Bring a product to market without certification?

[Kerlin]

For sure some of what fcb says above is true.

If there was a prosecution it would be against a local company.
I have been told the rules are mostly only enforced on local companies.

I know in many countries "they over there have" an official Favored Trading Nation status. Why?

I have always had a feeling feel there is some thing very sus about the whole compliance thing, remember I have worked over there and seen what goes on.
In fact when I worked there I was trained in how to go about it.

[penfold]

[...]
I have been told the rules are mostly only enforced on local companies.

I know in many countries "they over there have" an official Favored Trading Nation status. Why?
[...]

I was under the impression (at least from a UK perspective, generally, though some more special arrangements exist) that it is the responsibility of the retailer, the entity that puts the device in th market place, that assumes responsibiliy for the certification of a product, though could just as eaasily be the importer, talking mostly about small producers and resellers since larger corporate entities can have slightly more negotiated and delegated responsibilities - also ignoring medical and functional safety related products. If a product produced by company "X" is purchased by company "Y" and resold to company "Z" before being "marketed" to consumer "U" - the responsibility of ensuriing compliance lies with the reseller "Z": which formally is because the certification relates to the advertised useage of a product rather than its original designed function. In real-life, reseller Z may (informally) delegate the respopnsibility of compliance Y, who delegates to X - if a legal dispute arrises the first priority of conviction would be anybody making an unsubsantiated claim about the product.

The word "practicable" could very easily feature as can "risk"; a low-volume and low-price product doesn't necesarily warrant a full testing and certification process, but the manufacturer/designer would be well advised to document all the reasons and motivations for not formally-testing along with all the mittigations and design choices. What fcb said is still perfectly valid, but even an insurer would want to see that one wasn't being reckless or negligant.

[fcb]

[...]
I have been told the rules are mostly only enforced on local companies.

I know in many countries "they over there have" an official Favored Trading Nation status. Why?
[...]

I was under the impression (at least from a UK perspective, generally, though some more special arrangements exist) that it is the responsibility of the retailer, the entity that puts the device in th market place, that assumes responsibiliy for the certification of a product, though could just as eaasily be the importer, talking mostly about small producers and resellers since larger corporate entities can have slightly more negotiated and delegated responsibilities - also ignoring medical and functional safety related products. If a product produced by company "X" is purchased by company "Y" and resold to company "Z" before being "marketed" to consumer "U" - the responsibility of ensuriing compliance lies with the reseller "Z": which formally is because the certification relates to the advertised useage of a product rather than its original designed function. In real-life, reseller Z may (informally) delegate the respopnsibility of compliance Y, who delegates to X - if a legal dispute arrises the first priority of conviction would be anybody making an unsubsantiated claim about the product.

The word "practicable" could very easily feature as can "risk"; a low-volume and low-price product doesn't necesarily warrant a full testing and certification process, but the manufacturer/designer would be well advised to document all the reasons and motivations for not formally-testing along with all the mittigations and design choices. What fcb said is still perfectly valid, but even an insurer would want to see that one wasn't being reckless or negligant.

EMC/CE/etc..  It's all a bit like the parking at my business unit:
In theory we have clearly defined (in the lease) parking spaces.  The reality is that everyone parks in the most convienent space and we all just get along nicely & quietly - someone once suggested marking out the spaces, that idea was very quickly gelded - things just seem to work well as they are.

I think a good part of the reason things aren't enforced is that no one really complains about EMC.  FHSS, digital TV, DAB and streaming radio all seem to still work well enough so people don't correlate low-bit rate events with RF noise anymore - I certainly remember the days when next door vacuuming would show on the TV as black speckles, or you could hear a motorbikes ignition on the radio - these things may well be generating the same RF crud as before, you just wouldn't notice it now.

And since the legislation was introduced in Jan 1996 I have NEVER been asked for a copy of a products TCF.

[penfold]

[...]
EMC/CE/etc..  It's all a bit like the parking at my business unit:
In theory we have clearly defined (in the lease) parking spaces.  The reality is that everyone parks in the most convienent space and we all just get along nicely & quietly - someone once suggested marking out the spaces, that idea was very quickly gelded - things just seem to work well as they are.
[...]

