fact is that there are at least 961 other devices that share the same spectrum under FCC jurisdiction. The FCC isn’t trying to make things difficult for goTenna, they’re just trying to protect a valuable and very much in demand spectrum. That’s also another reason why I conclude the FCC is quite unlikely to approve a box chock full of goodies that would let people fiddle with much of what remains locked down under the hood of the GTM.

Those who pine for a turn-key, open source radio device similar to goTenna Mesh to experiment with are probably well advised to start working on that ham license, with the possible exception of a developer board that might be released at a substantially lower output. However, even there I think it’s being too simplistic to assume that is the only area of concern to the FCC. The message per minute limit of 5 is one example. I suspect there’s a range of choice that would be allowable for the designer, but that the FCC determines how much total in a minute is allowable based on giving other modulation formats a chance to get a word in edgewise. If you accelerate the message rate high enough, you could turn such an unlimited device into a jammer.

I’m sure there’s plenty more possibilities for why devices like this face a steep regulatory hurdle to utilize these very effective and efficient modulation schemes at anything other than the most modest power when available to the public in unlicensed form. The FCC has slammed this barn door shut to that, likely for good reason, so that we may see incremental change continue there, but only after being thoroughly justified and with the assumption that the limits the FCC specified continue to apply, although subject to modification.

This statement is confusing. For clarification, what are your specific objections to PyMesh based on LoRa?

It doesn’t require a ham license. And anything requiring a ham license would already not be similar to GTM. Well, except in the case that some of the hacked GTM devices deployed as base nodes are probably already in violation of FCC rules and should require ham licenses and removing all encryption and forbidding tranmission of any commercial messages. I’m not the RF police, but for comparison’s sake I do wonder whether these discussions always pit theoretically legal apples against practical [il]legal oranges.

LoRa is fairly well established at this point – both in terms of real-world use and institutional backing. The LoRa Alliance is supportrd by IBM, Alibaba, Google Cloud, Cisco, Duracell, Charter, Inmarsat, and a bunch of other legit companies.

There aren’t any observable FCC hurdles with the Pycom devices or the underlying tech, but you keep going back to that. I can see your point on the various other devices which remain vaporware. However, LoRa is known and documented and FCC compliant by design. It’s already been adopted as the de facto wireless technology in the IoT community for the augmentation of features WiFi and BLE can’t efficiently handle.

It would be one thing if LoRa didn’t work or was a mystery. I don’t see how that claim could be made at this point.

All I can tease out is that you’re assuming goTenna is inherently better than LoRa, largely based on one data point: transmission power. Based on my observations, that assumption may very well be incorrect. There are good explanations of why the CSS modulation LoRa uses works better than alternatives at low power, but I’ve been absorbing that stuff via YouTube so it’s not as easy to search/cite. I can say that one doesn’t have to look far to find end-users reporting poor range for GTM devices in normal circumstances. While I don’t think that’s fair, and that it likely reflects a gap between unreasonable expectations based on cell phone use, I don’t see any reason to believe GTM blows other technologies out of the water in the 1/4 to 1 mile ranges we might consider normal use.

Not a big deal. Having not yet committed funds to GTM or LoRa hardware, I don’t have a dog in the fight.

1 Like

No objections to LoRa at all. One of our node hosts is involved in disaster radio and it looks very interesting.

Power is somewhat of an issue, but perhaps not the way you might think. In building UMESH, I’ve been lucky enough to depend on connections based on long residency and reputation for getting things done in a relatively compact town with uncomplicated geography. A non-profit startup like that (and I have no ambitions other than setting up a mesh network for public benefit at minimal cost and essentially no fund raising because we just don’t solicit donations) is in good position to do a lot with relatively little.

That’s largely due to being able to call on supportive people who are well located in relation to each other.

Then imagine needing to have five or six times as many supportive parties to host nodes because of the shorter range that results from lower power. From an organizing perspective, the hill suddenly grows very steep.

Essentially, it’s the lesser density of adopters required to establish an effective mesh network that is favorable to goTenna Mesh for my needs. It doesn’t mean I don’t like the alternatives, I just have a hard time grokking how big an organizing effort is required to cover the same ground.

Personally, I’m very much in favor of adopting alternatives, because you don’t want to get boxed into any single form of communication.

BTW, in the middle of dinner, so will be back to complete these thoughts shortly…

The fact that GTM is a turnkey solution (except for that pesky relay power issue) to creating mesh is a big mark in its favor. The Maker folks are a blessing to anyone concerned about the younger generation getting their hands on good stuff they can make themselves. However, the idea of dozens of people soldering up their own devices to build a community network is a steep challenge. Someone could meet it, but it’s gonna take a dedicated effort. I’d certainly participate, I like building stuff, but the scarcity of folks like me are one reason why there’s no local Radio Shack anymore.

