Making OSHW design a viable occupation: The OSHW product "Introducer"

[OwO]

designing my own downconverting mixer based on the BFP740 transistor.  That is one reason why I have been prototyping mixer and amplifier circuits based on the BFP740 transistor.

Dude, that is exactly what I'm doing too, and exact same transistor too. Are you doing a switching mixer or Gilbert cell?

[OwO]

This paper goes into detail about the legality of copyright protection on PCB layout designs in US and China: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2337014
The consensus is the functional aspects of a layout design are not covered by copyright, and only if the board design has aesthetic aspects can it be protected, but then only those aspects. Similarly a schematic design document may be copyrightable, but only the document itself and not an implementation of the circuit.

[profdc9]

I am using a half-Gilbert cell or single-balanced mixer.  I may go to a full-balanced mixer but the distributed parasitics of a larger circuit may obviate the benefit of that.  It is part of a bigger project called the RFUtilityKnifeBFP740:

https://github.com/profdc9/RFUtilityKnifeBFP740

This is a combination of five circuits I am working on including the mixer.  Because it basically costs the same to make five circuits on one board from JLCPCB, that is what I'm doing.  The mixer is on the lower left.  I publish my experiments on github as I work on them.

I have attached the Qucs-S (qucs with ngspice) simulation of my mixer if you want to check it out.  You are free to use it under the CC-BY-SA 4.0.  I hope it may help you with a new VNA or SA design.

Dan

[rhb]

Story is very simple:

- 2016 - edy555 (aka ttrftech, Takahashi Tomohiro) designs the NanoVNA from scratch, with his own schematic, PCB layout, and firmware. The schematic and firmware are public and GPL licensed. The PCB layout is unreleased. edy555 sold NanoVNA kits for a short while.

Thanks.  That pretty much matches up with my understanding.  However, there is animosity  among some towards edy555 because he used Tom Baier's VNWA circuit.  My response to that was even a patent doesn't last forever.

I've tried to find a way to contact edy555 so I could send him some money via PayPal, but never found an address I was confident was his.

Reg

[floobydust]

Is there a way to make OSHW thrive- what would have to change?
I'm asking because if the concept is a fantasy, a bullshit notion, then can we end the dreaming about it already 

I am surprised to see OSHW concepts are largely communist philosophies:

"... that there should be no patent laws and an elimination of intellectual property protection entirely.  The idea of commercializing ideas goes against the public ownership of all property produced by society."
"the {patent} incentive offered is a monopoly which increases the private profits realized from the invention rather than the benefits to society in general."

So all designs are for the benefits to society in general- it's wrong to own any IP and have a monopoly for private profits. Sounds like OSHW to me.

[rhb]

I certainly don't endorse the communist doctrine.

My desire is to lower the cost of T&M kit such that  recent EE grads can afford to have a working lab of their own.  There are many EEs who just view it as a job and would never spend a nickel on it, but they will be left by the wayside fairly quickly.  My concern is those who would have their own lab if they could afford it.

I'm also concerned about the lack of component level repair information and lack of FW access to fix bugs.  While FW access was not available from Tek and HP, component level repair data was.  But that is now seldom the case.

OSSW works because it spreads the cost of SW across many corporate entities who make money selling services.  So what I am suggesting is spreading the cost of HW design across multiple manufactures by selling manufacturing support services with sufficient revenue going to the designer to allow them to make a living.

It may well be that to be successful OSHW only  schematics are OSHW and Gerbers are licensed.  And perhaps the UI is OSSW, but the FPGA bitstream is licensed.

I don't know what the answer is.   But proprietary only designs result in a lot of reinvention of the wheel.

It may be that OSHW is only for hobby projects where the designer does not need or want to try to generate revenue from the work.  I'm redesigning a 2 transistor and single IC 40 m transceiver.  I have *no* desire to sell it.  But I'd like for people to be able to buy it cheaply.

Reg

[floobydust]

I'm just looking at the politics of open-source hardware and the resulting clash. It purports the designs have to be open, source files available to everyone, for the greater benefit of society. No individual alone is allowed to profit. It's very noble to make a design and give away all the IP and files.

But you are advocating a monopoly- select manufacturers that can build a widget and make profit.
So the IP has to be protected somehow, otherwise the vultures swoop in and rip it off... which is what we're seeing happen with OSHW projects. Designers are upset because their sweat and toil gets bastardized or others profiteering from their work. So projects are limited to a flash in the pan happening, one-man band kind of complexity.

