AVR-Chat

I've been developing an embedded device for some time now, and have
decided to manufacture a batch for sale to the public.  Although I
have tested it as thoroughly as I possibly can, I would like to do
more before offering it for sale (an ounce of debugging is worth a
pound of bug fixing...).

Is there a standard testing procedure that should be used to test for
electrical and/or mechanical robustness (similar to testing for IP67
etc...)?

Obviously software testing is left up to me, which is fine; been
running 10,000 hrs non-stop on the prototypes so far.

-Don

You need to be aware of FCC testing rules! They are spelled
out in FCC part 15 and I will provide a link when I can
look it up. Technically, you CANNOT sell a digital device
with a "clock" in it that runs over a certain frequency
without FCC certification. You can produce up to 200 as
demos but you cannot sell them - you must retain ownership!

As for quality testing, while FCC does not require it the
way CE does, I would try some basic "ESD" testing,
especially if this is going into sites where human "static
zaps" can occur. For a real quick test, you can use one of
those "BBQ lighters" that makes a spark. Spark to the
chassis and see whether or not it resets or does other
strange things. That will save you lots of grief from
things not operating correctly in the field. 

Jim


On Tue, 15 Aug 2006 09:19:58 -0000
 "Don Kirby" <psyclopedia@yahoo.com> wrote:
> I've been developing an embedded device for some time
> now, and have
> decided to manufacture a batch for sale to the public.
>  Although I
> have tested it as thoroughly as I possibly can, I would
> like to do
> more before offering it for sale (an ounce of debugging
> is worth a
> pound of bug fixing...).
> 
> Is there a standard testing procedure that should be used
> to test for
> electrical and/or mechanical robustness (similar to
> testing for IP67
> etc...)?
> 
> Obviously software testing is left up to me, which is
> fine; been
> running 10,000 hrs non-stop on the prototypes so far.
> 
> -Don
> 
> 
> 

---------------------------------------------------------------
The Think Different Store
http://www.thinkdifferentstore.com/
For All Your Mac Gear
---------------------------------------------------------------

Thanks Jim, I knew there was some kind of legal testing protocol to
follow.  I'm checking out the FCC part 15 right now.

-Don



--- In AVR-Chat@yahoogroups.com, "Jim Wagner" <jim_d_wagner@...> wrote:

>
> You need to be aware of FCC testing rules! They are spelled
> out in FCC part 15 and I will provide a link when I can
> look it up. Technically, you CANNOT sell a digital device
> with a "clock" in it that runs over a certain frequency
> without FCC certification. You can produce up to 200 as
> demos but you cannot sell them - you must retain ownership!
> 
> As for quality testing, while FCC does not require it the
> way CE does, I would try some basic "ESD" testing,
> especially if this is going into sites where human "static
> zaps" can occur. For a real quick test, you can use one of
> those "BBQ lighters" that makes a spark. Spark to the
> chassis and see whether or not it resets or does other
> strange things. That will save you lots of grief from
> things not operating correctly in the field. 
> 
> Jim
> 
> 
> On Tue, 15 Aug 2006 09:19:58 -0000
>  "Don Kirby" <psyclopedia@...> wrote:
> > I've been developing an embedded device for some time
> > now, and have
> > decided to manufacture a batch for sale to the public.
> >  Although I
> > have tested it as thoroughly as I possibly can, I would
> > like to do
> > more before offering it for sale (an ounce of debugging
> > is worth a
> > pound of bug fixing...).
> > 
> > Is there a standard testing procedure that should be used
> > to test for
> > electrical and/or mechanical robustness (similar to
> > testing for IP67
> > etc...)?
> > 
> > Obviously software testing is left up to me, which is
> > fine; been
> > running 10,000 hrs non-stop on the prototypes so far.
> > 
> > -Don
> > 
> > 
> > 
> 
> ---------------------------------------------------------------
> The Think Different Store
> http://www.thinkdifferentstore.com/
> For All Your Mac Gear
> ---------------------------------------------------------------
>

You want 47cfr15.13 (part 15, section 13).

Unfortunately, testing costs pretty big bucks ($1K to $2K).
This is not something you can do yourself; it has to be
done by an FCC certified lab. Depending on where you are in
the U.S., I can offer at least one recommendtion. I can
also provide some sense of what is involved and how
to maximize the odds of passing the first time.

