Hi,
I thought I’d provide a quick update for the benefit of other Australian folk who stumble across this. There was an option for an Australian lab (EMC Bayswater) to review the Photon FCC report, so I paid $660 for that, the outcome being that extreme voltage and temperature testing needs to be performed (~2k) at minimum. It’s still not 110% clear if I still need to do EMI testing, will see how that goes.
@will, I must say I love the content in the building a product (certification) guide at https://docs.particle.io/guide/how-to-build-a-product/certification/. Not sure if its been recently filled out or added, but its great. If you’d like a copy of the Australian review report detailing the necessary extra tests then you’re welcome.
@mterrill thanks for the kind words! It hasn’t changed much since we’re still waiting on some of the final Electron certification numbers to come through, but it is definitely due for an update. I know very little about Australian certifications, and would love to learn more!
I’ll try to remember to post to this thread again once I’ve updated it for the Electron certifications that come through.
@will a few questions regarding certification of a product containing the P1.
Has particle done any SAR (specific absorption rate) testing on the photon/P1? Our cert lab brought it up since our product may be in close proximity to meatbags humans.
Our product requires FCC co-location testing (we have a second BLE radio on board) which requires firmware for FCC test cases (blast signal on band X with [b/g/n] modulation) is that firmware available anywhere?
We haven’t yet done SAR testing, so you’d have to to do that one separately.
I can work on getting the test firmware to you from the manufacturer of the P1, USI. We have worked with them in the past on verification of certification of products with the P1 on board.
Please, if the firmware could be put on github or the cloud libraries that would be most appreciated. I’ll need it within the next 10 days if that’s possible?
Firstly thank you for the information about your results from EMC Bayswater, it’s generous of you to share this!
I’ve only just stumbled on this thread again after last viewing it months ago and I have to say that, like you I’d assumed that the FCC compliance would carry over (I’m based in NZ looking at NZ/AUS certification and possibly CE).
I’m finding it pretty overwhelming trying to understand different countries requirements. Do you have any updates on your progress with this extreme voltage & temperature testing? I assume that’s what you’re after the firmware for?
Furthermore @mterrill, have you found that you are going to be able to reuse some of the ETSI reports in terms of intentional radiators?
My understanding from @hfiennes comment was that any product going for CE would require all of the intentional radiator tests to be redone - since CE does not accept modular approval, however from your comment regarding just doing the extra extreme voltage and temperature testing, I’m a little confused…
If your device is intended to be hand-held or body-worn, with a output power of >50mW (16.9dBm EIRP), you generally need to do SAR testing. There are different rules for different parts of the body - in particular hand-held is less stringent than body worn (think: walkie talkies).
As I remember, you can completely skip SAR if your EIRP is <50mW (this includes “worst case” (ie maximum) antenna gain). Depending on your product, you may be able to turn down the maximum transmit power - sacrificing range - to avoid the need to do SAR at all.
We did SAR testing on the first imp, and kept our ~80mW EIRP (+17dBm output power in CCK modes, +2.5dB antenna gain) - we passed device independent testing - which has a lower maximum absorption limit) at 20mm separation as I remember. SAR testing is quite cool (they use a robot arm and a tub full of “human body simulator” goo!) but also not cheap (~$10k).
On CE: CE is self certified. There is no such thing as CE modular approval, but modules can be tested to CE specs - that just doesn’t guarantee your product will pass, it only guarantees that on a test board, the module passed product level specs. There are plenty of modules with CE marks which are obviously lying about it - eg the ESP8266 ones which note a +25dBm output power and then also say CE, which is not possible as this is 5dBm over the legal CE power levels…
People DO ship stuff with modules in it without doing intentional emitter testing, but you should speak to a lab and determine whether you think this is a good path. If you CE mark your product and can’t produce any backup that shows that you tested it to CE specs, then you may end up in a bad place. See my post back in July’15.
Obligatory picture of SAR testing setup. The unit being tested is located under the tub of goo - the robot moves a field probe around in the liquid, mapping out the strength at specific points.
I haven’t looked too hard but the photon/P1 don’t list output power. The Core used to be 18dBm.
Our product isn’t “mobile/wearable” but might be mounted close enough to a person. The testing lab brought it up but I think it might have been an upsell, another lab I spoke with says we just need a warning blurb in the instructions somewhere.
Also, if anyone else is looking at certifying a device with another radio on board, the option that I’m most likely to go wtih is to ask for a Class II permissive change that allows co-location. No testing should be required. The lab will just do calculations based on the FCC testing docs of both modules and make sure the emissions are still within the limits.
If the product isn’t held or worn then you generally don’t need to do SAR; for example, whilst you can hug your router, this isn’t the normal use case and hence routers don’t need SAR approval.
(btw, I mis-remembered the 50mW ceiling for SAR - it’s actually 25mW at 2.4GHz, see the above link).
If you’re putting multiple radios in one product, you should also look at antenna coupling if they are both in the same band. This can put a serious dent in your wifi range under certain circumstances.
Hi @will, just following up, are you able to put up the firmware to do the radio testing? IE it walks through the bands, channels, low power, high power?
The USI instructions seem to be using an early board design with firmware instrumented I presume over the serial pins, whereas the other document shows a photon and JTAG programming.
I had thought (perhaps naively) that there would be an .ino (or multiple) that would go through the various scenarios, even potentially triggered via cloud functions to switch.
May I ask:
Could I simply zap one of those .bin’s to the photon via the particle CLI rather than JTAG? (do you need all three?)
I’m thinking that a lot of folk will need to do certification testing (or they should be…) so having this as a straight forward process would be quite advantageous.
Japan and France have the toughest standards for emissions mask compliance. That’s harder to pass than conducted emissions. Their mask skirts are so steep that many vendors have to restrict the Tx power to much less to pass in those regulatory domains.
There are two sets of instructions–one for flashing the firmware file to the Photon, and one for running the Wi-Fi test script that interacts with the device.
You can flash firmware to the device using the JTAG pins. We offer a Programmer Shield in our store that would allow you to do this if you do not have your own JTAG programmer.
The wl.exe script is used to send test commands to the firmware running on the Photon.
The green board in the instructions is for a different development PCB for the BM-09 (P0)–we’re using USI’s firmware that was originally developed for the development board that they used for FCC certification. That looks like a USB serial converter that they’re using to connect the device to their computer.
