Making an embedded project distributable

[dkonigs]

So there's an electronics project I've been working on for several months now. Its reached the point where I have several fully functional prototypes, and it honestly looks kinda like a finished product. (Mostly because I designed a nice enclosure for it, even if there is still room for improvement.)

The project is entirely open-source hardware and software, but I do not expect the many potential users to be willing or able to build it from the component level themselves. (Firmware tinkering is more likely, but still probably a minority of users.) It is also sufficiently niche, that I do not expect some random company to swoop in and start building it for their own profit.

I'm at the point where I really want to explore what is necessary to move forward with making this something I can actually distribute to other people (US, Europe, and elsewhere), both from a legal perspective and from a personal liability perspective. There are a couple of stages of this "distribution" I'm likely to want to go through, and its possible that each one may have a different level of issues/concerns:

• Limited number of prototypes distributed to friends/acquaintances for evaluation and feedback. (Maybe a dozen units, not sure.)
• Low rate of production, but technically selling the thing. (Dozens of units, still probably under a hundred.)
• Moderate rate of production, probably considered a real "product" at this point (but still relatively low volume)

(There's also the option of distributing as a partially-assembled "kit", but I'm not sure if that would actually help with any of the concerns discussed here.)

I'm currently just a private individual, though I'm fully aware that I may need to "incorporate" at some point in the process to have an entity through which to do everything. But I'm obviously not established in any sense.

Its entirely likely that the correct answer to many of my questions will involve "hiring a consultant" to walk me through the process and help me make critical decisions. I am willing to do that. That being said, I have absolutely no idea how to find such a person. Any actionable advice as to how I do that would certainly be helpful.

Now, on to the project itself... Its a fancy enlarger controller for the (resurging) area of analog darkroom photography. But I'm going to stick to what aspects of it are relevant to people here. In the most basic form, think of it as an embedded device that switches the power to a pair of mains-plug light sources. I've made a little block diagram that should hopefully illustrate:


I know "mains power" tends to raise all sorts of red flags when it comes to this whole process. However, being able to reliably switch those mains-powered devices is kinda important to making this device useful. And while all the components that touch the mains-side of the PCB have all the necessary certifications (AC-DC converter, relays, switch, etc) I also understand that this probably doesn't get me quite as far as one would hope.

Now I've started to explore my options for "outsourcing" the mains side of the system from my responsibility. Unfortunately, there aren't really any good off-the-shelf options that tick all the boxes for me here. And many of them actually make the finished product more dangerous for the user to properly hook up, even if I'm potentially less liable for that danger.

One idea that I've come across, is replacing the switched mains outlets with a DMX512 interface. That would theoretically offer me a lot of future flexibility in terms of what I can control. But unfortunately, no one actually makes a cheap "1-2 relay switched outlets /w DMX control" device. Sure you can cheaply build one, with parts bought off Amazon or AliExpress, but you're back to expecting the user to do their own mains electrical wiring to use your device. (And while I could provide a reference design for this, I still wonder if doing that opens me up to any liability.)

In any case, I think this is enough to get the conversation started. I'll gladly provide additional information, if it would be useful to the discussion.

For what its worth, I am actually willing to spend money to get through this process. I'm also more interested in making this project available to the relevant community, than I am in making a profit myself from it.

I'm mostly looking for actionable advice here, even if that advice involves recommendations for others I should contact/hire to help with this.

[ebastler]

To work around the "switching mains" challenge, couldn't you incorporate a transmitter to control a couple of wireless power switches -- you know, those plug adapters controlled via Zigbee or whatever the current protocol of choice is?

To sell in Europe, you would still need to deal with other major compliance headaches: CE declaration of conformity (which includes meeting EMC and RoHS requirements, and documenting that you do), and electronic waste collection (registering with the WEEE scheme). The latter responsibility could be transferred to a Europe-based distributor. CE compliance is formally also the responsibility of the European importer, but you will probably not find a serious reseller unless you can provide documentation to show compliance.

If you are thick-skinned you could sell to Europe directly from the US, which would put the import and compliance risk on the end customer: If customs picks up on the fact that the device is not CE compliant, they can seize the package, and you would be faced with an unhappy customer. If you expect to sell in small numbers, refunding the occasional customer might still be a more attractive option than paying for the overhead of CE compliance?

[dkonigs]

To work around the "switching mains" challenge, couldn't you incorporate a transmitter to control a couple of wireless power switches -- you know, those plug adapters controlled via Zigbee or whatever the current protocol of choice is?

Wireless home automation tech is really a non-starter for two reasons. First, I really do need reliable and consistent timing. Second, they're all closed proprietary "standards" where there's a fair bit of overhead to actually implementing them. (And they're generally not designed for this sort of use case.)
That's why I got so excited when I discovered DMX512. On paper, it looks like the perfect sort of interface standard for a project like this. Unfortunately, it seems to be lacking in "off the shelf products" on the lamp/device side that meet my needs.
(Again, not hard to build something from off-the-shelf parts, or to buy something expensive/overkill that could work, but neither of those is desirable.)

Of course there are future product goals where I still may want to consider a DMX512 output port regardless (e.g. for operating RGB LED controllers), but that doesn't really negate the need for also having simple switched mains outlets.

To sell in Europe, you would still need to deal with other major compliance headaches: CE declaration of conformity (which includes meeting EMC and RoHS requirements, and documenting that you do)...

Yeah, I know I can find plenty of threads around here where people will happily detail all the compliance headaches at the acronym level. What I'm really looking for is a way to get useful advice on how to actually navigate this landscape. Both in terms of what I need at the early stages (e.g. sending a dozen prototypes around the world) and later on. I think what I'm most desperate for is someone to simply have private conversations with on the subject, to get concrete actionable advice and guidance from. In theory, that means hiring some sort of consultant. But I still need to figure out how to find that consultant... Especially since I want someone who is familiar with my general situation, and will thus give advice appropriate to my scale.

[bjbb]

Why is this an 'embedded' project? What is 'distributable'?

Per NIST and OSHA, the ISM category also covers consumer equipment that does technical stuff. The NIST publishes a readable (and a bit over-simplified) guide (NISTIR 8118). For EMC and safety, this would probably fall under the regulations and standards that are scoped for ISM equipment.

The real reason that you do not want to do wireless control of power is to avoid being scope by Functional Safety requirements. There are several ANSI/UL/CSA standards that specifically scope this type of equipment. EMC for ISM stuff is in 47CFR part 18, but am not certain that you could claim Class A for stuff intended for consumer at home, so conducted emissions could be problematic.

Product safety for design of stuff that is connected to AC mains is not that difficult. But reading and understanding the construction and material requirements, and test methods, and the physics of the scoped safety standard is a leap unto the Dark Side. My only advice is to keep to control/secondary side to Class 2 voltage and power limits, and use a relay that is certified by an NRTL per the unit's label ratings, and use a PCB house having a UL mark and that is rated DSR.

There are a bazillion other small things to consider for the layout, components, recurring testing, user manual, labeling, etc.

Edited. oops.