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

[802.3af] RE: Gain in the DC/DC




Hi Roger,
I too don't like 77Hz with zero tolerance. In my presentation I live it TBD,
we need to spec tolerance.
Lets ask the chip vendor what tolerance they prefer (+/-10% ??) and we will
design the other parts accordingly.
I'll run sensitivity analysis again and verify that we still OK.

Regarding the DC/DC converter output to input ripple rejection ratio, I'll
Run a test with your numbers 70-130Hz any input ripple voltage and you will
get the Vout/Vin transfer function for any given input as function of
frequency.

The results from above are not function of the cable type due to the fact
that we are talking of very low frequency, almost DC compared to the cable
bandwidth if you still find that cable type is important for this question,
please let me know.

Regards

Yair



-----Original Message-----
From: R karam [mailto:rkaram@cisco.com]
Sent: Wednesday, March 20, 2002 12:46 AM
To: Yair Darshan; 'stds-802-3-pwrviamdi@ieee.org'
Subject: Gain in the DC/DC


Hi Yair

it seems like no vendor is willing to sign up for a freq F say the 77hz you
picked
with zero tolerance, we are talking about integrating the oscillators...
so if we add 30% tolerance and say for lack of better data today we go to
100hz
now we are looking at 70-130hz range, what is the 'gain' of dc/dc converters
in
this freq range?  what do we expect the 40-100mv at the input of the PD to
look like at the output of the converter where the audio is?

again, i am doing my homework, i mean work-work.

also Please note that what Dan brought up is important all testing must be
with
cat-3 cables... NOT CAT-5.


roger