EMC compliance question

Where is there such an exemption?

It's under the section on sub-assemblies:

A "CPU board" as defined previously is considered a type of sub-assembly.

As a hypothetical say someone sells a product that's a PCB with a PIC on it and some relays that has e.g. RS-232 port and terminal blocks to connect to other stuff. It's in some sense a functional product, but the user must at least connect it to some load of their choosing for it to actually do anything. And they can put it in a housing if they wish, or not, whatever.

Is this still a "sub-assembly"?

Do you really care?

My Brit friends say that CE means Can't Enforce.

It's been a long time since I saw an FCC truck bristling with antennas snooping around.

A sub-assembly is employed inside a listed device - the final device requiring compliance.

A subassembly is not a stand-alone product and cannot/neednot demonstrate compliance or even full functionality on its own, as the housing,interconnection and operational environment is not present.

If it functions stand-alone, it CAN be tested for compliance, is self contained and has well-defined IO sources/targets.

Products may contain parts that are self-compliant and list them as subassemblies, but the final product inherits and is stuck with that subassembly's red tape.

RL

One of those: "I'll tell the client 'if it sells then pay to get it tested and if it don't sell then its compliance is kind of a moot point'" kind of questions..

Or Chinese Export.

They may say that, but I bet they comply - it means if your competitor finds out that you don't comply you have problems. I've never worked anywhere where it's not taken seriously.

Same here. I've nursed many products through the EMC and safety compliance processes over the years. There have been few changes to EMC, but the safety regulations for audio and IT products and for test equipment have seen some radical changes - mostly for the better.

John

I've gotten by on the sub-assembly exemption for some things I've sold myself over the past couple years but I'm stating to do designs that probably can't get by on that exemption no mo, le sigh.

My thinking is that I should at least do some fashion of pre-compliance testing and then if the design sells go all in.

And if it don't sell then compliance is somewhat of a moot point, anyway.

In europe, "taking seriously" often means buying a reel of CE stickers and slapping them on everything.

Did you EMI lab test everything that you sold?

We mostly sell things that go onto bigger systems, and our customer worries about certifications. So we only need to not make them fail. Our box could be buried deep inside their system.

We only once were blamed for an important EMI failure, and that turned out to be our customers fault.

There must be ballpark a million shops in the USA that assemble PCs from parts, mostley Chinese parts, the other CE kind, and I doubt if

Yeah, riiight.

Every individual unit? No, neither is it necessary. Generally, you design a product with EMC in mind, and an independent lab/consultant tests and certifies. Iterate if needed.

I have never seen any developers in Europe sell products that are not tested and compliant.

I have seen many Chinese products that fails testing, but also seen Chinese products that comply with very ingenious solutions.

Right, that's the differentiation. You do not need to test products that go into a system, just as long as the complete unit complies.

But, if it fails, it can be hard to find out which one of the sub assemblies are the root cause, and time is money

A few pennies for a certified test lab to do full certs?

A wart relieves one of all the AC-line safety certifications. There are some big warts these days, including 48v ones.

One can resell a cheap wart with the usual molded-in (usually fake) UN/CE/CSA markings, or let the customer buy their own wart.

Software certs on top of hardware certs?

For a small company making a modest number of some test instrument, full certs will multiply development cost. That may be why I don't see a lot of small instrument companies in europe.

The guys I was working with in Oxford laughed at me when I asked if our atom probe system would need to be CE tested. "CE means Cant Enforce."

When the regulations change, do you have to recertify?

Just a little mindless optimism there.

I did get on product started on CE testing, and the mains connector was the weak point. I think the final solution was to buy a better one.

If your product can power usage is larger than 15W, then you get close to nothing by using external SELV supply, because then a lot of the demands on safety are back in play

Some just takes the risks. If you are caught it can be an expensive risk. On the other hand, I have never heard of a case where the company went bankrupt. Have heard of large fines, but nothing that killed the company

What's crazy is how expensive the CE specs are. I can buy one PDF for $600, and it will reference a bunch of others. Recursively.

These specs have the force of law. It's like being forced to pay to know what's legal or not.

With the vast accretion of laws and regulations we’ve seen in the past 30 years or so, we’re all breaking several every day.

That’s why the rate of guilty pleas is over 98%—they can always get you for something, and a plea bargain at least shortens the ordeal.

Phil Hobbs

A wart used in an EMC certification becomes part of it. Hence mrfs listing and retailing part numbers for suitable use.

Warts can be (and are) listed independently, to reduce potential testing and deployment gliches. A listed wart doesn't guarantee radiated compliance, only facilitates conducted performance on that one, main, port.

RL