Slashdot Mirror
Raspberry Pi Passes EU Electromagnetic Compatibility Testing
A week ago, we posted news of the delay that the Raspberry Pi Foundation faced because of a requirement that their boards be tested to comply with EU regulations. Now, the word is in, and the Raspberry Pi passed those tests without needing any modifications. From their post describing the ordeal: "The Raspberry Pi had to pass radiated and conducted emissions and immunity tests in a variety of configurations (a single run can take hours), and was subjected to electrostatic discharge (ESD) testing to establish its robustness to being rubbed on a cat. It’s a long process, involving a scary padded room full of blue cones, turntables that rise and fall on demand, and a thing that looks a lot like a television aerial crossed with Cthulhu."
Thank goodness it passed the cat rubbing test. We Europeans love rubbing electronic devices on our cats.
"We live in a global world" - Harvey Pitt, former Securities and Exchange Commission Chairman
Am I the only one that desperately wants to see pictures of the Cthulhu antenna?
Bad kitty! That's my Pot Pi! No! You're a bad kitty!
Also, cats.
WHERE'S THE CAT
...
*CLAP*
<xml><I><am><so><damn>Web 2.0</damn></so></am></I></xml>
We'll send you the cat as soon as Farnell finishes v7 of their price matrix, assuming they've read any of the last 20 mins emails they've been sent about the requirement that they sell to end users.
It was written by the PR person, not the engineers.
I actually thought the comment was tongue-in-cheek, so, no I am not scared by them.
The PR person is a food blogger who does this for free in her spare time because she is married to one of the foundation Trustees. Cut her a little slack.
The Daddy casts sleep on the Baby. The Baby resists!
Doesn't every product, everywhere, pass this test?
So, this is worthy of the front page why?
Irrispective of who wrote it I think that it was meant as a joke, which I personally found funny and lol'ed a bit. My my so touchy....
After the testing you will be baked and then there will be cake.
Time flies like an arrow. Fruit flies like a banana.
All electronics that are going to be sold, as finished products, in the European economic area (EEA) have to be tested and comply with European standards. It's the short answer, and I'm skipping a lot of details.
The problem the Raspberry-foundation faced was that it was initially not a "finished" product, more of a DIY kit. Once it became clear it was more of a "consumer" product it had to comply and be tested.
The same applies in the US where the FCC has the same role, but labs do the actual testing in both jurisdictions.
From TFA:
A cute story. Radiated immunity testing involves hitting the Raspberry Pi hard with narrow-band EM radiation, while checking (amongst many other things) that the device is still able to send Ethernet frames to a hub. The first time the team did this, the light on the hub stopped blinking: no frames were making it through. They did it again: still nothing. Finally, they discovered that the hub (which, I should point out, gave every appearance of being CE marked, so it should have been able to get through these tests itself) was being knocked out every time somebody pressed the button. Jimmy used a longer cable, put the hub outside the field, and found that the Raspberry Pi got through its immunity tests with no problems at all.
Too bad their CE certified ethernet hub failed the CE testing.... remember kids, this is what you get when you buy cheap stuff from cheap manufacturing countries.... oh wait!
This is a charity and this is the first time they have gone through this process. There are bigger companies still have this issue. With pi they are just being more transparent about it.
The interesting thing is they were forced to do this at all.
Farnell in Norway does NOT sell to end users, only companys and developers, and as such the boards are not supposed to be forced to adhere to this testing.
It smells like some competitors have gotten their will here, and it's nice to see that they didn't win. Now the sales will be even better from the get-go.
I already know of several projects that may use the pi as a base-platform now that it passed the tests.
Not when one of the FCC regulations is "must accept interference". No seriously, that's actually a requirement.
I don't think that pacemakers would qualify as Part 15 devices; but Part 95 ones. Those have a much shorter list of things they aren't allowed to interfere with; but they are still required to deal with interference(that and, obviously, building life-critical systems that can't handle a little RF would be a Bad Thing even if it were legal.)
The issue is with the distributors, not the RaspPi people.
Farnell and RS got nervous when they realized how many of the boards they would be shipping. There is not the same requirement for low volume eval boards they sell as engineering prototypes.
The Foundation always planned on obtaining the CE mark for the Raspberry Pi boards during the main launch, which will come in the future when schools have their curricula worked out and huge numbers of the finished devices (in enclosures mostly) will be going out to schoolchildren. Right now the boards are seen as a preliminary release. The Foundation had CE certification on the schedule. Just not this soon.
Its sort of a cabal arrangement. The big established device manufacturers collude with the regulatory bureaucrats. They do so to insure there is always a substantial barrier to entry to protect themselves from those meddlesome low budget startups with their market-disturbing innovations.