As for legalities, I’m simply trying to emphasize that the FCC has exerted its regulatory authority rather differently in the cases of devices with about a 1/4 watt vs those that have a nominal 1 watt output. That’s meaningful. But it’s also important to parse the differing legalities between the manufacturer of a FCC approved device and those who use it, whether or not a ham. Things like the requirement for a non-removable antenna are a very straightforward way to limit the possibility of consumers easily “upgrading” their signal. That doesn’t prohibit the consumer from doing so, just that they will be held responsible for significant violations that might result. What’s good about being a ham is that you pretty well can figure out that line and stay on the positive side of things, unlike someone hacking something with brief internet instructions. You can hack your antenna, goTenna just isn’t allowed to deliver an easily hackable antenna to you.

I’m not a FCC-experienced attorney, so my nuances may be a bit off, but it’s a distinction that is important.

There may be a gap in our communications. The objections you’re raising generally apply to projects discussed in the past, but not so much to PyMesh. They already have the following…


  1. GTM-esque turnkey devices. These are the PyGo1 (WiFi, BLE, LoRa, Sigfox, GPS, accelerometer) and PyGo2 (same + cellular/LTE data). Both are FCC certified. Both have OLED touchscreens for notifications.
  1. A range of dev boards with the combination of all of the above radios, all with external antenna connectors. All FCC certified.
  2. Dedicated gateway board shipping in June.


  1. PyMate for iOS and Android. Includes mesh chat, etc.
  2. PyLife for iOS and Android. Launching in June. Adds mapping, tracking, etc.
  3. PyBytes middleware for dev boards.
  4. PyMakr plugin for MicroPython coding in Atom and Visual Studio Code.

The uniqueness in the various bits of the Pycom ecosystem is that they have both turnkey devices for non-techy end-users, like GTM, and the dev boards for those into soldering, coding, hacking, and making, like (and the myriad others). There are paths and options for more types of users than the previously discussed projects.

1 Like

What I think is important to note – respecting this is the goTenna forum – is that goTenna could probably open source the software, and continue to monetize the hardware and services. LoRa, Sigfox and other hardware components remain proprietary, but work with open source software. The companies involved, including Pycom, accrue revenue from hardware sales. Since all the GTM software is already free, the incentive to keep it all closed is lower than it would be for a SaaS company. And the Plus upgrade could still be an add-on paid service.

I love the efforts to integrate GTM with Signal, CATAN, and any of the many other open source alternatives. There’s big potential to grow the user base by being fully integrated with some of them. Licensing hardware to hundreds of millions of users isn’t usually a bad thing for a company. Unfortunately, those projects, and the open source community generally, will always be wary of putting their own energy into integrating closed services with their projects. And no, I dont think the SDK solves this for most other projects. So @rmyers and others can do a ton of awesome integration work from the goTenna side (let’s call this “push”), and some people in the other projects will think it’s awesome and recognize some potential, but it will be an uphill battle getting traction with the other projects’ developers (let’s call this “pull”).

I could be wrong, and goTenna might fancy itself a SaaS company; in which case most of this is probably wrong.

Anyway, I didn’t intend to launch into a diatribe about business goals. I just think the PyMesh stuff is cool, and suspect others do as well.

1 Like

Yes, that’s a good thing. But looking at the choices already available sort of made my eyes glaze over. The Py folks are going to need a clear way into that ecosystem that’s easy to grasp for what they see as the average user. If I need to decide much beyond Mac or Android, which I suspect tends to be the thinking of most potential users, then I quickly am baffled by what I should do.

Accessibility for people who feel those limits, rather than those who are comfortable in making such choices, is the crucial barrier that needs overcome on the road to widespread adoption of any of these devices.

Perhaps the word “devices” is the inflection point. I suspect there are many more LoRa chips sold than GTM units – probably by an order of magnitude or more. The question becomes about getting one’s hardware into every device in an open network of potentially billions of users (how I’m interpreting “widespread”) versus trying to build and own all of it. I guess if a company thinks it can pull off owning it all, then the risk in slogging through until serious network effects kick in might be worth it.

Apple and Samsung would be the analog to goTenna (Pycom, or hardware company X) in this way of looking at it. They make billions in profit from hardware, but they don’t need to own the phone companies and network (mobile internet) too. And they don’t need to own the software running the internet, most of which is open source projects like Linux and WordPress.

I am getting ready to go refresh a couple of relays and put up another, so will probably have some more thoughts later. But I don’t think that goTenna sees itself a hardware company as much as it sees itself as a systems developer.