Jim

On Tue, 15 Aug 2006 20:28:32 -0000
 "Don Kirby" <psyclopedia@yahoo.com> wrote:
> Thanks Jim, I knew there was some kind of legal testing
> protocol to
> follow.  I'm checking out the FCC part 15 right now.
> 
> -Don
> 
> 
> 
> --- In AVR-Chat@yahoogroups.com, "Jim Wagner"
> <jim_d_wagner@...> wrote:
> >
> > You need to be aware of FCC testing rules! They are
> spelled
> > out in FCC part 15 and I will provide a link when I can
> > look it up. Technically, you CANNOT sell a digital
> device
> > with a "clock" in it that runs over a certain frequency
> > without FCC certification. You can produce up to 200 as
> > demos but you cannot sell them - you must retain
> ownership!
> > 
> > As for quality testing, while FCC does not require it
> the
> > way CE does, I would try some basic "ESD" testing,
> > especially if this is going into sites where human
> "static
> > zaps" can occur. For a real quick test, you can use one
> of
> > those "BBQ lighters" that makes a spark. Spark to the
> > chassis and see whether or not it resets or does other
> > strange things. That will save you lots of grief from
> > things not operating correctly in the field. 
> > 
> > Jim
> > 
> > 
> > On Tue, 15 Aug 2006 09:19:58 -0000
> >  "Don Kirby" <psyclopedia@...> wrote:
> > > I've been developing an embedded device for some time
> > > now, and have
> > > decided to manufacture a batch for sale to the
> public.
> > >  Although I
> > > have tested it as thoroughly as I possibly can, I
> would
> > > like to do
> > > more before offering it for sale (an ounce of
> debugging
> > > is worth a
> > > pound of bug fixing...).
> > > 
> > > Is there a standard testing procedure that should be
> used
> > > to test for
> > > electrical and/or mechanical robustness (similar to
> > > testing for IP67
> > > etc...)?
> > > 
> > > Obviously software testing is left up to me, which is
> > > fine; been
> > > running 10,000 hrs non-stop on the prototypes so far.
> > > 
> > > -Don
> > > 
> > > 
> > > 
> > 
> >
>
---------------------------------------------------------------
> > The Think Different Store
> > http://www.thinkdifferentstore.com/
> > For All Your Mac Gear
> >
>
---------------------------------------------------------------
> >
> 
> 
> 
> 
> 

---------------------------------------------------------------
The Think Different Store
http://www.thinkdifferentstore.com/
For All Your Mac Gear
---------------------------------------------------------------

Have you spoken with the FCC recently Jim?  I spoke with them a couple 
of years ago and with today's electronics being so low power they allow 
some producers and manufacturers to do self certification...or so I was 
told.  It seems to depend on the application...i.e. if it's connected to 
a telephone line, can cause interference with other equipment, the 
product's environment, and so forth.

The representative I spoke with said the gotcha is that if you certify 
that it meets their requirements and your product does not, open your 
wallet...or perhaps the better thing to do would be to just give them 
your wallet.

The best thing for Don to do would be to contact the FCC directly.  That 
way he doesn't depend on interpretation.  Testing labs, UL and ETL, are 
certainly going to tell you that everything needs to be tested and 
certified by them.  That's their business and means money in their pockets.


REB


Jim Wagner wrote:

>You want 47cfr15.13 (part 15, section 13).
>
>Unfortunately, testing costs pretty big bucks ($1K to $2K).
>This is not something you can do yourself; it has to be
>done by an FCC certified lab. Depending on where you are in
>the U.S., I can offer at least one recommendtion. I can
>also provide some sense of what is involved and how
>to maximize the odds of passing the first time.
>
>Jim
>
>On Tue, 15 Aug 2006 20:28:32 -0000
> "Don Kirby" <psyclopedia@yahoo.com> wrote:
>  
>
>>Thanks Jim, I knew there was some kind of legal testing
>>protocol to
>>follow.  I'm checking out the FCC part 15 right now.
>>
>>-Don
>>
>>
>>
>>--- In AVR-Chat@yahoogroups.com, "Jim Wagner"
>><jim_d_wagner@...> wrote:
>>    
>>
>>>You need to be aware of FCC testing rules! They are
>>>      
>>>
>>spelled
>>    
>>
>>>out in FCC part 15 and I will provide a link when I can
>>>look it up. Technically, you CANNOT sell a digital
>>>      
>>>
>>device
>>    
>>
>>>with a "clock" in it that runs over a certain frequency
>>>without FCC certification. You can produce up to 200 as
>>>demos but you cannot sell them - you must retain
>>>      
>>>
>>ownership!
>>    
>>
>>>As for quality testing, while FCC does not require it
>>>      
>>>
>>the
>>    
>>
>>>way CE does, I would try some basic "ESD" testing,
>>>especially if this is going into sites where human
>>>      
>>>
>>"static
>>    
>>
>>>zaps" can occur. For a real quick test, you can use one
>>>      
>>>
>>of
>>    
>>
>>>those "BBQ lighters" that makes a spark. Spark to the
>>>chassis and see whether or not it resets or does other
>>>strange things. That will save you lots of grief from
>>>things not operating correctly in the field. 
>>>
>>>Jim
>>>
>>>
>>>On Tue, 15 Aug 2006 09:19:58 -0000
>>> "Don Kirby" <psyclopedia@...> wrote:
>>>      
>>>
>>>>I've been developing an embedded device for some time
>>>>now, and have
>>>>
>  
>


