RE: PD power start up
See bellow my comments.
> -----Original Message-----
> From: Dieter Knollman [SMTP:djhk@xxxxxxxxxxx]
> Sent: ב מרץ 26 2001 19:09
> To: Yair Darshan
> Cc: stds-802-3-pwrviamdi@xxxxxxxx
> Subject: Re: PD power start up
> I believe we need a special start up state.
> I'm confused by your statement. See comments below.
> Yair Darshan wrote:
> > Dieter,
> > It looks O.K. however I would like to make some changes in the wording
> > choice 1:
> > Choice 1:
> > Normal operation:
> > The PD will will draw between 10mA to 350mA average current
> > Vin>30V.
> [Dieter] I don't like average. I believe 10 to 350ma are the limits that
> trigger the TBD timers.
[Yair Darshan] I have summarized why I think we need an average
definition, you probably know that my
intention is to allow some ripple current
around the average in order to support non-constant
load while keeping the average 350mA max. I
am aware of your and others concern that it
will put some burden on the power supply and
on the PSE switch. I am currently working on
the numbers that will allow us all the
goodies from all sides.
In addition see my response to David Kohl on
the reasons why we need it.
> > During startup :
> > The PSE will limit the current to 1.5 times the max. staedy
> > state value for TBD msec.
> [Dieter] I think you want more than this. Don't you want the PSE to
> 1.5 times Imax for about 100ms?
[Yair Darshan] Yes, that is what I have meant, 1.5 times max. steady
state = 1.5 times 350mA = 525mA ~ 500mA.
By stating max. steady state I mean max. average available=350mA.
> At present the PSE is only required to have a capacity of Imax available
> perform detection.
[Yair Darshan] I am not sure that I understand your meaning, why
during detection the PSE is limited to 350mA.
It should be disconnected from the port. Please advise of your
> I can set the current to limit to 1.5 times Imax, but supply only Imax.
> your PD start up?
1. Why you can supply only Imax? Is it hardware limitation that you
currently have or you are still referring to the case that the PD is
limiting to Imax=350mA and you set the PSE to 1.5 times Imax. Please
2. Its not my PD. It is just the fact that low cost PD power supply
can be designed with startup current higher than 350mA at less than
50-100mSec. Actually It can be design theoretically to almost any set of
numbers that you want by increasing the cost of the solution.
3. To your question: Some PD's will startup, depends for how long
you supply the Imax during startup.
> > During startup, PD may drop the line
> > to zero for TBD mesec max.
> [Dieter] This creates a conflict. Basic state machine has PD off for
> voltage <
> 30 volts.
> The new state machine will need a timer to switch from start up to ON.
> Plus some PD implementation issues. How does the PD maintain the start up
> state with the input voltage equal to zero? Looks like you need some kind
> monopulser. This raises the issue of false triggering.
> > Is it O.K. by you?
> [Dieter] No. Your PD design is in violation of the detection state
> We need a new state machine to allow this method of operation.
> > Yair.
> > > -----Original Message-----
> > > From: Dieter Knollman [SMTP:djhk@xxxxxxxxxxx]
> > > Sent: ?, ??? 24, 2001 8:21 PM
> > > To: Yair Darshan
> > > Cc: stds-802-3-pwrviamdi@xxxxxxxx
> > > Subject: PD power start up
> > >
> > > Yair,
> > >
> > > I believe we have two choices.
> > >
> > > Choice 1: Define a special start up case for the PD.
> > > In this state the PD is allowed to violate steady state
> > > The PD may drop the line voltage to zero.
> > > The PD may draw any current level.
> > > The PSE will supply current at a level about 1.5 times the steady
> > > state
> > > value.
> > > The time for the start up state is limited to X ms.
> > >
> > > Choice 2: The PD meets steady state requirements.
> > > In order to receive power
> > > the PD input voltage is >30 volts and
> > > the PD current is >10ma and <350ma.
> > > Some violations are allowed for Y ms.
> > >
> > > Can we have a straw poll?
> > >
> > > Dieter
> > >
> > > << File: Card for Dieter Knollman >> << File: Card for Dieter
> Knollman >>