Enforcing any copyright or licensing without funds for legal support and all the while dealing overseas with a country that does not support IP protection, seems doomed. The project has no money to protect or look after itself in any way, it's destined to get ripped off by the vultures.
"Creative Commons is only a service provider for standardized license text, not a party in any agreement."  "... All copyright owners must individually defend their rights..."

I don't have an answer because I see it as a clash of political systems - either we do it as a flat society and let the work be for anyone and get pilfered... or set it up to profit in a Capitalist kind of way.

[profdc9]

I originally wanted to use edy555's NanoVNA as a model for my VNA, but he has an ambiguous license and did not response to inquiries about his device.   There were also aspects of his design that make no sense to me.  Therefore I felt that I should not use his design or software.  Instead, I started with the EU1KY analyzer design.  But I have gradually worked it into my own version.  I have tried to keep track of the provenance of the various components that are incorporated into the software I wrote.

As for HP and Tektronix, I have used at the lab $150,000 HP VNAs that have perfectly good hardware but no software updates and therefore succumb to planned obsolescence.  I have much other otherwise good hardware with obsolete software.  Open source hardware would be nice because it might avoid ending up in landfills prematurely.

[OwO]

My opinion is simply that ideas do not deserve protection, but implementations do. If I had it my way, patents would be highly limited (5 years maximum term and higher standard for novelty), but functional designs including PCB layouts would be copyrightable.

Protecting your design by simple secrecy is fine in my books, keeping a secret is a basic right. However PCB copying is trivial in China, almost every stall in the shenzhen market advertise PCB cloning services, so releasing schematics but not layout files is futile IMO.

I will go with a compromise in my future contract projects, which is closed source hardware for 1-2 years, and everything released to public domain afterwards. During the protected period you can use techniques like epoxy potting to protect your layout design.

[KE5FX]

I am using a half-Gilbert cell or single-balanced mixer.  I may go to a full-balanced mixer but the distributed parasitics of a larger circuit may obviate the benefit of that.  It is part of a bigger project called the RFUtilityKnifeBFP740:

Interesting.  What's the benefit to designing your own mixer from discrete transistors?  Gilbert-cell RF/microwave mixers are kind of a solved problem, aren't they? 

[OwO]

Mixers above 6GHz are hard to get and expensive. After some looking around it seems using a single transistor as a switching mixer was pretty common in satellite LNBs, e.g. NE4210S01 (JFET). I'll be doing some experiments with that.

[profdc9]

This is true, but I am trying to learn how to do it, and secondly I think that the mixer should work into the X-band which is a little better than most of the low-cost MMIC mixers, and at a fraction of the price.  For example, here's an example of the circuit (in simulation) with a BFP740 model mixing 10.0 and 10.4 GHz.  Of course the parasitics are going to matter a lot but the circuit is designed to have a minimal layout on a single layer above a ground plane.  I am trying to take advantage of JLCPCB's placement of tiny 0402 components to get everything very small.

I am using a half-Gilbert cell or single-balanced mixer.  I may go to a full-balanced mixer but the distributed parasitics of a larger circuit may obviate the benefit of that.  It is part of a bigger project called the RFUtilityKnifeBFP740:

Interesting.  What's the benefit to designing your own mixer from discrete transistors?  Gilbert-cell RF/microwave mixers are kind of a solved problem, aren't they?

[EEVblog]

I'm just looking at the politics of open-source hardware and the resulting clash. It purports the designs have to be open, source files available to everyone, for the greater benefit of society. No individual alone is allowed to profit. It's very noble to make a design and give away all the IP and files.

This is such a problem that I came up with a Creative Commons like OSHW logo variant that allowed creators to show exactly what parts of the design they are licensing.
Was quite a popular idea, and many have started to use it, but it resulted in nothing but crickets from the OSHW association.  I think they regard me as some sort of non-pure OSHW heretic 

[EEVblog]

But you are advocating a monopoly- select manufacturers that can build a widget and make profit.
So the IP has to be protected somehow, otherwise the vultures swoop in and rip it off... which is what we're seeing happen with OSHW projects. Designers are upset because their sweat and toil gets bastardized or others profiteering from their work. So projects are limited to a flash in the pan happening, one-man band kind of complexity.

Designers making money in the OSHW space isn't rocket science, you just need to do several things:

1) Have the design ready for sale at a reasonable price when you release it. It usually takes time to copy a design and sell it. Of course if it's mainstream profitable enough then the mas market cloners will come in quickly. But with niche items, it's not hard to protect yourself through reputation. Once everyone know to get it from your site then that's what gets linked everywhere.

