Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

[802.3af] RE: Surge test with 0.3u/50u



Hi Roger,
 
I did not run a test with 0.3u/50u 1KV through two 404 ohms in parallel.
 
I did run a test with 1.2u/50u 1KV - 4KV  through 2ohms to 12ohms and in some tests I pass without external protection
and in some tests I failed without external protection (bad luck for PCB and its inhabitants...). all of these tests were part of immunity tests and not part of IEEE802.3.
 
I tested only the behavior from 2ohm to 12 ohm at 1.2u rise time and saw significant difference at this range alone (2ohms 12ohms).
So Increasing the source impedance from 2-12 ohms to 202 ohms improve the outcome dramatically which is not a big surprise.. Ip max =Vp/Z... 
 
Regarding the equipment required: I am checking with labs around here for 0.3us rise time generator.
 
I don't know personally of somebody who pass this tests (0.3u , 404ohms) usually it is implementation and layout dependent and above all, it is done at system level not at chip or module level.
I assume that there are many that pass this test due to the fact that it is old requirement in 802.3 and now that I have notice in drawing 33-18 that the source impedance is 202 ohm I believe that it was not a big challenge.
 
If you are doing surge tests as part of immunity tests 1K-2KV, 1.2u/50us through 2-12 ohms as part of EMI package (required only for environment B)
than meeting : 1K, 0.3u/50u, through 202 ohms is much easier and you are OK for 802.3.
 
IN ANY CASE YOU SHOULD TEST IT IN LAB DUE TOO SENSITIVITY OF THE RESULTS TO LAYOUT AND IMPLEMENTATION.
 
 
Regards
 
Yair.
 
 
 
 
 

 ----Original Message-----
From: Roger Karam [mailto:rkaram@xxxxxxxxx]
Sent: Wednesday, January 29, 2003 6:44 PM
To: Yair Darshan
Cc: 'Geoff Thompson'; 'mike_s_mccormack@xxxxxxxxxxx'; 'Steve Carlson'; 'Schwartz, Peter'; 'Dave Dwelley'; 'Chris DiMinico'; 'ccullin@xxxxxxxxx'; David Pincu (Dudi); 'stds-802-3-pwrviamdi@xxxxxxxx'
Subject: Re: Surge test with 0.3u/50u

Hi Yair
did you or anyone else do this test?
did you pass it?
and where can we find the equipment to test this.
roger


At 08:31 AM 1/29/2003, Yair Darshan wrote:
Chris, Roger and all,
 
 
We in Vancouver wanted to delete the requirement of 0.3u/50u 1000V fault tolerance test.
I have also mentioned that this is a destructive test.
It is true if it is done with low source impedance 2-12 ohms which is the standard surge test when it forms part of the immunity tests during EMI tests.
 
However, this requirement is specified by figure 33-18 with two 404 ohms source resistance which make life much easier.
So if we can easily remove this requirement from the draft lets do it.
If it cause to much trouble (old requirement from 802.3) than it doesn't worth the time, I don't believe we will have problems with 1KV pulse 0.3u/50us
through 202 ohm source resistance (two 404 ohms in parallel according to figure 33-18).
 
The above info was sent to Chad as part of my comments to draft 4.01 as informative material to Roger and others who had concerns regarding this issue.
 
Regards
 
Yair.