3

I got a job to do a household appliances with wireless remote control. I’m working as embedded developer in current company, but we haven't had to certificate our solutions and I'm looking for advice.

The first level of this job is to make a prototype of the machine. However I want this prototype not to demand too many modifications during the certifications process before entering the EU market.

As for my internet research, I know that the necessery things to do are: certification EMC, RED, LVD and additionally norms IEC-60335 and IEC-60730 for Class-B machine (household appliances).

Has anyone of you designed or made a household device controled by MCU?

Is following norms IEC-60335 and IEC-60730 demanded? Are making the tests obligatory for the norms? Or only the producer statement is enough to let the product enter the market?

What documents are needed to conduct the test of this kind of appliances in external laboratory (PCB circuict schematic, list of norms, elements list, software architecture, code source, code/libraries certifications etc.)?

For Class-B device necessery is to use mechanisms self-test peripherals of MCU (such as CPU, RAM, Flash, Watchodgs etc.)?

But the producer of the chip (Nordic Semiconductor - nRF5) I chose doesnt provide the self-test library or the support for this kind of tests. Can I go through the certification process implementing this kind of self-test by my own or using for example only watchdogs?

For example, ST Electronics provides certified library of this type of self-test for free. However its wireless modules are more difficult to get and much more expensive than nRF modules.

I would like to use nRF modules because its precertified with FCC/IC etc. And here I have another question: whether the certification of my own remote controller with precertified module on my own PCB circuit with battery (CR2032 3V3) supply and 4 buttons is as expensive as using a not precertified radio module?

UPDATE:

To precise: my goal is to designing and making a prototype a remotely controlled cooker hood. I have wanted to use RF modules in remote controller and mainboard cooker hood. The LED lights and turbine (low voltage 10V) would be controlled by the MCU. The voltage to mainboard would be provided by the 12V power supply.

As I found out the norm IEC-60730 divides the household devices in to 3 classes:

  • Class A - Control functions, which are not intended to be relied upon for the safety of the equipment. Examples are: room thermostats, humidity controls, lighting controls, timers, and switches.
  • Class B - Control functions, which are intended to prevent unsafe operation of the controlled equipment. Examples are: thermal cut-offs and door locks for laundry equipment.
  • Class C - Control functions, which are intended to prevent special hazards (e.g., explosion of the controlled equipment). Examples are: automatic burner controls and thermal cut-outs for closed, un-vented water heater systems. [https://www.nxp.com/docs/en/application-note/AN3257.pdf]

Theoretically it would be possible to define that control as Class A - lightning control, timers and switches. It would make the process of designing easier, because the software systems of safety aren't demanded in this class. Unfortunately the control of a turbine makes this project as device of class B.

Thank you for your answers, I hope they would be useful to others in the future, because the subjects of meeting the standards, safety regulations, certification of products and the right way of conducting the project are not well described on the internet and this is the next step after the own DIY projects.

martin_embed
  • 63
  • 1
  • 6
  • 1
    Regarding your update, you have to make a risk assessment and I suppose address the concern of fire. Could a malfunctioning board lead to a fire hazard etc if smoke spreads through the kitchen instead of going up through the chimney? If so then you end up in the B or C category. I would imagine there's all kinds of building & fire protection regulations. – Lundin May 06 '21 at 08:28
  • 1
    As a side note, I really don't understand why you'd want to make such a product wireless. Seems like it would only serve to turn the product needlessly complex and expensive. – Lundin May 06 '21 at 08:30
  • Wireless control is not up to me. It is the customer's decision. I have to use expert advice, it's too complex topic. Thank you. – martin_embed May 07 '21 at 08:59

2 Answers2

6

The detailed technical information on your product would be needed to advise on exactly what standards may or may not be applicable. But I can talk in general terms.

I know that the necessery things to do are: certification EMC, RED, LVD...

This is not anywhere near as simple, you also need to declare compliance with the correct parts of each. For example, you cannot declare compliance with the safety parts of the LVD if you need to declare compliance with RED.