2) Keep ahead of the competition for new designs and become the "go to" original source. A lot of people like to buy from the original source, so make it clear that doing so is supporting the original creator. i.e make it personal, put a face behind the project.

3) Protect with Trademarks on names, and enforce that. "NanoVNA" is (was) a perfectly trademarkable name. Let the cloners sell your OSHW product, but if they try and use your Trademarked name then send them cease and desist notice. BTW, you don't have the actually apply for and pay for a Trademark for it to be enforceable. If you mark all names with TM from day one then you have commercial use rights (I forget the exact legal terms here).
Of course this works less with generic no-name Aliexpress cloners, but works fine with all the major players in the industry like Adafruit, Sparkfun, Seeed et.al.

4) Advertise widely that there are clones and that you don't make any money from it and they are using your Trademarked name illegally.


For example, I have been selling my uCurrent for over 12 years now, and it's been OSWH from day one. I provide all the original design files so you can make your own. But I make sure to send a cease and desist to anyone that tries to use my uCurrent name on clones. And those people have always been happy to comply. I've even let them sell the remainder of their stock with the trademarked name on it for no commission. I'm still the go-to source 12 years later.

Also, if a design uses specific single source high cost parts, then there is not as much margin to under cut you than there is on say a generic Arduino clone.

[ionberkley]

Dave already nailed many of the arguments that show OSHW is a very viable day job.

But to add a few more:

1) If your "idea" is to make a low spec version of something that already exists at a fraction of the legacy price then you are already a cloner to some degree and the barrier to entry to those who excel at this to come take your spin on the product space and drive the cost lower is low to non existent. If your idea is something that is significantly new, original and sophisticated, then you are immediately differentiated and the barrier to someone else understanding the subtlety underlying your product is still high, even with the source materials in the public space. Frankly it's no different in the closed source world, if your new idea doesn't pass the high barrier to entry test for competition then think hard before spending more time on it.

2) You are not putting dinner on the table with a new $9.99 widget, so if that's your thing, be real with yourself that you *may* cover your costs but likely not, and you are doing this to make the world a nicer place in some way. If at all possible, you want your new OSHW "Thing" to be something whos available market involves significant commercial purchasers...it's a different price point, and crucially when you make something that some company buys N of, instead of Joe the hacker and his mates buying one each, your support burden per unit is dramatically lower.

3) Just because you open sourced it all, doesn't mean you can't also sell the same IP under other closed source terms to those who wish to embed/modify/rebrand it in some way to fortify there deep pocketed but stagnant brand. The initial open source viral spread is your low budget marketing outreach.

4) There is an almost limitless opportunity to sell contracting services to people wanting to use your widget, with some tweak so that it works with there own product line/idea. If your widget is sophisticated and complex don't assume that the rest of the world can run with it *and* understand it just because they have the source files.

5) You don't have to add much in the way of a black box to an "Open Source" design to make it unattractive to the cloners, and it can often be a piece of the system that has little to no value being Open Source anyhow.

Personally I have worked on a number of OSHW designs that own their market against all comers indefinitely, even in the presence of cloners, and my inbox constantly overfloweth with people who want me to help make Widget X do something new for them and are happy to exchange plentiful cash for that.

[ebastler]

He wants it to be OSSW/OSHW but lacks the capital and no one seems to have  a clear idea of how to suppress piracy

I don't understand. What would be the point of making it open-source, if at the same time you want to avoid "piracy"?

Is the idea to publish the design files and source code, but with a license that does not allow others to use them for their own purposes? That would simply mean that you are aiming for free community support for your own commercial, proprietary product. That's not "open source" in my book, but rather "nice try!". 

[OwO]

"NanoVNA" is (was) a perfectly trademarkable name.

USPTO thinks otherwise (good thing it was refused registration, because it wasn't edy555 applying for it):
https://tsdr.uspto.gov/documentviewer?caseId=sn88701756&docId=OOA20200226121814#docIndex=1&page=1

SECTION 2(e)(1) REFUSAL - MERELY DESCRIPTIVE

Registration is refused because the applied-for mark merely describes a feature, characteristic and function of applicant’s goods.  Trademark Act Section 2(e)(1), 15 U.S.C. §1052(e)(1); see TMEP §§1209.01(b), 1209.03 et seq.
...

Applicant’s applied-for mark is NANOVNA for “Electronic devices for measuring electric current; Instruments for measuring length; Optical inspection apparatus; Precision balances; Protractors; Radiation gauges used for measuring the physical properties of materials; Surveying instruments; Surveying machines and instruments; Surveyors' levels; Teaching robots.”

