7 ms·

Raspberry Pi faces distribution troubles over CE certification

33 points by GigabyteCoin 15y ago

chrisacky 15y ago
This seems like one of those things that should of been negotiated before they appointed the contract to RS/Farnell, right? Presumably RasPi or Farnell/RS had previously discussed the distribution of the devices and I'm assuming that they at least made some suggestion that they aren't going to ship until CE certification was granted. So this shouldn't really be a surprise. If on the overhand Farnell/RS have both just said that they aren't going to ship out of the blue, then the cynic in me would probably guess that there would be some other underlying reason why there is a delay, such as other manufacturing problems. Either way, I don't think too many people are going to be too concerned. They will get here when they get here.
bprater 15y ago
Non-hardware hacker question: as the Pi is modified over time, does it need to be re-certified each time? Or is it certain components on the board that require this?
- radiowave 15y ago
  Simple answer is "it depends". The device has to be assigned a risk category, and depending on that there's a range of possibilities from the manufacturer being able to self-certify (so long as they stand ready to demonstrate the design process that went into a new revision of the product), and at the other extreme, every revision having to be independently verified. My employer is able to self-certify, but our products are much simpler than a Raspberry Pi.
  - polshaw 15y ago
    What are your products? PCBs?
    - radiowave 15y ago
      Shackles, and other steel hardware.
latch 15y ago
I cannot reconciliate the fact that I believe regulations kill innovation in our industry, while at the same time believing that lack of regulation in finance is destroying the economy.
- andybak 15y ago
  Is it possibly as simple as the fact that you can smart regulations and dumb regulations?
- furyg3 15y ago
  While this is annoying for the Raspberry Pi guys, I don't think their innovation is being stifled... Requiring certification when you're planning on making a few hundred units and selling them yourself to a few schools (or whatever) in the UK would be stifling innovation. Requiring certification when the project explodes, scaling up new distribution channels and production, shipping internationally, seems reasonable. Considering how popular it's gotten, and how many, many thousands of these are going to be sold in the near future, it's nice to know that these meet some basic standards.
- retube 15y ago
  > while at the same time believing that lack of regulation in finance is destroying the economy It's not a lack of regulation in finance that's destroying the economy, it's too much. The rules and regulations now are so punitive, so lengthy and so complex that it's making some (many) banking services non-viable or at best very expensive. And ultimately it's the end consumer which bears these increased costs as they get passed on. This means banks make less money, customers have to pay more, staff get paid less or cut, shareholders get reduced dividends, pension funds lose out from suppressed stock prices etc. Many of these rules don't make sense or make the bank "safer" - they're really just a penalty for conducting finance-orientated business - with such rules sadly being driven by politicians attempting to appease the baying mob. It's unfortunate that these regulations will have the opposite effect of what they are trying to achieve. Indeed some very large names may actually go out of business as a result which would be catastrophic for the economy.
  - 15y ago
ukdm 15y ago
Eben states in the comments on the original post that both distributors didn't include compliance people during the contract stage. Raspberry Pi assumed that they could be like the Beagleboard and not require CE marking. So everyone involved overlooked it. More of the blame lies with the distributors I feel. I mean, they are meant to be the experts here, and have done this many times before.
polshaw 15y ago
Can we use this opportunity to actually discuss what is involved in CE certification in this case? It doesn't have any radios so electromagnetic compatibility directive shouldn't apply, and the low-voltage directive states it's for 'equipment designed for use with a voltage rating of between 50 and 1000 V for alternating current (A.C.) and between 75 and 1500 V for direct current (D.C.)'.. since it runs on 5v DC and has no PS it should be ok there too [?: If it shipped w/ an integrated PS, this would apply? would the whole device need testing here or just the PS??]. So what does that leave and what does this testing involve?? (E: it seems to be EMC testing from rPi.org, is this all, and what would it involve?) FWIW, to me it sounds a little like F/RS want this to be certified as a finished product so they can sell it in bundles, and profit maximise that way.
- stephen_g 15y ago
  In Australia at least, pretty much any electronic device has to be tested to ensure unintended radiation emission between 9kHz and (I kid you not) 400GHz is within acceptable limits before it gets a C-tick (similar to the CE certification) and be sold. I think most of the time they stick the device in an anechoic chamber and take measurements at certain distances. For example, under the CISPR 22 standard (that information technology equipment with a supply voltage of less than 600V falls under) it must radiate less than 30dB (uV/m) of electromagnetic radiation between 30 to 230MHz, measured at a distance of 10 metres with a quasi-peak detector, less than 37dB at 230 to 1000MHz, and so on. There are also restrictions of how much interference is caused from the power supply (which shouldn't apply to the Raspberri Pi) and how much EM is emitted from telecommunications ports on the board (which will apply to Ethernet, but probably not USB).
  - polshaw 15y ago
    I am guessing that most PCB/PS -like electronics would pass this without any specific attention, given there were no major manufacturing defects? I'm a little unsure why TI made efforts to avoid it.. or just because they could?
    - stephen_g 15y ago
derda 15y ago
I am wondering, why they haven't thought about those problems before. CE certification, EMC testing, WEEE (european electronic waste regulations) are a big part of why hardware is hard. I don't know about specific regulations in the UK, but I guess they are similar to those here in germany. You can get around those regulations if your product is for use by professionals and/or for development purposes, that means there is some professional knowledge needed to make the product work (soldering, programming,...). For the rPI you only need a power adapter and a downloadable image on a SD card, thats not really professional knowledge. For CE certification, you can however put a CE badge on anything you want, and refer to some regulation (I don't have the paragraph at hand atm). This basically means "We think it will be alright". A frightening number of Chinese imports do this. This is absolutely legal, the problem comes when something bad happens, e.g. your product causes a pacemaker to fail, then you have to prove that you took all necessary precautions to prevent this, which is basically impossible if you haven't done proper EMC testing and such. Designing complex circuits (like the rPI) that do not make any EMC problems is extremely hard. I don't think most computer hardware would go through testing/certification without problems, thats why your mainboard sits in a metal case. My worst case guess is that they have to ship the boards in cases or with some shields. Disclaimer: I am a ECE student. My father is a consultant in the CE/WEEE field. Statements above are based on memorys of conversations with him, when I told him I had an idea for some hardware product (and he told me to forget it unless I want to spent a lot of money on regulatory stuff)
RLG_RLG 15y ago
Sounds like this is going to be $99 by the time it is de-bugged (on manufacturing side) and certified. I am angry and frustrated because I want one ;-)
ChuckMcM 15y ago
Hopefully they aren't trying to sell them in Japan, the certification process there is waaaaay painful.

