Searching \ for '[OT] Anyone done FCC certification?' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: techref.massmind.org/techref/index.htm?key=anyone+done+fcc
Search entire site for: 'Anyone done FCC certification?'.

Exact match. Not showing close matches.
PICList Thread
'[OT] Anyone done FCC certification?'
2006\09\13@102518 by Gerry Duprey

flavicon
face
Howdy,

I have a person/client who wants to use one of my boards in a commercial
product.  They are concerned about getting FCC certification for it and
I really have no experience with it.

I think they can "dodge" the UL stuff by simply using an external UL
approved power supply/wall wart (so everything inside the product is
5volts), but as I understand the FCC stuff, if you have a
clock/oscillator over 9Khz, it needs to be certified.

If anyone has gone through this process, I'd really like to hear some
details about it.  Particular questions:

1) Are there any exemptions for small batch/lot products (these will
mostly be "art works")?

2) Did you have to do a lot of shielding?  The controller is a PIC
running at 8Mhz doing a lot of RGB LED PWM stuff (http://www.rgbled.org).  I
don't think it throws off that much RF, but...

3) Did you use one of those FCC certification shops?

4) Do you have a rough idea how much the process costs?  This could be
for class B certification (residential)?

5) Did you also get a CA certification too?  If so, was the pricing similar?

Of course, while all the questions are important, #4/5 are really what
are going to sink/float the product.

Thanks again,

Gerry
--
Gerry Duprey
Ann Arbor, MI 48103
http://www.cdp1802.org

2006\09\13@110240 by Russell McMahon

face
flavicon
face
> I  don't think it throws off that much RF, but...

You should KNOW.

Almost everything electronic with a processor in produces SOME
radiated noise.

Take one standard AM/FM transistor radio.
Turn it on.
Place it literally on top of the operating product.
Tune across AM and FM bands.
Note where you hear product produced noise.
If in doubt turn product on and off and note what happens.
On each noise peak move radio away and vary orientation and note how
strong each one is and how far away you can hear it.

This wont qualify as any sort of formal test but afterwards you'll
have a much better feel for what you are up against.

A manually tuneable TV makes an even better test receiver as you can
SEE the interference as well and this can be most useful.



       Russell





2006\09\13@112115 by David VanHorn

picon face
>
>
> 1) Are there any exemptions for small batch/lot products (these will
> mostly be "art works")?


Possibly. There is an exemption for prototypes, as long as you maintain
ownership.

2) Did you have to do a lot of shielding?  The controller is a PIC
> running at 8Mhz doing a lot of RGB LED PWM stuff (http://www.rgbled.org).  I
> don't think it throws off that much RF, but...


The rules of noise:
1: Don't make it.
2: Deal with it at the source.
3: Shield it.

Try #1 and #2 first.  Return currents to where they came from, not somewhere
else.
Place crystal caps and route them, so that they return ONLY to the
processor's ground pin, not to other places.
Avoid dumping noisy signals into ground planes, give them direct and low
impedance paths back to their sources.



3) Did you use one of those FCC certification shops?


Many times.  I do my own pre-scans using an Icom IC-R8500 receiver, which
has "one band" coverage from 0.1 to 1999 MHz

4) Do you have a rough idea how much the process costs?  This could be
> for class B certification (residential)?


B is more, A runs about $2000 for a half-day.

5) Did you also get a CA certification too?  If so, was the pricing similar?


No


--
Feel the power of the dark side!  Atmel AVR

2006\09\13@113436 by M. Adam Davis

face picon face
My limited experience:

> I think they can "dodge" the UL stuff by simply using an external UL
> approved power supply/wall wart (so everything inside the product is
> 5volts)

UL tests for more than just AC line volatge and isolation.  However,
depending on the market, you may not legally need UL testing to sell
your product in a given area.  For instance you can buy computer power
supplies that are not UL listed, and are much less expensive than the
UL listed power supplies.  Many retailers don't carry them though,
since they are exposed to some liability for resulting losses.

> but as I understand the FCC stuff, if you have a
> clock/oscillator over 9Khz, it needs to be certified.

If the device is battery powered and less than 1MHz or so (it's not
exactly 1MHz, read the regulations) then FCC certification is not
required.  Since it connects to a powerline, however (which is really
a poor antenna, as far as the FCC is concerned) then it's more strict.
Check out the regulations - they are actually pretty clear as to what
is excepted under Class B.

