The draft copy of IEC 63315 I have states
The PD shall not create any hazard when supplied with
– 130 % of the nominal rated input voltage of the PD at the relevant DC power transfer port for circuits greater than 5 V (with a minimum of 7,5 V)
The reference designs for the network powered device power conversion circuit often stipulate a SMAJ58A shunt overvoltage protector after the polarity correction diode bridge. The lowest voltage threshold SMAJ58A is 71.1 V. The SMAJ58A clamping conditions and using the highest threshold voltage give a linearized slope impedance of about 1.5 ohm. Apply 1.3× voltage to the network powered PD and allowing a 2 V bridge voltage drop gives a voltage of about 72 V across the SMAJ58A. This would give an initial current of about 0.6 A. SMAJ58A power loss is 43 W. As the SMAJ58A heats up the avalanche voltage will rise reducing the current and power loss.
As the thermal resistance is 120 deg C/W, 43 W implies a temperature rise 5000 degC. Clearly impossible, as the thermally generated carriers will exceed the silicon carrier doping level at some temperature and the silicon will go intrinsic (short circuit). What happens then depends on the SMAJ58A package construction. Soldered chip connections usually hold good. Bonded chip connections will often fuse leading to package fragmentation (explosion). Other things like PW vaporisation can happen.
The problem here is that the specified voltage source doesn't have a specified current capability and so very high currents will be able to flow if the SMAJ58A goes low impedance. This is a test condition that needs to be looked at carefully to avoid a safety fail.
On 15/03/2021 20:56:50, Chad Jones (cmjones) <00000b60b3f54e8d-dmarc-request@xxxxxxxxxxxxxxxxx> wrote:
All, I have received confirmation that I can share the 63315 draft in our password protected area. I have placed the first draft in the PDCC/63315 private area:
https://www.ieee802.org/3/ad_hoc/PDCC/private/63315/index.html
With our next meeting occuring in less than two days, I can’t expect that anyone can read and comment on the document. I will defer this to the next PDCC meeting, asking for comments to be sent to me by EOD
March 22. I will then consolidate them into one document that we can review on March 24th or 31st.
Regards,
Chad Jones
Technical Leader, Cisco Systems
Chair, IEEE P802.3cv Task Force
Chair, IEEE P802.3da Task Force
Principal, NFPA 70 CMP3
To unsubscribe from the STDS-802-3-PDCC list, click the following link: https://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-802-3-PDCC&A=1