Group,
Here is a reply from Monrad Monsen
For those that went to the 98 Denver EMC Conference
Monrad presented a Measurement Uncertainty Workshop.
He is also a Certified Quality Engineer and teaches a
stats course. Please direct any furthur questions directly to him.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
Posted for Monrad Monsen
Mr. Selva,
You have asked an interesting question. I am not aware of too many
companies that have time (and money) to perform testing required for the
80/80 calculations. Therefore, I can only answer your question from an
academic perspective as opposed to industry trends.
> How do you perform this calculation ?
Assuming you are measuring the same worst-case maximized frequency, you may
do the 80/80 calculations based on the margin from the limit with the same
statistical accuracy as if you had used the measured emission levels. The
arithmetic mean (average) of the margin is calculated with the normal method
giving a passing negative average margin from the limit, and the standard
deviation is also calculated using the same formula as listed in the
standard. Then the compliance is judged from the following relationship
gained from the standard:
(x-bar) + (k)(s) <= 0
(x-bar) + ks * 0
Where "x-bar" is the arithmetic mean (average) which should be negative, "s"
is the standard deviation, "k" is from the table in CISPR 22, and "L" is 0
since you are calculating the statistics for margins where negative margin
meets the standard. Of course, if the arithmetic mean (average) is actually
over the limit (i.e. positive margin), then there is no need for any further
calculation since it fails.
In fact, I would contend that performing this 80/80 calculation using the
margins actually makes the calculation more meaningful to the customer since
he really wants to know by how much does his product pass or how much
comfort should he have regarding emissions for his product.
> On which value do you perform it (margin, level, azimuth, height of
antenna, ...) ?
The 80/80 calculation can only be calculated on the measured result (either
a measured level or a margin) as compared to a standard. If you calculate
the 80/80 using margins, then the standard is L=0 (i.e. the standard is the
same as a zero margin). To take a measurement, you must maximize the
emission using the same procedure for each sample including maximizing
azimuth, the height of the antenna, and the cable layout. It is meaningless
to do statistics on the azimuth or height of antenna since these are merely
a part of the measurement process, and there is no agency standard regarding
these values.
I hope this helps.
Monrad L. Monsen
Senior EMC Engineer
Product Compliance
303.673.2438 phone
303.673.2431 fax
StorageTek
INFORMATION made POWERFUL
> -----Original Message-----
> From: Grasso, Charles (Chaz)
> Sent: Thursday, August 26, 1999 6:23 PM
> To: SAFETY-EMC (Adresse de messagerie); 'Pierre Selva'
> Cc: Monsen, Monrad L
> Subject: RE: 80/80 calculation rules for CISPR22/EN55022
>
> Hello,
> I have taken the liberty of forwarding this to
> Mr Monsen of StorageTek. He presented a
> Measurement Uncertainty course at the Denver
> 1998 EMC Symposium & is a stats instructor.
>
> Thank you
> Charles Grasso
> StorageTek
> 2270 Sth 88th Street
> Louisville CO 80027
> Tel: (303)673-2908
> Fax(303)661-7115
>
>
> ----------
> From: Pierre Selva[SMTP:[email protected]]
> Sent: Thursday, August 26, 1999 8:12 AM
> To: SAFETY-EMC (Adresse de messagerie)
> Subject: 80/80 calculation rules for CISPR22/EN55022
>
>
> Dear colleagues,
>
> My subject of interest is the statistical rule explain in chapter
> 9.2 of EN55022 (also called the 80/80). Reading the standard, I don't know
> on which value I have to apply the rule.
> In fact, I usually apply the rule on each suspect frequency (about 5
> by product) and I compare the calculated LEVEL to the Limit, in dBµV.
>
> One of my customer is asking me to make this calculation on the
> MARGIN. For each product, I have to take the worst margin, and I make the
> calculation for the x products I have. The calcul gives a result which has
> to be compare to 0 (zero). The resulting margin has to be less than 0.
>
> How do you perform this calculation ?
> On which value do you perform it (margin, level, azimuth, height of
> antenna, ...) ?
> What are your own experience with your products or your customer ?
>
> In advance, I thank you a lot for your answer, which, I'm sure, will
> be of great value (80/80 calculation, of course !!)
>
> Best regards,
> ================================================================
> Pierre Selva
> Laboratory responsible EMC and Safety laboratory
> SMEE Actions Mesures Ph : 33 4 76 65 76 50
> ZI des Blanchisseries Fx : 33 4 76 66 18 30
> 38500 VOIRON - France e-mail :
> [email protected]
> ================================================================
>
>
>
> ---------
> This message is coming from the emc-pstc discussion list.
> To cancel your subscription, send mail to [email protected]
> with the single line: "unsubscribe emc-pstc" (without the
> quotes). For help, send mail to [email protected],
> [email protected], [email protected], or
> [email protected] (the list administrators).
>
>
Thank you
Charles Grasso
StorageTek
2270 Sth 88th Street
Louisville CO 80027
Tel: (303)673-2908
Fax(303)661-7115
---------
This message is coming from the emc-pstc discussion list.
To cancel your subscription, send mail to [email protected]
with the single line: "unsubscribe emc-pstc" (without the
quotes). For help, send mail to [email protected],
[email protected], [email protected], or
[email protected] (the list administrators).
