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

Re: [802.3_4PPOE] Channel P2P RUNB



Hi George,

 

Thanks for the good summary.

1.       We all in agreement that we need to focus and specify that Channel Pair to Pair resistance unbalance , PSE PI Pair to Pair resistance unbalance, PD PI Pair to Pair resistance unbalance from the standard point of view.

2.        

 

From: George Zimmerman [mailto:george@xxxxxxxxxxxxxxxxxxxxxxxxxxxxx]
Sent: Wednesday, April 30, 2014 1:57 AM
To: STDS-802-3-4PPOE@xxxxxxxxxxxxxxxxx
Subject: Re: [802.3_4PPOE] Channel P2P RUNB

 

Yair – thank you for the clarification on acronyms.  Being clear on this stuff stops a lot of misunderstanding.

 

With that for clarity, I will try to focus on the text we have to write, since it will clarify the numbers and definitions that we have to agree on:

 

From the standpoint of the standard, what we need specify are:

Channel Pair to Pair resistance unbalance  is the P2P resistance unbalance of the cables and 4 connector model.

 

PSE PI Pair to Pair resistance unbalance is the P2P unbalance measured at the PSE  PI and include PI interface circuitry such RDSON, Current sense resistor, equipment connector, magnetic winding resistance. This is included in the " end to end channel resistance unbalance" and need to be extracted from it to be separate definition for PSE PI P2PRUNB.

 

PD PI Pair to Pair resistance unbalance is the P2P unbalance measured at the PD PI and include PI interface circuitry such DIODE BRIDGE Voltage drop difference and dynamic resistance, equipment connector, magnetic winding resistance. This is included in the " end to end channel resistance unbalance" and need to be extracted from it to be separate definition for PD PI P2PRUNB.

 

Yair: Agree. See ad hoc report with more details.

-----

As far as text, I believe that should be easy, the numbers are the hard part.  For Text, I would propose:

 

For channel pair-to-pair resistance unbalance, that we add a section 33.1.4.3 (modeled on 33.1.4.2), called “Pair-to-Pair Resistance Unbalance”, and retitle 33.1.4.2 “Within-Pair Resistance Unbalance”

 

33.1.4.3 Pair-to-Pair Resistance Unbalance

Type 3, 4 pair operation, requires the additional specification of resistance unbalance between each two pairs of the cable shall be 5%(TBD) or less.  Resistance unbalance between the pairs is a measure of the difference of resistance of the common mode pairs of conductors used for power transmission.  Measurement of resistance unbalance is described …. (if you agree, we will need to complete this text similarly to 33.1.4.2)

Yair: Agree. I already started to craft the baseline wording since this was the planned agenda for this week Thursday meeting and I'll use your suggested wording and add some missing parts.

-----

For PSE PI Pair to Pair resistance unbalance and PD PI Pair to Pair resistance unbalance we have a different problem.

The existing standard does not specify PSE or PI resistance unbalance, but rather specifies the unbalance current.  (sec 33.2.7.11 and Table 33-11).

Yair: PSE PI or PD PI we cannot use current unbalance definition, we need to use only resistance unbalance and the reasons are:

1.       In  sec 33.2.7.11 and Table 33-11 this is the pair current unbalance and it is implementation independent. It is computed from Pair Runb=(Rmax-Rmin)/(Rmax+Rmin) and multiplied by the pair current. This is required for helping designing the magnetics and limiting their bias current so OCL will not drop.

2.       In Pair to Pair resistance unbalance, if we specify current unbalance instead of resistance unbalance we will force implementation and we must ensure that the P2P unbalance definition will not be implementation dependent but only the effect of the channel physical limitations and PSE , PD PI P2PRUNB limitations.

EXAMPLE of what will happen if we specify pair to pair current unbalance: It will force any PD or PSE to use active or passive circuitry (current sharing/balancing) to ensure that the  power drawn from the PD is splitted evenly within "End to End Channel P2PRUNB" range,  between the two power channel. This is not necessary,  has cost issues and thermal issues and again not necessary and limiting applications such:

a)       will not permit PD applications that required 50mA on pairs A and 150mA on pairs B. This is (150-50)/(150+50)=50%P2PUNB  while nobody cares since it below maximum pair current so nobody care has no technical advantage or other.

b)      Any PD that draws power less than Type 2 current per pair, will have to have current sharing circuitry with no reason.

 

Maybe with clever text we may use Iunb instead of Runb terms however we can start with resistance terminology and if we will find that current terminology will improve things without limiting PD applications then it is fine. I don’t believe that its should be our focus now.

 

3.       I am not worry about what the existing standard specify or not, since this is totally new requirement that need to be handled differently since their physical impact  on systems/sub systems and component is different. Any way if you look at 33.1.4.2 you will see that resistance unbalance do specified by resistance unbalance equation for the pair and the current unbalance per   33.2.7.11 and Table 33-11 is derived from it and is not the source requirement it is just the result to help with magnetics.