[Non-text portions of this message have been removed]

From what I'm reading in part 15, only devices drawing less than 6nW
are exempt, although that might actually be a typo (6*m*W).  There are
other exempt devices, which mine might actually be one of.  I;m not
much for the legaleze of government documents, but I'm muddling
through it.  I can do the required testing in house, and there's
nothing that I've found so far that says a certified testing facility
be used.

FCC verification/certification and IP notwithstanding, can anyone
suggest any other test methods that I should impliment before shipping
devices to John Q. Public?  

-Don

--- In AVR-Chat@yahoogroups.com, "Roy E. Burrage" <RBurrage@...> wrote:
>
> Have you spoken with the FCC recently Jim?  I spoke with them a couple 
> of years ago and with today's electronics being so low power they allow 
> some producers and manufacturers to do self certification...or so I was 
> told.  It seems to depend on the application...i.e. if it's
connected to 
> a telephone line, can cause interference with other equipment, the 
> product's environment, and so forth.
> 
> The representative I spoke with said the gotcha is that if you certify 
> that it meets their requirements and your product does not, open your 
> wallet...or perhaps the better thing to do would be to just give them 
> your wallet.
> 
> The best thing for Don to do would be to contact the FCC directly. 
That 
> way he doesn't depend on interpretation.  Testing labs, UL and ETL, are 
> certainly going to tell you that everything needs to be tested and 
> certified by them.  That's their business and means money in their
pockets.

> 
> 
> REB
> 
> 
> Jim Wagner wrote:
> 
> >You want 47cfr15.13 (part 15, section 13).
> >
> >Unfortunately, testing costs pretty big bucks ($1K to $2K).
> >This is not something you can do yourself; it has to be
> >done by an FCC certified lab. Depending on where you are in
> >the U.S., I can offer at least one recommendtion. I can
> >also provide some sense of what is involved and how
> >to maximize the odds of passing the first time.
> >
> >Jim
> >
> >On Tue, 15 Aug 2006 20:28:32 -0000
> > "Don Kirby" <psyclopedia@...> wrote:
> >  
> >
> >>Thanks Jim, I knew there was some kind of legal testing
> >>protocol to
> >>follow.  I'm checking out the FCC part 15 right now.
> >>
> >>-Don
> >>
> >>
> >>
> >>--- In AVR-Chat@yahoogroups.com, "Jim Wagner"
> >><jim_d_wagner@> wrote:
> >>    
> >>
> >>>You need to be aware of FCC testing rules! They are
> >>>      
> >>>
> >>spelled
> >>    
> >>
> >>>out in FCC part 15 and I will provide a link when I can
> >>>look it up. Technically, you CANNOT sell a digital
> >>>      
> >>>
> >>device
> >>    
> >>
> >>>with a "clock" in it that runs over a certain frequency
> >>>without FCC certification. You can produce up to 200 as
> >>>demos but you cannot sell them - you must retain
> >>>      
> >>>
> >>ownership!
> >>    
> >>
> >>>As for quality testing, while FCC does not require it
> >>>      
> >>>
> >>the
> >>    
> >>
> >>>way CE does, I would try some basic "ESD" testing,
> >>>especially if this is going into sites where human
> >>>      
> >>>
> >>"static
> >>    
> >>
> >>>zaps" can occur. For a real quick test, you can use one
> >>>      
> >>>
> >>of
> >>    
> >>
> >>>those "BBQ lighters" that makes a spark. Spark to the
> >>>chassis and see whether or not it resets or does other
> >>>strange things. That will save you lots of grief from
> >>>things not operating correctly in the field. 
> >>>
> >>>Jim
> >>>
> >>>
> >>>On Tue, 15 Aug 2006 09:19:58 -0000
> >>> "Don Kirby" <psyclopedia@> wrote:
> >>>      
> >>>
> >>>>I've been developing an embedded device for some time
> >>>>now, and have
> >>>>
> >  
> >
> 
> 
> [Non-text portions of this message have been removed]
>

I can only say that I have had products with basic 8051
micros that failed.
I have also had ones with switching power supplies fail.
 Power consumption is not a good indicator.