I like that analogy, but I wonder what lengths people would go to if, in the unlikely event of some upset, the penalty for having caused the upset was getting their car crushed - though I'd also suspect that one of the reasons to not mark parking spaces would be that it increases council tax rates. From my perception of it, EMC is just one of those nightmare issues thats invisible, often counter intuitive, potentially high penaltied, expensive and time consuming to measure and correct, and very difficult to portray a pragmatic approach to without sounding slap-dashed. Something I forgot from my previous post was that the user manual and any exclusions to the devices useage are also quite important, i.e. making clear whether it is commercial or industrial, for integration within a larger system or not, listing suitable external power sources etc

[Psi]

If you fit a nice TDK PSU you are just as liable if it squirts all over the spectrum as if you fit a NONAME unit - however the reality is no one will really care and you are fairly safe. YMMV.

True, but the TDK PSU is MUCH less likely to squirt all over the spectrum than the $1 aliexpress PSU.

My main recommendation would be setup an LTD company (£13+ongoing accountacy!), get your company some product liability insurance - Hiscox for about £2-300/yr. depending on product.  Remember that regardless of paying an EMC lab or not - you are always self-certifing the product.

Pretty sure in the event of an issue, the liability insurer would refuse cover you if you didn't have all the certs with all the I's dotted correctly. It's insurance, they are always looking for a way to avoid having to pay out.

[fcb]

If you fit a nice TDK PSU you are just as liable if it squirts all over the spectrum as if you fit a NONAME unit - however the reality is no one will really care and you are fairly safe. YMMV.

True, but the TDK PSU is MUCH less likely to squirt all over the spectrum than the $1 aliexpress PSU.

My main recommendation would be setup an LTD company (£13+ongoing accountacy!), get your company some product liability insurance - Hiscox for about £2-300/yr. depending on product.  Remember that regardless of paying an EMC lab or not - you are always self-certifing the product.

Pretty sure in the event of an issue, the liability insurer would refuse cover you if you didn't have all the certs with all the I's dotted correctly. It's insurance, they are always looking for a way to avoid having to pay out.

TDK vs NONAME - You pays your money and makes your choice - ultimately the only way to prove this is to test it. I'd sure trust TDK's documented test results more than a NONAME unit.

As for the insurer covering you - they are covering the liability the product goes wrong and injuries people or property, not that it fails EMC.  Unlikely they are protecting you if by some weird turn of events you get sued for emmisions (unless of-course they are the reason people/persons/property get injured).  Also, it's important to keep pointing out - the EMC lab report is not a certificate - it's an opinion, there is no such thing w.r.t. EMC in the UK/EU as certification, there is in the US (FCC), but not in the UK - you are always self-certifying.

Also, I'm not actually suggesting that manufacturers don't test, but that they understand the role it plays in self-certification.

[fcb]

[...]
EMC/CE/etc..  It's all a bit like the parking at my business unit:
In theory we have clearly defined (in the lease) parking spaces.  The reality is that everyone parks in the most convienent space and we all just get along nicely & quietly - someone once suggested marking out the spaces, that idea was very quickly gelded - things just seem to work well as they are.
[...]

I like that analogy, but I wonder what lengths people would go to if, in the unlikely event of some upset, the penalty for having caused the upset was getting their car crushed - though I'd also suspect that one of the reasons to not mark parking spaces would be that it increases council tax rates. From my perception of it, EMC is just one of those nightmare issues thats invisible, often counter intuitive, potentially high penaltied, expensive and time consuming to measure and correct, and very difficult to portray a pragmatic approach to without sounding slap-dashed. Something I forgot from my previous post was that the user manual and any exclusions to the devices useage are also quite important, i.e. making clear whether it is commercial or industrial, for integration within a larger system or not, listing suitable external power sources etc

Agree, difficult to not sound slap-dashed.  Probably also worth pointing out that I do test products I design and do advocate testing at a pucker lab (we only do precompliance - own EMC chamber, spec-ans analysers, LISN's, EMC receivers, antennas and a clean field site with tripod mounted York antenna, PMM receiver etc...).

Defining in the manual who uses the products is a good idea.

[mikeselectricstuff]

the EMC lab report is not a certificate - it's an opinion, there is no such thing w.r.t. EMC in the UK/EU as certification, there is in the US (FCC), but not in the UK - you are always self-certifying.

An EMC lab once told me that the only difference between a "quick-look, try what you can, tweak and repeat" day  and a "formal" test is the lab's level of liability

[coppice]

the EMC lab report is not a certificate - it's an opinion, there is no such thing w.r.t. EMC in the UK/EU as certification, there is in the US (FCC), but not in the UK - you are always self-certifying.