Big companies have whole departments dedicated to "handling" the compliance/regulatory stuff. Its just good (borderline monopolistic) practice to operate that way, and defend the regulatory requirements in a bellicose tone any time the requirements are questioned.
I worked for years in the medical device field. The big companies know very well how to design and package a device with no more complexity than a simple AM/FM Radio into a product Insurance companies can be billed $800-1000 for.
You're both semi-wrong here: The testing was for EU regulations, nothing to do with the FCC. Though they may want their own tests done in addition.
In case anyone's interested, the chamber referred to is fairly similar to this: http://www.flickr.com/photos/ethicsgradient/sets/72157606434322104/ which is the EM anechoic chamber at my old job. No cthulhu antenna but all the spikes you can eat.
There you go http://www.raspberrypi.org/wp-content/uploads/2011/12/mooncakehelpsout.jpg
(from a forum thread 'Pi's and small fuzzy animals. How will you keep them apart? ')
Your welcome.
"they claimed they already gave 10000 boards to distributors"
Errr.... NO.
The first batch of 10000 was produced. When these arrived in the UK, the Foundation performed full functional testing on samples from the batch (as compared to electrical tests carried out at the factory) and discovered that the manufacturers had substituted a certain component between the first manufacturing samples and batch production. All the production was returned to the manufacturer for rework, which has taken some time. Reworked stock is now in the UK and going into distribution. Delivery is now contingent on RS and Farnell receiving copies of the test results so they are satisfied in their own minds that the Raspberry Pi can be sent out. Neither RS Components or Farnell have had "a 10k batch" to hand. Ever.
Its all on the Raspberry Pi forum you know. Whats "nice" is your complete lack of knowledge and comprehension. :-)
The details vary but ultimately there does have to be some firmware located in a place that is non-volatile and directly accessible to kick off the boot process.
Older devices used paralell flash directly on the data bus. The bootloader was then executed directly from this flash and went on to load the kernel.
Most recent arm devices have a very small boot program on the chip itself. This chip then reads a bootloader from somewhere (usually NAND flash or SD card) which in turn loads the kernel.
The Pi is a bit unusual in that the GPU starts first. The GPU starts from a small peice of code on the chip itself, reads it's firmware from the SD card (in several stages) and then loads the kernel, resets the SD card device and starts up the arm core.
note: i'm known as plugwash most places but i screwd up registering that here somehow in the past and now can't register
Generally there is a small ROM embedded in CPU that loads another bootloader from NAND, SD card, SPI Flash, etc. On Atmel ARM chips that bootloader must be small enough to fit into embedded SRAM. Than bootloader initializes SDRAM and fetches U-Boot into it. U-Boot in turn may initialize wider range of devices and then load Linux kernel.
All boot process is very SoC- and board-specific. Bootloaders must be compiled for selected CPU and board components, and Linux kernel should also have board description down to what types of chips are installed as autodetection is usually very limited.
I'm sorry, I only know what I read on their website. As far as I can tell they're shipping them out as soon as possible subject to their certification papers being "approved" by the distributors in question. The CE certification itself is official and valid so there really is no reason for further delays.
If you're suggesting FM broadcast, you're dealing with a step phase shift every tenth of a second. Yes, you can "time it right" for one point (say, equidistant from all ten) but at other points it will vary considerably. I'd have to run the math (which I can't be arsed to do) to put numbers to it, but I'm pretty sure you'd get audible buzz at some multiple of 10 Hz on some receivers.
More importantly, note that a lot of cheap FM receivers have pretty poor AM rejection, thus the volume will "throb" through ten values related to the distances, repeating every second. Even for receivers with perfect AM rejection, the noise floor will throb in the opposite pattern (louder for faint signals), so I really don't see this as a feasible solution.
Oh, and people tracking you down? With a high-gain antenna (e.g. Yagi-Uda), guess what, the needle on my S-meter is pulsing up and down at 1 Hz -- So I sweep more slowly and just watch for the strongest peak. Then I'll move toward that signal (making it even stronger), until I get to it. Yes, I could be "stuck" equidistant between the two nearest ones -- in which case I'll pick one of the directions at which I get an equally strong signal, and go. Then it's stronger, so I keep following it.
It would pose a problem for old-school DF with a loop antenna, where I'm trying to null out the signal, and these are preferred for DF per se (because antenna nulls are steeper than peaks, it gives you more accurate direction), but it's not like the FCC guys only have one trick in their bag, and a lot of fox-hunters do primarily use Yagi-Udas because the waving an antenna in front of you is a lot easier/more intuitive, particularly while moving, than holding a loop overhead and sweeping for the null.
They also did FCC (US), Canadian and Australian (don't know the acronyms) testing at the same time.
You, obviously. We put them up to annoy you. I vote we have a Raspberry Pi week to annoy all the haters!