> 1) Are there any exemptions for small batch/lot products (these will
> mostly be "art works")?

There are exceptions for prototypes, test setups, and some forms of
kits.  Other than that a one-off that's sold to be used as-is is just
as regulated as a device mass produced in the millions.

> 2) Did you have to do a lot of shielding?  The controller is a PIC
> running at 8Mhz doing a lot of RGB LED PWM stuff (http://www.rgbled.org).  I
> don't think it throws off that much RF, but...

Testing really doesn't start until you get into the higher
frequencies.  If nothing in your device runs at greater than 4MHz,
then it's pretty easy to pass.  Make sure that any high frequency
tracks on the PCB are short, without sharp corners.  Make sure you
have good bypass caps across all VDD/VSS pins on all chips.  If you
have long wires (going to LEDs or other equipment) then try to keep
signals as slow as possible, perhaps consider sending sine waves
instead of square waves (use a cap/coil to filter out the higher
frequency components of the square wave) as long as the signal is
still meaningful on the other end of the line.  Any wires going into
or out of the box need to be filtered to some degree.  The box needn't
be shielded if the circuit itself is radiating very little energy. In
this case I don't think shielding would help or hinder the project.
But I don't know the particulars, so don't take my word for it...

> 3) Did you use one of those FCC certification shops?

My expereince is with TUV.  They do a good job, fairly easy and straightforward.

> 4) Do you have a rough idea how much the process costs?  This could be
> for class B certification (residential)?

If you're _only_ doing very simple FCC B testing then it could be as
little as $2,000.  Shop around - they are like any other service
industry and they'll be more than happy to answer all your questions
about certification.

> 5) Did you also get a CA certification too?  If so, was the pricing similar?

Generally once you test for FCC, you can get some other emissions
certificates for an extra $500 report fee, as long as the emissions
testing itself isn't more complex.  CE emissions and immunity are much
more rigorous than FCC, so if you go for CE you cover a lot more
ground and may be able to get more certifications for the simple
"additional report fee"

> Gerry Duprey
> Ann Arbor, MI 48103

Hah!  I never noticed.  We'll have to go to lunch sometime - I work on
main street in downtown A2.  Let me know if you have any other
questions about certification.

-Adam

2006\09\13@115152 by Paul Hutchinson

picon face
> -----Original Message-----
> From: spam_OUTpiclist-bouncesTakeThisOuTspammit.edu On Behalf Of Gerry Duprey
> Sent: Wednesday, September 13, 2006 10:25 AM
>
> Howdy,
>
> I have a person/client who wants to use one of my boards in a commercial
> product.  They are concerned about getting FCC certification for it and
> I really have no experience with it.
>
<snip>
>
> If anyone has gone through this process, I'd really like to hear some
> details about it.  Particular questions:

I've been involved with FCC approvals since 1985.

> 1) Are there any exemptions for small batch/lot products (these will
> mostly be "art works")?
>

No, quantity sold does not enter into the regulations.

> 2) Did you have to do a lot of shielding?  The controller is a PIC

Rarely, shielding should only be a last resort.

> running at 8Mhz doing a lot of RGB LED PWM stuff (http://www.rgbled.org).  I
> don't think it throws off that much RF, but...

Driving LED's with high power can be a significant source of RF, I've worked
with two products like that and both required extra testing and some
PCB/circuit redesign to meet standards. I recommend getting into a
laboratory early for a less expensive ($200 to $8000) quick scan to identify
problems before scheduling a final test.

> 3) Did you use one of those FCC certification shops?

Always, the regulations can be difficult to understand and most importantly
if a problem does occur having experts to help will save many hours/$$.

> 4) Do you have a rough idea how much the process costs?  This could be
> for class B certification (residential)?

It depends on many factors especially whether the device needs verification
or actual certification. In general $700 to $1500 if the device is
verification only and meets requirements without modification. For
certification, with a design that takes extensive modifications over six lab
days to meet requirements, it can cost upwards of $10,000.

IME, the most important factor for choosing a laboratory for FCC testing is
location. If you can go to help with the testing it can reduce the costs on
the worst case by half.

My favorite lab, Curtis-Straus, has a number of documents on their web site
that will help you with preparation for testing.
http://www.curtis-straus.com/piqs.html

>
> 5) Did you also get a CA certification too?  If so, was the
> pricing similar?

Assuming CA = Canada, I've never gone ahead with getting Canadian
certification but, have been quoted roughly a 50% increase maximum for
adding Canadian certification.

