|Thread Links||Date Links|
|Thread Prev||Thread Next||Thread Index||Date Prev||Date Next||Date Index|
I think what you saying is generally reasonable.
We should test radiated emissions with both CAT-3 and CAT-5.
Both discovery and power excite the pair to pair mode, but there is always some coupling into the true common mode,
which is what really radiates.
Also, like you say, it is beneficial to test the following loads: identity network, open, and short.
As far as doing conducted on the data cables, I'm not sure that EN 55022 calls for it.
In any case testing conducted emissions generally involves using a LISN on the AC power line input.
How is it performed for a data cable?
Conducted Immunity uses a common cable clamp instead.
At least the conducted tests cover frequencies below 30 MHz, which is important.
By the way, for the diode discovery radiated emissions test, did you notice that the discovery spectrum
was only about 2 dB above the noise floor of the chamber and test equipment?
I'm not trying to imply that the risk is low or anything.
From: R karam [SMTP:rkaram@xxxxxxxxx]
Sent: Friday, September 22, 2000 1:39 PM
To: Brooks, Rick [SC5:321:EXCH]; stds-802-3-pwrviamdi@xxxxxxxx
Subject: Re: FW: Composite list of tests for DTE detection
Unless I am missing something, what is reasonable here is just the detection
over a CAT-3 Cable, on their own. I am going to assume that one has the supply
under control, the Switcher's noise under control in the DTE, I would be happy
knowing that the spectrum of the detection pulses will not cause problems.
my take on this is should be another scan for the EMI person, As if he was
for CLASS B - business as usual.
10/100 Data now that is a problem we all know how to solve, By now.
NOTE: Another data point here is Detection Pulses going into a Box that is OFF
and Grounded (just a quick scan) the idea is that the load seen is
the ID network, but something else with a possible ground loop back
initiating supply, where the loop of detection is infinite since the
DTE END is
OFF and can not be ID-ed.
Also a note on some of the "Warranty" - like requests, there will be risk at
end of the day no matter which approach we take, and unless real systems are
we may not uncover all the issues at this point. Let's GO! (give and take a
Also in Reference to the CISPR 22 Conducted, now that is a Must I think also.
hope- I have reduced the test time to a day or Two worst case.
I am hoping to test CISPR 22 Conducted soon, but would like a confirmation from
someone else on my test results since this is new business...
No more cents for today, gotta hit the bank.
At 01:20 PM 9/22/00 -0700, Rick Brooks wrote:
> I was reviewing this list of tests.
> Now, I have some questions regarding part 5, EMI scan:
> Is this only radiated emissions, or do you want susceptibility also?
> What antenna orientation do you want?
> What turntable orientation do you want?
> What antenna height do you want?
> Do you always want the data maximized?
> Test via the setup in the EN 55022 spec?
> Now I'm looking for volunteers who are willing to do these 42 EMI scans that
> have been requested in section 5.
> Once you have all the equipment, this might only be several days in an
> anechoic chamber to collect the all of the data.
> First, you have to put together a full communication system where you are
> measuring only what you want to,
> and not Smartbits boxes, for instance.
> I think that this is a lot of testing!
> What is it going to buy us really?
> One suggestion: go with one cable length, with physical routing suggested by
> EN 55022.
> Also, test only discovery, maybe add 100 meg traffic.
> just my opinions...
> - Rick
> -----Original Message-----
> From: Nick Stapleton [SMTP:Nick_Stapleton@xxxxxxxxxxxx]
> Sent: Thursday, September 21, 2000 3:42 AM
> To: stds-802-3-pwrviamdi@xxxxxxxx
> Subject: Composite list of tests for DTE detection
> The compiled list of tests for DTE detection mechanisms that Mike McCormack
> actioned to collate has been placed at the URL:-
> It has been placed in the private area as it is in IEEE document format with
> IEEE copyright notice. We are not certain it really needs to be here, but
> is the usual rule for IEEE copyrighted material, so will leave it there for
> Please bear in mind that the document is a collation of different tests posed
> various people during the course of our work both in committee, and on the
> reflector. If you have any issue with any of these tests, the proper place to
> take it up is on the reflector, in order that the question reaches whoever
> the test.