I would strongly advise getting consultancy from a test house (or a dedicated compliance consultant), they are usually very helpful. I would also look at the Declarations of Conformity from other similar products already in the market. This will give you an idea of what standards you may have have to declare conformity against (a test house will always tell you that what you need is more tests, having an idea of what may possibly be required will allow you to ask intelligent questions).

Are making the tests obligatory for the norms? Or only the producer statement is enough to let the product enter the market?

Technically, yes. You can self-declare compliance with only a Technical File. I can't emphasise enough that I think this would be an extremely bad idea, as you don't seem to have the knowledge required. One of the possible things that can happen if you are found to be non-compliant in the market is forcible recall from customers at your cost. These regulations have teeth and can bankrupt companies.

whether the certification of my own remote controller with precertified module on my own PCB circuit with battery (CR2032 3V3) supply and 4 buttons is as expensive as using a not precertified radio module?

It should be the same price, the same tests need to be made. The advantage is that you have a good chance of passing emissions with a pre-certified module. It is one less, major, uncertainty in your compliance process.

Cursorkeys
  • 1,118
  • 7
  • 12
  • 2
    "a test house will always tell you that what you need is more tests, having an idea of what may possibly be required will allow you to ask intelligent questions" Yep this is important to know, since it might save you a fortune in test costs. Always check the scope of every standard they suggest. – Lundin May 05 '21 at 10:11
4

Has anyone of you designed or made a household device controled by MCU?

Not as such, but industrial applications with very similar requirements. IEC 60730 is one of the "functional safety" ("SIL") standards, there exist different ones but all of them are pretty much the offspring of the industrial IEC 61508 functional safety standard.

I can't tell you if IEC 60730 compliance is required by law for your particular application. There's usually product-specific standards stating such. If it is then you'll need help from a test house, preferably at design stage of the product. These standards are complex and incredibly time consuming, which in turn makes them expensive, with lots of consultant hours. Some rule of thumb is that the cost and time to market are both multiplied by at least x3. In worst case maybe ten times the time & cost.

What documents are needed to conduct the test of this kind of appliances in external laboratory (PCB circuict schematic, list of norms, elements list, software architecture, code source, code/libraries certifications etc.)?

For 60730 you'll need to demonstate compliance to every part of the document. This includes your quality system, your documentation, your hardware design, your MCU and your software design. All depending on what safety level of the standard that is required. You will likely have to use a safety MCU, I'd recommend checking out NXP parts (I'm not affiliated) since they are big in this area.

As for software, it likely means MISRA-C plus misc other functional safety practices. I wrote a little check list here which sums up the most common requirements. MISRA-C is a major undertaking too if you have never worked with it and have no C programming veterans (10+ years of C experience) in the company.

For Class-B device necessery is to use mechanisms self-test peripherals of MCU (such as CPU, RAM, Flash, Watchodgs etc.)?

From what I vaguely recall from checking out that standard at some point, then yes. You need lots of self tests in the MCU hardware. ECC memory also helps a lot, since implementing flash CRCs is complex and a pain in general. Why picking a MCU with all this baked in is pretty much mandatory, re-inventing the wheel is a very bad idea here.

Another important thing to be aware of is that the functional safety standards are roughly 50% utter nonsense and 50% really good stuff. The key is to figure out which... You can usually argue about various nonsensical requirements in your compliance documentation. Test houses usually don't have very specific technical expertise of these requirements either and it is them that you have to convince. You can do this by using the same methodology as the safety standards themselves, here is an example where I make an argument against some nonsense NASA software rule by applying "FMEA" (which safety people love).

EMC/RED/FCC/IC etc testing will be a breeze compared to functional safety. I wouldn't worry as much about picking some part just because it has pre-certified radio. If you have a requirement for IEC 60730 then that's your main concern and everything else is secondary. It's far easier to manually make some radio compliant to a standard, though I guess that depends on what competence you have in house.

Yet another thing that might be a major pain in the neck is electrical safety testing for LVD/UL, in case that applies to your product.

Lundin
  • 17,577
  • 1
  • 24
  • 67