Hope this helps,
Paul

> Of course, while all the questions are important, #4/5 are really what
> are going to sink/float the product.
>
> Thanks again,
>
> Gerry

2006\09\13@120639 by Bob Axtell

face picon face
Gerry Duprey wrote:
> Howdy,
>
> I have a person/client who wants to use one of my boards in a commercial
> product.  They are concerned about getting FCC certification for it and
> I really have no experience with it.
>
> I think they can "dodge" the UL stuff by simply using an external UL
> approved power supply/wall wart (so everything inside the product is
> 5volts), but as I understand the FCC stuff, if you have a
> clock/oscillator over 9Khz, it needs to be certified.
>  
Mostly correct; there are a few exceptions. Be careful with Chinese
wallwarts; they are famous
for stealing certs from a competitor. UL maintains a list of complying
manufacturers. It is up to
you and your client to do the due diligence work here. "I didn't know"
won't pass muster here.
A very well-known client of mine had to destroy thousands of
bogus-certified wallwarts when
we investigated a few bad supplies.
> If anyone has gone through this process, I'd really like to hear some
> details about it.  Particular questions:
>  
Many times.
> 1) Are there any exemptions for small batch/lot products (these will
> mostly be "art works")?
>  
basically, no, unless they are special commercial products, such as
diathermy machines, etc. My
robotics client was exempt since the product was used in an underground
sewer (unable to radiate
in normal operation). If you are not selling the product, some protyping
is OK; unless it interferes
with airplane comm services.
> 2) Did you have to do a lot of shielding?  The controller is a PIC
> running at 8Mhz doing a lot of RGB LED PWM stuff (http://www.rgbled.org).  I
> don't think it throws off that much RF, but...
>  
Actually, PMW DOES throw off RF. Any signal that changes state rapidly
on a long conductor throws off RF when
the sharp edge occurs (a simplistic explanation, but it covers it). The
PCB layout is the most important factor; if
you do much PWM, you will probably need to have a 4-layer PCB to reduce
radiation (at least one layer is GND).
The 4-layer PCB usually drops radiation by 50% immediately... think
about it, you'll see why. If you can internalize
the signals to an inner layer, you might crush the radiation entirely. I
normally make 4-layer PCBs, but only a GND
layer is used; I place noisy signals on the internal signal layer, then
provide a copper pour on top (or bottom, as
needed) to encase the offender.

I have never had to shield anything to pass FCC part 15 or 68, but I
HAVE had to shield stuff to stop external
noise from getting into the product. Strong nearby radios can swamp
uP-based prducts, and metal boxes don't
always keep noise out. Police-band long-range radios are especially bad.
> 3) Did you use one of those FCC certification shops?
>  
Yes, my favorite is Sid Sanders' shop in Tampa, FL: Timco Engineering.
They are extremely competent and will
tell you what to do to get your product into compliance, yet their fees
are very reasonable. Call 1-888-472-2424.
> 4) Do you have a rough idea how much the process costs?  This could be
> for class B certification (residential)?
>  
Its been 5 years since I rode herd on it, but then it was about $1800
for the complete service on a still camera
accessory. That project passed immediately, took less than a month.
> 5) Did you also get a CA certification too?  If so, was the pricing similar?
>  
CE Norms? or the Canadian Standards Tests? No, but later CE Norms tests
passed, I was told. BTW, CE Norms is NOT
exactly the same as FCC Part 15 or 68. Canadian Standards is literally
identical to FCC, and my remembrance was that CA
was included with FCC. Canada and the US keep their standards VERY
close, so that free trade is easy to maintain.

> Of course, while all the questions are important, #4/5 are really what
> are going to sink/float the product.
>  
Your client is right to do this; the FCC has become quite militant in
enforcing radiation testing. I was at a Comdex show
one year when the FCC seized untested PC accessories and arrested a few
executives on the floor. I think it was 1980,
if I recall. Surprised a lotta folks.

- - -

You can save a LOT of grief by doing some testing yourself. Get an
allband radio or a "bug sniffer" and try to pick up
your product with it. I paid less than $100 for a nice "bug sniffer"
from a kit company a few years ago, and it does a
great job of detecting radiation. If you can pick up a strong carrier
more than 4 meters away, you probably have a
problem.


--Bob Axtell

2006\09\13@121918 by Tamas Rudnai

face picon face
> If the device is battery powered and less than 1MHz or so (it's not
> exactly 1MHz, read the regulations) then FCC certification is not
> required.