33 points by GigabyteCoin 15y ago

This seems like one of those things that should of been negotiated before they appointed the contract to RS/Farnell, right? Presumably RasPi or Farnell/RS had previously discussed the distribution of the devices and I'm assuming that they at least made some suggestion that they aren't going to ship until CE certification was granted. So this shouldn't really be a surprise. If on the overhand Farnell/RS have both just said that they aren't going to ship out of the blue, then the cynic in me would probably guess that there would be some other underlying reason why there is a delay, such as other manufacturing problems. Either way, I don't think too many people are going to be too concerned. They will get here when they get here.

bprater 15y ago

Non-hardware hacker question: as the Pi is modified over time, does it need to be re-certified each time? Or is it certain components on the board that require this?

radiowave 15y ago
Simple answer is "it depends". The device has to be assigned a risk category, and depending on that there's a range of possibilities from the manufacturer being able to self-certify (so long as they stand ready to demonstrate the design process that went into a new revision of the product), and at the other extreme, every revision having to be independently verified. My employer is able to self-certify, but our products are much simpler than a Raspberry Pi.
- polshaw 15y ago
  What are your products? PCBs?
  - radiowave 15y ago
    Shackles, and other steel hardware.

latch 15y ago

I cannot reconciliate the fact that I believe regulations kill innovation in our industry, while at the same time believing that lack of regulation in finance is destroying the economy.

andybak 15y ago
Is it possibly as simple as the fact that you can smart regulations and dumb regulations?
furyg3 15y ago
While this is annoying for the Raspberry Pi guys, I don't think their innovation is being stifled... Requiring certification when you're planning on making a few hundred units and selling them yourself to a few schools (or whatever) in the UK would be stifling innovation. Requiring certification when the project explodes, scaling up new distribution channels and production, shipping internationally, seems reasonable. Considering how popular it's gotten, and how many, many thousands of these are going to be sold in the near future, it's nice to know that these meet some basic standards.
retube 15y ago
> while at the same time believing that lack of regulation in finance is destroying the economy It's not a lack of regulation in finance that's destroying the economy, it's too much. The rules and regulations now are so punitive, so lengthy and so complex that it's making some (many) banking services non-viable or at best very expensive. And ultimately it's the end consumer which bears these increased costs as they get passed on. This means banks make less money, customers have to pay more, staff get paid less or cut, shareholders get reduced dividends, pension funds lose out from suppressed stock prices etc. Many of these rules don't make sense or make the bank "safer" - they're really just a penalty for conducting finance-orientated business - with such rules sadly being driven by politicians attempting to appease the baying mob. It's unfortunate that these regulations will have the opposite effect of what they are trying to achieve. Indeed some very large names may actually go out of business as a result which would be catastrophic for the economy.
- 15y ago

