RE: Discovery process
I am not sure that the two are mutually exclusive.
Geoff has talked about having an LED at a powered wall
plate to indicate that power is available. With the "diode
detection" I have proposed, that LED would be powered by
the detector during its normal process and would blink
constantly at a very low rate.
The LED would not impact the function of the diode detector
because it would allow a high forward and reverse voltage
which would not impact detection. If you recall, it was the
clamping action of a forward biased schottky diode that
enabled detection by limiting the voltage to < 1.5v when
Once power was applied, that LED (there could be two in
opposing polarities so that the LED would go on constant when
power is applied regardless of polarity) would draw < 10mA.
So you could set the requirement of the PDTE to draw 20-30mA
so that it is possible to ensure detection of a removal of the
PDTE when the current draw drops below 20mA.
_________ _/ ___________ Daniel Dove Principal Engineer __
_______ _/ ________ dan_dove@xxxxxx LAN PHY Technology __
_____ _/ ______ Hewlett-Packard Company __
____ _/_/_/ _/_/_/ _____ Workgroup Networks Division __
____ _/ _/ _/ _/ _____ 8000 Foothills Blvd. MS 5555 __
_____ _/ _/ _/_/_/ ______ Roseville, CA 95747-5555 __
______ _/ ________ Phone: 916 785 4187 __
_______ _/ _________ Fax : 916 785 1815 __
__________ _/ _________________________________________________________
I n v e n t !
> -----Original Message-----
> From: verger [mailto:verger@xxxxxxxxxxxxxxxxxx]
> Sent: Friday, April 28, 2000 6:22 AM
> To: Geoff Thompson
> Cc: Raymond Gass; Gerard Vergnaud; stds-802-3-pwrviamdi@xxxxxxxx
> Subject: Discovery process
> I'm a concern raised by the two mails here below.
> On one side, it is necessary to take into account a possible
> low current induced by an indicator and in the other side
> it is necessary the power sink draw a minimum current
> to hold the power sent via MDI.
> Question : how will it be possible to distinguish these two
> types of current? and how will it be possible to detect the
> disconnection of the terminal?
> Objet: Re: Discovery process
> Date: Wed, 19 Apr 2000 11:07:56 -0700
> De: "Geoff Thompson" <gthompso@xxxxxxxxxxxxxxxxxx>
> A: verger <verger@xxxxxxxxxxxxxxxxxx>
> Copies à: stds-802-3-pwrviamdi@xxxxxxxx, Raymond Gass
> This issue was discussed in committee. The decision was that:
> 1) If the "discovery characteristic" was found then
> power would
> be sent
> 2) For a device to KEEP the ability to draw power without
> restarting discovery there would be a requirement for the
> load to draw a minimum amount of power (5 ma was
> mentioned) over the link.
> >Hi guys,
> >I am thinking about the different proposals done, concerning
> >the discovery process, that in case where the terminal is
> >locally powered then the power via MDI is not sent.
> >Unfortunately, this case will be unavoidable since very often
> >the terminal is installed by the end user who is not awared
> >about the properties of the network infrastructure.
> >Moreover the network is able to provide a safe power
> >distribution (power with battery-back-up) and therefore
> >it will be better to provide such a power to any terminals.
> >What are you thinking about this problem?
> >Our solution is to analyse the impedance under alternative
> >voltage and this analyse stays efficient even the terminal
> >is locally powered and therefore this allows providing a
> >safe power to terminal whatever its state and in case of mains
> >breaking holding the current communication with this terminal.
> Objet: Re: Continuous Current Capacity of Panels, Jacks, and Plugs
> Date: Thu, 10 Feb 2000 16:52:24 -0800
> De: "Geoff Thompson" <gthompso@xxxxxxxxxxxxxxxxxx>
> A: Raymond Gass <Raymond.Gass@xxxxxxxxxxxxxxxxxx>
> Copies à: stds-802-3-pwrviamdi@xxxxxxxx, jhkee@xxxxxxxxxx
> This proposal sounds like it is in line with the discovery proposals
> that the group to be homing in on. The one item that I would take
> direct issue with would be the null value. I believe that we will end
> up defining a minimum, non-zero current that we will decide
> represents the non-connected condition. This will be to make
> provision for a small amount of current to be leaked into the
> TO/socket itself so that an indicator can be lit.
> >we have been looking at the various proposals for
> >remote powering / discovery process, and we would
> >like to stimulate discussion about a simplified
> >the benefits of this proposal are as follows:
> >- it does not introduce extra hardware in the power
> >sink, and thus no extra cost in the terminal part
> >- no impact on the data transmission
> >- it can be used whatever choice for powering
> >(phantom or direct mode, on data or on spare wires)
> >the method is based on a measurement of the
> >value of the current, performed by the "check
> >box" associated to the power source:
> >- during discovery process, a low frequency (some kHz),
> >low voltage (2-3 V) is sent, that makes it possible to
> >measure the line impedance (whatever line), and to
> >decide if this particular line needs to be fed or not,
> >based on the peak value of the AC current
> >- during feeding phase, DC current is being monitored.
> >Power feeding is maintained as long as DC current
> >is not null. Discovery process is re-initiated when
> >DC current is null
> >we have prepared some slides to describe further
> >this solution, and we are ready to make a presentation at the
> >next 802.3 meeting
> >thanks for your comments