OK, my PIC has a 4MHz internal osc. I suppose for that Microchip has to pass
the FCC, not me -- is that true? Now, if I use clock out, I have to pass the
test as well?

Tamas

2006\09\13@123223 by M. Adam Davis

face picon face
No, component parts sold for incorporation into another product do not
need to be FCC tested if they are not intended to be used alone.

The end product itself has to be tested.

As far as I understand, if any frequency above 1MHz or so is generated
by a device, regardless of the source and use, then it still requires
certification.

So Microchip is not responsible for testing their product - as sold it
does nothing.  Your product does need certification - even if the
clock never leaves the chip package it is still present in the device.

But don't rely on my words - the FCC regulations are not difficult to
understand, you just have to read a lot and figure out what doesn't
apply so you can focus on what does apply.

-Adam

On 9/13/06, Tamas Rudnai <.....tamas.rudnaiKILLspamspam@spam@gmail.com> wrote:
> > If the device is battery powered and less than 1MHz or so (it's not
> > exactly 1MHz, read the regulations) then FCC certification is not
> > required.
>
> OK, my PIC has a 4MHz internal osc. I suppose for that Microchip has to pass
> the FCC, not me -- is that true? Now, if I use clock out, I have to pass the
> test as well?
>
> Tamas
> -

2006\09\13@125106 by Harold Hallikainen

face
flavicon
face
For the FCC rules on unintentional radiators, see sections 15.101 through
15.123 at http://www.hallikainen.com/FccRules/2006/15/ .

Your test lab may be able to design a test series that demonstrates
compliance with US FCC, Canadian, and European standards.

Good luck!

Harold



--
FCC Rules Updated Daily at http://www.hallikainen.com - Advertising
opportunities available!

2006\09\13@133334 by Bob Axtell

face picon face
Tamas Rudnai wrote:
>> If the device is battery powered and less than 1MHz or so (it's not
>> exactly 1MHz, read the regulations) then FCC certification is not
>> required.
>>    
>
> OK, my PIC has a 4MHz internal osc. I suppose for that Microchip has to pass
> the FCC, not me -- is that true? Now, if I use clock out, I have to pass the
> test as well?
>
> Tamas
>  
No, Microchip doesn't have to test anything. You must get your product
tested and obtain a certification.

FCC part 15 specs are actually an easy read. Gargoyle for it.


--Bob

2006\09\13@134356 by Paul Hutchinson

picon face
> -----Original Message-----
> From: piclist-bouncesspamKILLspammit.edu On Behalf Of M. Adam Davis
> Sent: Wednesday, September 13, 2006 12:32 PM
>
> No, component parts sold for incorporation into another product do not
> need to be FCC tested if they are not intended to be used alone.
>
> The end product itself has to be tested.
>
> As far as I understand, if any frequency above 1MHz or so is generated
> by a device, regardless of the source and use, then it still requires
> certification.

The frequency limit is 9kHz and it applies to any device that uses digital
techniques. A 555 timer IC running at 9kHz is covered by the regulations.

For exclusively battery operated devices (absolutely no way to power from AC
without hacking the circuit) the limit is 1.705MHz.

Paul

{Quote hidden}

2006\09\13@162945 by David VanHorn

picon face
On 9/13/06, Tamas Rudnai <.....tamas.rudnaiKILLspamspam.....gmail.com> wrote:
>
> > If the device is battery powered and less than 1MHz or so (it's not
> > exactly 1MHz, read the regulations) then FCC certification is not
> > required.
>
> OK, my PIC has a 4MHz internal osc. I suppose for that Microchip has to
> pass
> the FCC, not me -- is that true? Now, if I use clock out, I have to pass
> the
> test as well?


Nope, microchip isn't selling finished goods.
You can do whatever you like with "components".
If it's intended to be used by Joe Public, then that's when you need the
cert.

2006\09\13@163130 by David VanHorn

picon face
>
>
> Actually, PMW DOES throw off RF. Any signal that changes state rapidly
> on a long conductor throws off RF when
> the sharp edge occurs (a simplistic explanation, but it covers it). The
> PCB layout is the most important factor;



You can PWM into an LC filter, which then provides smoothed DC to the LED
strings.
The high speed edges then are confined to a couple square inches of PCB at
most, and that makes life a LOT easier.

More... (looser matching)
- Last day of these posts
- In 2006 , 2007 only
- Today
- New search...