-----

My question to the group is whether this presently works to partition the unbalance between PD and PSE on the within-pair model?  If so, then I would recommend that rather than specifying the P2P resistance unbalance of the PSE PI, we may specify the DC current imbalance pair-to-pair. 

Yair: See above why specify resistance and not current.

----

If specifying just the unbalance current is sufficient (not sure how), then the way is clear -  we should add a line to Table 33-11, specifying the value, which I will allow others to put in (X%).  Additionally, we should add a line to section 33.2.7.11 Current Unbalance, stating:

Yair: See arguments above why specify resistance and not current. Regarding Table 33-11, I would suggest to address it in other round of discussion after we agree on the principles which is what we are doing now.

----

“For Type 3 operation, the specification for Ipp_unb in Table 33-11 shall apply to the current unbalance between the two pairs of conductors over the current load range.”

Yair: Here I would change your proposal to fit my arguments above, let's discuss it at the adhoc or over mail after you see my contribution on the subject that I sent few minutes ago.

------

If, however, we do need to partition for the electrical specification for interoperability, as a PSE and PD electrical spec, my proposal would be to add new specifications to Table 33-11, a new subclause for the PSE unbalance, and a statement of the PD unbalance requirement, in addition to appropriate test procedures.  I would propose to reuse the existing equation 33-1 which defines unbalance, perhaps with some clarifying language so that Rmax and Rmin are generalized to be between the two conductors under consideration.

Yair: Agree.

----

 

Specifically,

I would suggest, following 33.2.7.11, stating:

“For Type 3 operation, the resistance unbalance, as defined by equation 33-1, between the pair combinations for Mode A and Mode B operation of the PSE shall be less than or equal to the value of Rpp_unb in Table 33-11 over the current load range”.

(we will have to specify a test procedure as well)

 

The PI section, will simply require a new subclause, since there is no table of existing electrical parameters.  I suggest that the best place to put this at the end of 33.3.1, describing the PD PI.

 

“For Type 3 operation, the resistance unbalance, as defined by equation 33-1, between the pair combinations for Mode A and Mode B operation of the PD shall be less than or equal x% over the current load range”.

 

(and, we will have to add a test procedure as well)

 

 

George Zimmerman

Principal, CME Consulting

Experts in Advanced PHYsical Communications Technology

george@xxxxxxxxxxxxxxxxxxxxxxxxxxxxx

310-920-3860

 

(PLEASE NOTE NEW EMAIL ADDRESS.  THE OTHER WILL STILL WORK, BUT PLEASE USE THIS FOR CME BUSINESS)

 

From: Darshan, Yair [mailto:YDarshan@xxxxxxxxxxxxx]
Sent: Tuesday, April 29, 2014 2:21 PM
To: STDS-802-3-4PPOE@xxxxxxxxxxxxxxxxx
Subject: Re: [802.3_4PPOE] Channel P2P RUNB

 

Good question.

Here is the acronyms that we plane to use so no confusion will arise (I hope..):

 

Pair resistance unbalance : Is the resistance unbalance between two wires in the same pair as specified by IEEE802.3 and other standards. This is 2% for cable and 3% maximum for the channel. Channel is a 4 connector model (cables and connector only).

 

Pair to Pair resistance unbalance: is the resistance unbalance between two wires of the same pair connected in parallel to another two wires of other pair connected in parallel. It is 5% for a cable.

 

End to End channel pair to pair resistance unbalance" it is 26.2% worst case calculation on a worst case data base that we have generated. The 26.2% was calculated at 25degC. The channel is including components at PSE PI and PD PI that affects the whole end to end channel.

 

PSE PI Pair to Pair resistance unbalance is the P2P unbalance measured at the PSE  PI and include PI interface circuitry such RDSON, Current sense resistor, equipment connector, magnetic winding resistance. This is included in the " end to end channel resistance unbalance" and need to be extracted from it to be separate definition for PSE PI P2PRUNB.

 

PD PI Pair to Pair resistance unbalance is the P2P unbalance measured at the PD PI and include PI interface circuitry such DIODE BRIDGE Voltage drop difference and dynamic resistance, equipment connector, magnetic winding resistance. This is included in the " end to end channel resistance unbalance" and need to be extracted from it to be separate definition for PD PI P2PRUNB.

 

Channel Pair to Pair resistance unbalance  is the P2P resistance unbalance of the cables and 4 connector model.  This need to be excreted from the  " end to end channel resistance unbalance"  and specified separately.

 

So (PSE PI +Channel + PD PI)p2prunb ALL TOGETHER IS 26.2%.

 

I hope it is clear now. Please let me know if it is not.

 

Yair