[SI-LIST] Re: Ethernet emissions.

  • From: Ken Wyatt <ken@xxxxxxxxxxxxxxxx>
  • To: "bryan@xxxxxxxxxxxxxx" <bryan@xxxxxxxxxxxxxx>
  • Date: Fri, 5 Apr 2013 22:21:05 -0600

See comments below...

Kenneth Wyatt
Wyatt Technical Services LLC
Woodland Park, CO
ken@xxxxxxxxxxxxxxxx
www.emc-seminars.com
(Sent from my iPad)

On Apr 5, 2013, at 9:08 PM, Bryan Ackerly <bryan@xxxxxxxxxxxxxx> wrote:

> Keith. Thanks for the advice.
> Not sure what more i can do with the layout. The traces to the connector are 
> rather short (<10mm) and run as equal length differential pairs on an outer 
> layer to preserve trace to ground impedance. No other digital signal tracks 
> for the PHY run on the outer layers. The board is 10 layers with paired 
> supply and ground planes to maximize plane capacitance. The PHY runs off 
> these planes and the supplies and grounding are well stitched to them.

Sounds like the layout is OK, at least between the PHY and connector. Be sure 
no other high speed traces on the board cross any gaps in the return plane..
> 
> Close Inductive probing using a spectrum analyzer shows these parts are the 
> "hottest" on the board at these frequencies.

I would expect near field probing those lines would appear hot. Not sure that 
proves anything.

> Use of a shielded ethernet cable drops most of the emission levels by about 
> 10-15dB.

Now this is interesting. Is the Ethernet connector shielded as well, or not? Is 
it CAT 5e or CAT 6?

> But this is not a practical solution as we have no control over what the 
> customer may use as a cable.

True, but if you specify shielded cables in the user or installation manual and 
the customer chooses not to follow your guidance, you're covered. On the other 
hand, if the product could be compliant using unshielded cables, that would 
obviously be best.

> This also applies to ferrite clamps, which so far i have not tried.

This may be your quick fix, but the above answer would apply.
> 
> This seems to suggest that the signals are common mode out of the PHY.

Maybe you could explain more, as I don't quite understand your reasoning.

> I thought this would largely be removed by the magnetics in the connector. 
> The connector is well bonded to the board lower layer which is a continuous 
> ground.

Good.

> The case does NOT form a shield as this is not practical.

This will make fixing the product more of a challenge.

> Though i would have thought if connector flange grounding was important, the 
> shielded cable would not have improved things so much.

That's definitely puzzling me. One thing that may help is installing an image 
plane close-spaced under the circuit board similar to the ESD protection idea I 
wrote up for www.theconnectingedge.com. I'll try to find the exact link, but 
you should be able to search it out.
> 
> The test was done as a loop back on the same product so i am measuring only 
> its emissions, not those coming from any connected equipment.
> 
> Regards,
> Bryan Ackerly
> 
> 
> On 06/04/2013, at 11:48 AM, keithK EPD <epd2001usa@xxxxxxxxxxx> wrote:
> 
>> Layout - I've done this for 15 years..
>> 
>> keith
>> 
>> 
>>> Subject: [SI-LIST] Ethernet emissions.
>>> From: bryan@xxxxxxxxxxxxxx
>>> Date: Sat, 6 Apr 2013 10:39:45 +1100
>>> To: si-list@xxxxxxxxxxxxx
>>> 
>>> I have a question regarding a (10/100) ethernet implementation that failed 
>>> radiated EMC. The phy is an SMSC LAN-8720. Due to space requirements a 
>>> connector with integral magnetics is used.
>>> 
>>> The main problem seems to be emissions at 375, 500 and 625MHz being 
>>> radiated from the UTP cable. 
>>> 
>>> For an expeiment i added an LC lowpass filter on each line from the PHY 
>>> cutting off from 125MHz which seems to improve things a bit, especially at 
>>> the higher frequencies.
>>> 
>>> But i dont know what impact this will have on the ethernet signalling 
>>> itself.
>>> 
>>> Has anyone else done this to pass emissions? Is there a better way of doing 
>>> it?
>>> 
>>> Regards,
>>> Bryan Ackerly
>>> 
>>> ------------------------------------------------------------------
>>> To unsubscribe from si-list:
>>> si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
>>> 
>>> or to administer your membership from a web page, go to:
>>> //www.freelists.org/webpage/si-list
>>> 
>>> For help:
>>> si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
>>> 
>>> 
>>> List forum is accessible at:
>>> http://tech.groups.yahoo.com/group/si-list
>>> 
>>> List archives are viewable at: 
>>> //www.freelists.org/archives/si-list
>>> 
>>> Old (prior to June 6, 2001) list archives are viewable at:
>>> http://www.qsl.net/wb6tpu
> 
> ------------------------------------------------------------------
> To unsubscribe from si-list:
> si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
> 
> or to administer your membership from a web page, go to:
> //www.freelists.org/webpage/si-list
> 
> For help:
> si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
> 
> 
> List forum  is accessible at:
>               http://tech.groups.yahoo.com/group/si-list
> 
> List archives are viewable at:     
>        //www.freelists.org/archives/si-list
> 
> Old (prior to June 6, 2001) list archives are viewable at:
>        http://www.qsl.net/wb6tpu
> 
> 
------------------------------------------------------------------
To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field

or to administer your membership from a web page, go to:
//www.freelists.org/webpage/si-list

For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field


List forum  is accessible at:
               http://tech.groups.yahoo.com/group/si-list

List archives are viewable at:     
                //www.freelists.org/archives/si-list
 
Old (prior to June 6, 2001) list archives are viewable at:
                http://www.qsl.net/wb6tpu
  

Other related posts: