The Open Source Hardware Conundrum

It's Tricky

I came across an interesting post from the blog of Julian OH8STN (one of my favorite amateur radio YouTubers) that briefly talks about a problem that DL2MAN, the developer of the (TR)uSDX transceiver is having. Unfortunately, the original video from DL2MAN discussing the issue is removed, but from what I can suss out from Julian’s post and in doing some more research, DL2MAN was getting frustrated with clone manufacturers getting the benefit of his (and I assume team’s) work in developing firmware for the (TR)uSDX. He seems to have moved to a model where the firmware is now keyed to a serial number in the (TR)uSDX, and you can only install firmware updates with firmware coded with the internal serial number of the radio, in order to thwart free-riding cloners

¹

.

Photo by Joshua Earle on Unsplash

If I understand the (TR)uSDX radio correctly, the hardware design is open source but the firmware is closed source. From the DL2MAN website:

Software, Schematics and PCB Layouts are licensed under a Creative Commons Attribution-NoDerivatives 4.0 International License.

This means: Even commercial use is allowed as long as you don´t change anything! No Royalties charged, donations welcome.

However, on the firmware page, no license terms are given and the ability to download is restricted to those who enter a valid serial number.

According to Julian’s post, someone has created a method of circumventing the firmware lockout restriction:

Now, some guy (Rob) comes along, trying to profit from the project, circumventing license protections with his own firmware release. The result is that Chinese clones can now get updated firmware, while the project team gets stuck with all the support requests, from clones not attached to the project.

In trying to do more research on this bootleg firmware, I had some real difficulty finding out more about it, using Brave Search, Google Search, and Bing Search. I don’t know if that’s because their algorithms suppress these results or the bootleg firmware is just genuinely that obscure that it’s not even indexed yet. It doesn’t really matter for the purpose of this post however, as I’m not here to specifically write about the (TR)uSDX, but about the big problem of selling open source hardware in general: the amount of free-riding clone sellers who will profit from your work as a developer, and the large amount of free support you’ll be expected to provide to the users of any and all products that are clones of yours or even related to yours.

I’ve long been a proponent of open source hardware in amateur radio, because I believe we hams should have real ownership of our radios and other gear. Watching a few Louis Rossmann videos will quickly show you the many downfalls of proprietary, closed hardware systems, if you haven’t personally ran into them yourself. However, running a very small open source hardware (and firmware/software) business has shown me that the great weakness of that business model may be very difficult to overcome. It’s not like this is a new problem and that it hasn’t been pointed out from the beginning of the entire concept, that it would be difficult to successfully pull off. It’s been done in the software domain, but hardware (and perhaps even more specifically the combination of hardware plus firmware) is a tough nut to crack.

With Etherkit, none of my products even became popular enough to warrant exact clones, although there were certainly a lot of places making Si5351A boards, which was my most popular product. That wouldn’t have been a problem, fair competition and all that, but when you pair it with the fact that I had authored a library to use the clock IC with the Arduino IDE, and that library became fairly well-used, then that’s where the OSHW problem began to manifest for me. One can say that it’s fair play for someone to buy the Adafruit board and then come to me expecting free support to use the library I authored to run on their hardware. However, it’s also disheartening to have so many people put their money with someone else and then come to you only when they need something from you for free. And while some folks were polite, others were more entitled and demanding. Larger operations like Adafruit have the resources to deal with these kinds of requests; smaller operations such as myself or DL2MAN do not. Others get the reward of pumping out a product with little-to-no development cost, no support cost at all, and yet we get to do the heavy lifting.

Now, all of this is not a way for me to moan about my lot in life. I made my choices with open eyes, and I don’t regret dipping my toes in the water even though I’ve failed to really get it to take off fully. I’m writing this so that those of you who may also support the idea of a more open hardware ecosystem can more fully understand the difficulties that the few open source hobbyist developers face. So that you can adjust your spending habits and your expectations if you want to nurture the idea further. As Julian says at the end of the previously-mentioned blog post:

For those of you who may not understand what the problem actually is. This is how the mcHF was utterly destroyed. Greedy Chinese copycats not honoring the license, and cheap ham radio operators, not supporting the original developer. Perhaps this is why development in ham radio, is practically stagnant! Food for thought anyway.

So is the small open source hardware business model sustainable at all? I’m still not certain. I’m going to give it one more try, because I hate giving up after failing to achieve my goals. It’s worth trying to do. I am going to have to figure out a better way to navigate this thorny problem. Especially as I most likely won’t have any employees (a least for a while) and will have to do everything myself or subcontract it out. If I make a successful product, I’ll never be able to undercut the cloners, so it’s probably going to be up to education and persuasion, to encourage people to not buy dodgy clones from a no-name Aliexpress shop. I do think that the way that the open source software industry monetized through offering paid support was a great way to go about it, but I’m not sure that would ever fly in this business. This is also one reason why I’m using Substack for this project, because I’d like to find new ways to make this type of business self-sustaining. Whatever form a future successful business of this type will take, it most likely won’t look quite like something you see now.

Project Yamhill Progress

So you’re getting the philosophical post this week because I was unable to make much progress on the Project Yamhill front panel PCB. Some more pressing issues came up on the homestead which required my attention, and which unfortunately did wonders to my back as well. Ah, middle age!

I did receive the tinySA Ultra that I ordered a while back (and yes, I did order from the official store linked from tinysa.org in order to properly support the developer). I’ve played with it a little and so far I’m pretty impressed at the power you get for the price. I didn’t realize that it had built-in measurements for many parameters such as IP3, phase noise, SNR, 3 dB bandwidth, and noise figure. At this point, I don’t know how accurate these measurements are, but even if they are reasonably accurate at the frequencies that we hams typically homebrew gear for, then it’s an amazing value. My Rigol SA won’t make these measurements unless I shell out a lot of money for the key to turn those features on in firmware.

tinySA Ultra measuring a simple carrier from Si5351 Breakout Board

In the above photo, you can see that I did a simple check to see if I could correctly hook up a RasPi Pico to a Si5351A Breakout Board and get the basic example sketch to output a signal at 14 MHz. The instructions for Project Yamhill will be written so specifically to cater to hobbyists using equipment like the tinySA and NanoVNA, instead of assuming that they have more traditional (and expensive) test equipment. I will be validating the measurements with my bench equipment, but I haven’t seen anything to suggest that these more affordable options won’t work, but there may be some limitations such as with RBW, noise floor, and dynamic range that may need to be worked around in certain circumstances.

At first glance, the tinySA Ultra is impressive. What would make it a killer app for me would be a version with a larger screen, such as 7, or even 10 inches. I can read the 4 inch display but it’s a little uncomfortable for my aging eyes. The same device in a large tablet form factor would be so nice to have. That’s my only real “complaint” so far. Otherwise, I’m thrilled that we hobbyists have access to measurements that were nearly out of reach for most of us just a decade ago.

I’ll be back at work on the front panel starting tomorrow and only plan to be distracted by the maiden launch of Starship this week, since the weather continues to be cold and wet here, meaning I won’t have to be able to do work on the greenhouse that Jennifer wants us to build.

1. If I’m wrong about this, please correct me in the comments. I don’t know much about this project other than hearing about it briefly. I’m mainly using this as a story hook more than writing about this specific product.