The attached entry from Merriam Webster’s Dictionary defines NANO as “one billionth part of something” and the attached entry from Acronym Finder describes VNA as “Vector Network Analyzer.” The attached webpage from www.tek.com (Vector Network Analyzers are used to test component specifications and verify design simulations to make sure systems and their components work properly together) explains the beneficial nature and usage of vector network analyzers in the technology industry. Descriptiveness is further evidenced by the attached webpages from http://nanovna.com/?page_id=21 (NanoVNA is very tiny handheld Vector Network Analyzer (VNA)), www.rtl-sdr.com (Reviews of the NanoVNA: An Ultra Low Cost $50 Vector Network Analyzer), and https://hackaday.com/ (nanovna is a $50 vector network analyzer) which shows that NANOVNA describes a small vector network analyzer. Moreover, Applicant has stated for the record that the wording NANOVNA has no significant meaning in the relevant industry; however, the above attached webpages shows that this wording has relevant descriptive meaning in the applicable industry. Thus, this wording is merely descriptive of the characteristic and feature of Applicant’s electronic products.

If you don't register your trademark the consequences vary by country. In the US you gain trademark rights by using the mark when doing commerce, and unregistered trademarks usually still get some protection if it's actually unique. In China it's mostly first-to-file. Someone who tries to register your trademark before you do may even get it, but typically they can't stop you continuing to use the mark if you can prove you were using it in commerce before the other applicant filed for a trademark. That is exactly what happened to many Japanese brands here, copycats registered first, and the original brand get no more protection although they can continue to use the brand.

[EEVblog]

If you don't register your trademark the consequences vary by country.

Yes, of course, always consult a local Trademark attorney if you are serious.
But often just having the TM there and cease and desist notice is enough to work.

[EEVblog]

He wants it to be OSSW/OSHW but lacks the capital and no one seems to have  a clear idea of how to suppress piracy

I don't understand. What would be the point of making it open-source, if at the same time you want to avoid "piracy"?

I explain here:

Is the idea to publish the design files and source code, but with a license that does not allow others to use them for their own purposes? That would simply mean that you are aiming for free community support for your own commercial, proprietary product. That's not "open source" in my book, but rather "nice try!". 

No, you could be doing it for genuine intent for learning, repair, modification etc.
This is why I created my OSHW logo idea. For example, if you just want to release the schematic, you can just release the schematic, and that's OK (actually a good thing).

[floobydust]

But you are advocating a monopoly- select manufacturers that can build a widget and make profit.
So the IP has to be protected somehow, otherwise the vultures swoop in and rip it off... which is what we're seeing happen with OSHW projects. Designers are upset because their sweat and toil gets bastardized or others profiteering from their work. So projects are limited to a flash in the pan happening, one-man band kind of complexity.

Designers making money in the OSHW space isn't rocket science [...]

You've left me a bit confused - the 121GW started out as open-source, but not where it is today. You closed the design I believe. What's going on, other side of the fence?

[EEVblog]

You've left me a bit confused - the 121GW started out as open-source, but not where it is today. You closed the design I believe. What's going on, other side of the fence?

No I did not. I never said nor promised it would be open source. I said it would be as open as possible, and that I would see what the manufacturer wanted to do. They decided only to release the schematic. Nothing I can do about, I do not own the rights.
The app software that I paid for and own the rights to is fully open source.

[floobydust]

I read the Kickstarter text and thought that was how the project started out "It's fully open source though, so you can take it and run with it." oops it's the app section
So I thought open vs closed changed and wondered why.

[nuclearcat]

First sorry for my bad english, i'm sure it's hard to read in long post.

IMO for opensource you need to pick model you like. I am writing on experience of several friends and customers, who succeed and failed on this matters.
I usually helped them with the crypto part.
Also my examples are somehow limited "use case", they wont work well for devices without "data" interface, like uCurrent.

First and main question - do you plan to make profit from it?
If NO or you don't mind if someone will make better product - fire and forget. Thats simplest case, and there is a lot of those. Open donations if you like, if you want to make support of product more self-sustaining, so it can buy you some beer. Be ready, that at any moment someone make better version.
And thats totally fine, if you want to share something cool with world.

If YES, then it's getting more interesting.

First rule in all of this:
Don't release up to date recipe how to make competing, successful product, especially if your product have high complexity+value, and you don't have new version/revision prepared,
 that have significantly better features and most of users will desperately want to upgrade to this new release.
Remember and keep in mind, if you plan to make opensource, your main priority to make life of users, and especially contributors easier, not life of unfair competitors.
Another thing, if someone enough big decide to invest in copying your product, without agreeing with you, you need to have all ready to hit him hard,
at moment when he start mass sales. He will have bunch of obsolete devices stock that he can't sell and he should be hurt on marketplaces he uses.