The lab report is a little more than an opinion, but not much. The lab can only attest that one sample of the product on one day passed the compliance tests. Its the maker that is responsible for ensuring a production run consistently passes. Considering how much minor variations, like the torque on certain screws, can affect the outcome of a test, the lab's liabilities should be limited to ensuring they actually performed the tests according to the approval spec.

[SiliconWizard]

You're not always "self-certifying". There are a number of regulated domains, such as medical devices, for which you can't self-certify. A notified body certifies your product. But you must provide all relevant technical documentation and test results. The notified body emits the certificate.

A properly accredited lab can provide test reports that conclude on compliance with some given standards. Of course, it's not up to the lab to then conclude if the whole product is conforming to any standards or directives it's supposed to be compliant with.

For classes of products that can be self-certified, then it's indeed the full responsibility of the company selling the product. You may even not have anything tested in a lab. But since you're claiming that your product is compliant, if you have no proof of the fact it is, then you'll obviously get into trouble if anything goes wrong in the field.

[coppice]

You're not always "self-certifying". There are a number of regulated domains, such as medical devices, for which you can't self-certify. A notified body certifies your product. But you must provide all relevant technical documentation and test results. The notified body emits the certificate.

A properly accredited lab can provide test reports that conclude on compliance with some given standards. Of course, it's not up to the lab to then conclude if the whole product is conforming to any standards or directives it's supposed to be compliant with.

For classes of products that can be self-certified, then it's indeed the full responsibility of the company selling the product. You may even not have anything tested in a lab. But since you're claiming that your product is compliant, if you have no proof of the fact it is, then you'll obviously get into trouble if anything goes wrong in the field.

The lab's certificate merely says one sample of the product worked to spec on one day. You still self certify that that the production run is up to the same standard. Nobody else can. Smart customers sample test production batches. You can't hold the lab liable in any way for samples of a product that don't meet the spec. You can only hold them liable if they didn't perform the tests properly.

[penfold]

the EMC lab report is not a certificate - it's an opinion, there is no such thing w.r.t. EMC in the UK/EU as certification, there is in the US (FCC), but not in the UK - you are always self-certifying.

The lab report is a little more than an opinion, but not much. The lab can only attest that one sample of the product on one day passed the compliance tests. Its the maker that is responsible for ensuring a production run consistently passes. Considering how much minor variations, like the torque on certain screws, can affect the outcome of a test, the lab's liabilities should be limited to ensuring they actually performed the tests according to the approval spec.

That's an interesting point. Firstly with the "little more than an opinion", I guess, as a type-test or design proving or verification exercise, it's only a statement considered universally valid on the day of the test with particular test gear, operator, room, weather, etc for that single (or few) DUT(s) built to whatever spec they were built to... I'm not sure I'd say opinion, but there is a fair amount of tenuousness in saying the results are true for every product coming off the production line. But with the ensuring all products off the production line meet the same spec, I wondered about this previously, because, even with so much weight put on the importance (or as it appears within the electronic engineering community) of good compliance data and test results, etc, my first instinct is product consistency is almost more important. I don't know if that's a common instinct, but it does concern me a little bit that "EMC testing" can start to appear more like a project manager's milestone whilst "making sure the documentation genuinely defines the product" can get pushed back. And yet, from last time I read up on CE/UKCA marking, I couldn't find anything that really settled my inner argument...

But yeah, if a product turned out to be a disaster, gotta be able to work out the extent of the disaster.

[karpouzi]

Also, worry about it when you get enough sales to justify it. Unless your product is actually dangerous enough that you are personally worried.

Are there any people who simply delay full compliance for a time until the product sells? Even if it means a new board revision and label for the next batches. Do regulators really go after small-volume startups?

Compliance testing may be just a "business expense" for corporations, but it squashes small startups who don't have the money, or the knowledge whether their idea will actually sell or not. Even using a "pre-certified" Bluetooth module, your compliance starts at $20K (Bluetooth qualification fees, CE+FCC testing) just for US and EU market access.

[coppice]

Do regulators really go after small-volume startups?

If that small startup annoys one of the big players, its not unknown for them to set the regulator's on the startup as a way of shutting them down - at least for a while.

[Kasper]

Do regulators really go after small-volume startups?

If that small startup annoys one of the big players, its not unknown for them to set the regulator's on the startup as a way of shutting them down - at least for a while.

What does 'shutting them down' mean?  Just can't sell any more until you get appropriate certifications? or fines or jail?

[Kasper]

I made a table with some addressable LED strips in it, now I want to try selling some.  Doubt I'll ever sell more than a few a month.  Will probably only sell in Canada.