ukdm 15y ago

Eben states in the comments on the original post that both distributors didn't include compliance people during the contract stage. Raspberry Pi assumed that they could be like the Beagleboard and not require CE marking. So everyone involved overlooked it. More of the blame lies with the distributors I feel. I mean, they are meant to be the experts here, and have done this many times before.

polshaw 15y ago

Can we use this opportunity to actually discuss what is involved in CE certification in this case? It doesn't have any radios so electromagnetic compatibility directive shouldn't apply, and the low-voltage directive states it's for 'equipment designed for use with a voltage rating of between 50 and 1000 V for alternating current (A.C.) and between 75 and 1500 V for direct current (D.C.)'.. since it runs on 5v DC and has no PS it should be ok there too [?: If it shipped w/ an integrated PS, this would apply? would the whole device need testing here or just the PS??]. So what does that leave and what does this testing involve?? (E: it seems to be EMC testing from rPi.org, is this all, and what would it involve?) FWIW, to me it sounds a little like F/RS want this to be certified as a finished product so they can sell it in bundles, and profit maximise that way.

stephen_g 15y ago
In Australia at least, pretty much any electronic device has to be tested to ensure unintended radiation emission between 9kHz and (I kid you not) 400GHz is within acceptable limits before it gets a C-tick (similar to the CE certification) and be sold. I think most of the time they stick the device in an anechoic chamber and take measurements at certain distances. For example, under the CISPR 22 standard (that information technology equipment with a supply voltage of less than 600V falls under) it must radiate less than 30dB (uV/m) of electromagnetic radiation between 30 to 230MHz, measured at a distance of 10 metres with a quasi-peak detector, less than 37dB at 230 to 1000MHz, and so on. There are also restrictions of how much interference is caused from the power supply (which shouldn't apply to the Raspberri Pi) and how much EM is emitted from telecommunications ports on the board (which will apply to Ethernet, but probably not USB).
- polshaw 15y ago
  I am guessing that most PCB/PS -like electronics would pass this without any specific attention, given there were no major manufacturing defects? I'm a little unsure why TI made efforts to avoid it.. or just because they could?
  - stephen_g 15y ago

derda 15y ago

I am wondering, why they haven't thought about those problems before. CE certification, EMC testing, WEEE (european electronic waste regulations) are a big part of why hardware is hard. I don't know about specific regulations in the UK, but I guess they are similar to those here in germany. You can get around those regulations if your product is for use by professionals and/or for development purposes, that means there is some professional knowledge needed to make the product work (soldering, programming,...). For the rPI you only need a power adapter and a downloadable image on a SD card, thats not really professional knowledge. For CE certification, you can however put a CE badge on anything you want, and refer to some regulation (I don't have the paragraph at hand atm). This basically means "We think it will be alright". A frightening number of Chinese imports do this. This is absolutely legal, the problem comes when something bad happens, e.g. your product causes a pacemaker to fail, then you have to prove that you took all necessary precautions to prevent this, which is basically impossible if you haven't done proper EMC testing and such. Designing complex circuits (like the rPI) that do not make any EMC problems is extremely hard. I don't think most computer hardware would go through testing/certification without problems, thats why your mainboard sits in a metal case. My worst case guess is that they have to ship the boards in cases or with some shields. Disclaimer: I am a ECE student. My father is a consultant in the CE/WEEE field. Statements above are based on memorys of conversations with him, when I told him I had an idea for some hardware product (and he told me to forget it unless I want to spent a lot of money on regulatory stuff)

RLG_RLG 15y ago

Sounds like this is going to be $99 by the time it is de-bugged (on manufacturing side) and certified. I am angry and frustrated because I want one ;-)

ChuckMcM 15y ago

Hopefully they aren't trying to sell them in Japan, the certification process there is waaaaay painful.