I also suspect that goTenna, like any company with considerable investment, has some reservations about just giving things away. But I also think that’s the lesser part of their reluctance to go open source.

As I’ve tried to argue, it’s mostly the FCC that is riding herd on such ambitions. An open source goTenna with 1/4 watt output would likely get FCC approval, but then it’s competing with all the projects, open source or not, already available at that lower power.

My conclusion: It’s the FCC that people should be pressuring, as I doubt they would approve an open-source device at 1 watt at the moment. That might change, it might not. But without that restriction, an open source version would tend to raise an issue like message per minute limits to the forefront. I presume there’s no way to lock down such limits with an open source version or it wouldn’t really be open source? That’s what I see as the roadblock here.

This conversation seems to be muddied by excessive breadth. That’s fine; there’s a lot of interesting topics at play. But I don’t think we’re using “open source” the same way.

Unless I’m understanding it wrong, goTenna could sell the GTM units exactly as they are, and open source the software. The firmware and/or the apps could either/both be open-sourced. None of this would necessarily change the capabilities of the hardware in the eyes of the FCC. There still wouldn’t be an antenna connector, for example.

It certainly may be the case that changing the firmware would change the capabilities in significant ways. In that instance, the apps could be open-sourced and not the firmware. Whether there’s value in that is an {cough} open question. And it’s hopefully the case that there will be new hardware versions, and future versions could be tailored accordingly.

A minimal step might be for goTenna to rebuild the Android and iOS apps in open source, but only using the existing SDKs. This would keep all of the currently proprietary bits proprietary while laying the groundwork for others to contribute in a meaningful way.

I’m not going to try to define open source. It’s use seems to vary according to what one wants to include or not. Like all things, it’s use can be somewhat subjective.

Here, I’m talking pretty much about the firmware in the device. The app may have some interaction with what I’m pointing to, but it’s the firmware that defines the unit’s interaction with the mesh. If you look through the FCC certification documentation, to me it looks like the firmware defines many of the limits of the unit and these are pretty sharply defined, number of frequencies used, and how they’re used. This is noted in the documentation starting on page 4 at this link:

As a frequency hopping spread spectrum radio. the GTM accesses the limited amount of airtime available in such a way that others are able to likewise do so at the same time without interference. If you dig a little, you’ll find built-in to these requirements are the use of at least 50 frequencies if using 1 watt of power, or between 25 and 50 freqs if using 0.25 watt. Note this distinction as I’ve mentioned it before. If the Py devices are certified at the lower power, then there are fewer freqs available to use. The user may or may not notice a difference there, depending on other traffic. It’s nonetheless one of the regulatory limits I’ve been referring to that affect all users and hence are determined by the FCC, but written in the firmware.

Other measurements of significance seem to refer to how the firmware negotiates its management of access to airtime as its signal is sliced into digital bits and broadcast on many freqs so it can be reconstructed by the receiver. These include, besides RF output, the number of channels used, average time in use, and various measurements that seem intended to ensure adequate performance with digital signals that will not interfere with other users of the spectrum.

There is a range of acceptable values for many of these, but likely little room for some of the most requested changes, like increases in number of hops allowed, manipulation of the transmitter to out of allowed bandwidth, or even operation at higher than authorized power. We know that these are controlled by the firmware and can only presume that, where there was open source code involved, that the end-user would have access to change these values at will.

Concerns over that aren’t restricted to the FCC. goTenna assuredly would want to preserve a positive user experience for those who consume its products. Their chief scientist has explained why uncontrolled numbers of hops quickly makes things unmanageable at higher hop counts. Turn something that would have an easily manipulable hop limit loose at the same time the rest of us are using our GTMs seems a recipe for creating problems for not only other GTM users, but other users of this spectrum.

So I guess my question about how you define open source would be If goTenna permitted access to all the goodies EXCEPT for those limited by the FCC, would it be considered open source? I suspect that we’re most of the way there already on that. If that is the case, goTenna has likely done about as much as it can as a manufacturer - and still be able to produce and sell a 1 watt FHSS device.

Consider that the FCC chases down and seizes CB radios that can be easily modified to operate out of band and at higher power than legal power. I really can’t imagine the FCC keeping that hardware off the market, while making it easy to use soft- or firmware to do basically the same thing with a software defined radio - able to be hacked with a few keystrokes into illegal territory.

1 Like

I made a post asking how Gotenna compares with a competing product, The whole thing got deleted without notice or comments.

I was exited to setup a local mesh and have brought and given away about 16 unit.

I’m cancelling that project until they have a better unit for stationary mode and more open software.

Similar device was Radacat