Story 1
Once manufacturer of wireless links based on ath5k chipset released build that was allowing to run some cheap D-Link Wireless AP as his product, as response to a trolls,
they required proof that he really worked hard to firmware and not just copied it from vendor.
This is how he ended his sales. People just kept flashing D-Links and most of them didn't wanted to buy his products, just because of a bit better reliability.
And it was just binary release, single firmware file, that killed everything.

Story 2, not really mine, so not sure if i understand all facts properly
You have to keep an eye also that where is most of your property kept, like ELM327 most of value was in chip firmware. They are not opensource, but
most of their hardwork was in decoding all those countless car protocols, was kept in firmware. PCB was trivial to copy, so everything that was waiting
when cloners will be able to extract firmware, and by not protecting copy they just gave them a gift, almost same way as in story 1.
And new products was not that much significantly better.
https://www.elmelectronics.com/products/ics/obd/#ELM327
Low power, settings retained, buffer size, its not very convincing.
Worst part, they missed moment and market of apps built around _CLONES_, and not original hardware.
Probably

Rule 2
Make your product such way, that users can quickly identify effing clone and legitimately demand their money back.
Primary trick - even it is fully opensource, nobody force you to release private keys that can confirm product genuinity.
First of all register some quick to remember domain that resemble your product name. Make some community-friendly features around it, forum, where people can talk, support system etc. Don't invest much,
but keep it ready, just in case.
Use crypto, components unique identification parts to build defense, but dont expose it early.

Several scenarios:

There is many ways how you can protect your device cryptographically. It's all depends how your manufacturing is done, i wont list it here as it will make post huge.
Then, as example - during update or first supplementary software run - show that this product is COUNTERFEIT.
Important it must hit at right moment, most likely directly after purchase!
Perfect if you can enable this "surprise" for moment when "cloners" will run mass sales.
They will receive wonderful chargebacks, might be kicked out of marketplaces like amazon, aliexpress and others. Hurt them as much as you can.

If you discovered presence of clones too late, also you can also add routine to verify hardware/serials in new releases,
and just show already existing users (dont hurt innocent victims of counterfeit products!) that their hardware is not original
with suggestion where to send info, such as where they bought it, for sake of naming and shaming or takedown requests.
You can send them what particular flaws their hardware/software have.
You can even make some profit from it, and if copied design is OK - release custom update for it, for small fee that will make device legit and workaround some bugs introduced by "cloner".
(add crypto signature, so this update wont be copied!)

Rule 3
Make sure, if someone try to reverse engineer your product, he will spot warning, that surprises are waiting for him and he gonna lose money, and better
if he make deal with you and sell legit product.
And make sure your gun loaded with multiple bullets. You might need to make warning shots too.

There is another model, that you make useable opensource product, and optional "closed" commercial features based on it. It is not truly opensource, however, it is very successful too.
It is very reasonable if you product can be used for hobbyists, where they can spend more time using free version, but paying much less or even making their own, and those who want to
use it for business, can afford to pay for "commercial" one.
But this model also has its own nuances too.

P.S. GPL. Don't idealize it. IMO it's not very enforceable, and really works only if there are large customers in your market who are concerned about licensing cleanliness. Others just dont care.

[EEVblog]

Don't release up to date recipe how to make competing, successful product, especially if your product have high complexity+value, and you don't have new version/revision prepared,
 that have significantly better features and most of users will desperately want to upgrade to this new release.

That's called the Osborne Effect
https://en.wikipedia.org/wiki/Osborne_effect

Rule 2
Make your product such way, that users can quickly identify effing clone and legitimately demand their money back.

Easiest way to do that is to put your name, your company logo, your Trademark name etc on it. If it has any of those things and it's not from an authorised reseller then it's an illegal clone.

[nuclearcat]

Easiest way to do that is to put your name, your company logo, your Trademark name etc on it. If it has any of those things and it's not from an authorised reseller then it's an illegal clone.

Partially agree. Will work for most, but not for all.
Many people is just too lazy to check, who is authorized and who is not.
Even more want to buy product on comfortable for them marketplace, amazon, aliexpress, ebay. Authorized resellers often too hard to deal with in anywhere else than "First world".
And most importantly, "cloners" may not dare to copy the brand/product of a famous brand/blogger, but they do not care about the reputation of a small developer who has much less leverage to force them to stop selling.