I'm hesitant to sell any before sending it to an EMC lab but I'm hesitant to pay for EMC testing before I see if anyone wants to buy them.  I don't have much appetite for risk.  After working for about 20 years, I've amassed enough savings to worry about getting sued and losing it.

It's been a while but last time I talked to an EMC lab, they wanted the final version of the enclosure and said any significant changes, particularly to metal components like screws should get retested.  Sounded more like a recommendation than a requirement but I can't really remember, it was just a conversation years ago.

Each piece of furniture will be hand made from live-edge wood, different shapes and sizes, different location of metal table legs and fasteners.  Different lengths of LED strips.  LED strips adhered to a metal plate that will be different size in each unit.

Should I send each one for EMC testing or can I just get the electronics certified and ignore their 'enclosure'?

If I just certify the electronics, would it be reasonable to go with the most LEDs I intend to use and expect some units will have less LEDs?

If first unit is far enough below the limits, would that justify not sending the others?

Should I sell a few to see if anyone wants them before I do EMC testing?

Are there any other regulations for me to satisfy as well?

I have a sole proprietor business right now focused on electronics contracting.  Should I make a separate business, incorporate and get insurance before I sell any?

What's the worst that can happen?

What would you do?

Power supply: USB-C (5V, dumb) with removable off the shelf USB battery.
Regulator: 3.3V linear
MCU: RP2040
LED strips: 3 wire addressable (WS2812B).  Up to 8 strips in parallel.  Each about 1m to 2m long.
LED wires: data from RP2040 goes through level shifter then 470ohm resistors then a foot or 2 of wire to LEDs strips.
UX: trimpots and power switch
Intentional radiators: none

[mikeselectricstuff]

After working for about 20 years, I've amassed enough savings to worry about getting sued and losing it.

That's what limited companies ( or equivalent in your country) are for.

What would you do?

Design it in such a way as to be reasonably confident that it would pass, but not waste money on testing.

In jurisdictions like US  and Canada I'd probably be more concerned about things like fire safety (overcurrent protection etc.).
 

[BrianHG]

What would you do?

Make sure you use an 'enclosed' fused power supply which has authentic UL/CA approvals.  (No Aliexpress/Amazon Chinese garbage clone supplies, use authentic name brand stuff only.)
This will get you around everything except FCC.
For US FCC, just make sure you don't have any intentional radiators and when driving the LED strip's data lines, use the proper series resistor at the IO pins of your MCU so that you aren't driving a harsh super fast square wave into the wiring feeding the LED strip.  (I think you said you have a 470 ohm series resistor already...)

Follow these rules and you shouldn't have any problems in Canada, or the US.


Fail on the UL/CA approvals for the power supply, then if your product creates a fire somewhere, the insurance company who pays out the damage and finds out during an investigation may come after you for some $$$.

[Kasper]

Thanks for your responses.  Here is what I'm thinking of for the power supply:

Battery:
https://www.amazon.ca/Anker-Portable-Charger-High-Speed-Charging/dp/B0BYP4Y1N8

Wall plug:
https://www.amazon.ca/Charger-Anker-Compact-Foldable-Included/dp/B0B2MMYZPL

User can leave the battery out and plug USB-C from wall plug into my device or they can plug battery into my device and charge it as needed.

Does Anker count as cheap garbage?  They rank well on lots of review sites and I've used them at home without problems.

[BrianHG]

Thanks for your responses.  Here is what I'm thinking of for the power supply:

Battery:
https://www.amazon.ca/Anker-Portable-Charger-High-Speed-Charging/dp/B0BYP4Y1N8

Wall plug:
https://www.amazon.ca/Charger-Anker-Compact-Foldable-Included/dp/B0B2MMYZPL

User can leave the battery out and plug USB-C from wall plug into my device or they can plug battery into my device and charge it as needed.

Does Anker count as cheap garbage?  They rank well on lots of review sites and I've used them at home without problems.

Anker appear to be a modern compact Gan wall plug.
However, what's important are the 'approvals' like UL, CE, CA listings.
Their approvals arent shown on the Amazon website.

[Kasper]

Anker appear to be a modern compact Gan wall plug.
However, what's important are the 'approvals' like UL, CE, CA listings.
Their approvals arent shown on the Amazon website.

It seems approvals aren't often shown online.  Couldn't even find them for Apple or Samsung, though I didn't look too hard.  The ones in my home do have a variety of stamps at least.