I would NOT use UL as a FCC testing lab. ETL I don;t know
about. There is a group of test houses in the Pacific NW
and N. California that I do trust. I have worked with them
for over 15 years on both FCC and CE testing.

Jim


On Tue, 15 Aug 2006 20:00:32 -0400
 "Roy E. Burrage" <RBurrage@bellsouth.net> wrote:
> Have you spoken with the FCC recently Jim?  I spoke with
> them a couple 
> of years ago and with today's electronics being so low
> power they allow 
> some producers and manufacturers to do self
> certification...or so I was 
> told.  It seems to depend on the application...i.e. if
> it's connected to 
> a telephone line, can cause interference with other
> equipment, the 
> product's environment, and so forth.
> 
> The representative I spoke with said the gotcha is that
> if you certify 
> that it meets their requirements and your product does
> not, open your 
> wallet...or perhaps the better thing to do would be to
> just give them 
> your wallet.
> 
> The best thing for Don to do would be to contact the FCC
> directly.  That 
> way he doesn't depend on interpretation.  Testing labs,
> UL and ETL, are 
> certainly going to tell you that everything needs to be
> tested and 
> certified by them.  That's their business and means money
> in their pockets.
> 
> 
> REB
> 
> 
> Jim Wagner wrote:
> 
> >You want 47cfr15.13 (part 15, section 13).
> >
> >Unfortunately, testing costs pretty big bucks ($1K to
> $2K).
> >This is not something you can do yourself; it has to be
> >done by an FCC certified lab. Depending on where you are
> in
> >the U.S., I can offer at least one recommendtion. I can
> >also provide some sense of what is involved and how
> >to maximize the odds of passing the first time.
> >
> >Jim
> >
> >On Tue, 15 Aug 2006 20:28:32 -0000
> > "Don Kirby" <psyclopedia@yahoo.com> wrote:
> >  
> >
> >>Thanks Jim, I knew there was some kind of legal testing
> >>protocol to
> >>follow.  I'm checking out the FCC part 15 right now.
> >>
> >>-Don
> >>
> >>
> >>
> >>--- In AVR-Chat@yahoogroups.com, "Jim Wagner"
> >><jim_d_wagner@...> wrote:
> >>    
> >>
> >>>You need to be aware of FCC testing rules! They are
> >>>      
> >>>
> >>spelled
> >>    
> >>
> >>>out in FCC part 15 and I will provide a link when I
> can
> >>>look it up. Technically, you CANNOT sell a digital
> >>>      
> >>>
> >>device
> >>    
> >>
> >>>with a "clock" in it that runs over a certain
> frequency
> >>>without FCC certification. You can produce up to 200
> as
> >>>demos but you cannot sell them - you must retain
> >>>      
> >>>
> >>ownership!
> >>    
> >>
> >>>As for quality testing, while FCC does not require it
> >>>      
> >>>
> >>the
> >>    
> >>
> >>>way CE does, I would try some basic "ESD" testing,
> >>>especially if this is going into sites where human
> >>>      
> >>>
> >>"static
> >>    
> >>
> >>>zaps" can occur. For a real quick test, you can use
> one
> >>>      
> >>>
> >>of
> >>    
> >>
> >>>those "BBQ lighters" that makes a spark. Spark to the
> >>>chassis and see whether or not it resets or does other
> >>>strange things. That will save you lots of grief from
> >>>things not operating correctly in the field. 
> >>>
> >>>Jim
> >>>
> >>>
> >>>On Tue, 15 Aug 2006 09:19:58 -0000
> >>> "Don Kirby" <psyclopedia@...> wrote:
> >>>      
> >>>
> >>>>I've been developing an embedded device for some time
> >>>>now, and have
> >>>>
> >  
> >
> 
> 
> [Non-text portions of this message have been removed]
> 

---------------------------------------------------------------
The Think Different Store
http://www.thinkdifferentstore.com/
For All Your Mac Gear
---------------------------------------------------------------