Re: catalyst 6500 power

Your "Major UK Provider" might know networking, but they sure as hell don't
know anything about electricity. The power supplies in the 6000 chassis are
industrial strength (which is why you pay an arm and a leg for them!) and
are full isolated. What this means is that power supply A is completely
isolated from power supply B. Imagine you have two toasters, each plugged
into two different receptacles on different phases. Will toasting bread at
the same time cause "strange toasting behavior" to occur because of
"harmonics" of the different phases? That sounds wacky because it is! This
is also true for your power supplies because even though they are in the
same chassis, they are completely independent of each other. On the bigger
power supplies that have two AC inputs you in fact have two power supplies
in one "box" and each is isolated from the other.. If you had two of these
in one chassis, you would have 4 power supplies, each completely distinct
from the other.

The power supplies take AC in and convert it to DC power that is then
supplied to the chassis backplane, which is then used to power the blades
in the chassis. The power supplies do not provide any AC power to the
backplane. Since we are running on DC power there is no chance of
"harmonics" affecting any component in the system (harmonics only affect AC
systems, not DC systems) Since the power supplies are isolated, one power
supply can't affect the other. If the power supplies were not isolated, a
problem on one power supply could affect the other, and the power supplies
would no longer be redundant, which is why they are isolated in the first
place.

This whole idea about using the same AC phase is crazy. Many of the power
supplies can use two phase AC input. (three wire, one neutral and two "hot")
Each of the "hot" wires are on a separate phase (This is how 208v is
supplied in the US). For full redundancy, you should have power supplied
from two different PDU's, and power supplied from each of the PDU on
different phases. This way, if you loose a PDU, you still have power, if
you loose a phase A, and power is supplied from phase A on both PDU's you
loose power to both. If you get power from phase A, on PDU 1, and phase B
from PDU 2, the only way you are down is if you loose power on both phases.

Scott

"CiscoHeadsetAdapter.com" <for-spam@xxxxxxxx> wrote in message
news:cdSdnYhaZ8yCTWTe4p2dnA@xxxxxxxxxxxxxx
<simon.marsden@xxxxxxxxxxxx> wrote in message
news:1140431028.818493.28070@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
we are having 2 new 6513's and 2 new 6509's, with 2 4000watt and 2
3000watt PSU's respectively.

Cisco state this:

All AC power supply inputs are fully isolated.
-Source AC can be out of phase between multiple power supplies in the
same chassis, which means that PS1 can be operating from phase A and
PS2 can be operating from phase B.
-Source AC can be out of phase between AC inputs on power supplies
that are equipped with multiple AC inputs, which means that power cord
1 can be plugged into phase A and power cord 2 can be plugged into
phase B.

http://www.cisco.com/en/US/products/hw/switches/ps708/products_installation_guide_chapter09186a008020e0d3.html#wp1026596

But our installer (a major UK provider) recommend the following.

"Both power supplies in the same chassis need to be on the same phase,
they can be fed from different PDUs and this is probably more
resilient. But must be on the same phase otherwise strange things
happen to the catalyst because of the harmonics."

Who's advice should we listen to?

Rgds, Big Si.


Hi Simon,

Probably you should change your "a major UK provider". :-)))

Cisco clearly states in their specification, that PSs are FULLI INSULATED.
Plus, even looking logically, Power Supplies provide DC power to the
switch. Switch has no synchronization from the AC power. And even if
"strange things happen to the catalyst", it's Cisco's problem and this
should be resolved by the Cisco. I would trust more to the Cisco's words,
than to an installer, unless they can provide proven track of a Catalyst
failure "because of the harmonics".

Good luck,

Mike
CCNP, CCDP, CCSP, Cisco Voice, MCSE W2K, etc...
-------
www.ciscoheadsetadapter.com




.

Relevant Pages

  • Re: HP Compaq power supply keeps failing
    ... DC7700 computer at work that keeps blowing power supplies - think I'm ... Dust does not cause damage. ... line voltage may need to exceed up to 600 volts for a certain time ...
    (comp.sys.hp.hardware)
  • Re: new computer question
    ... I find power supplies often missing essential functions. ... BTW there is no relationship between watts and quality. ...
    (microsoft.public.windowsxp.hardware)
  • Re: HP Compaq power supply keeps failing
    ... DC7700 computer at work that keeps blowing power supplies - think I'm ... Many blame dust and heat for failures that are created ... Dust does not cause damage. ...
    (comp.sys.hp.hardware)
  • Re: Power supplies keep crapping out on me
    ... It might jist be the off/on switch just went bad. ... I've now had three power supplies fail on me. ... The most recent failure was sometime last week. ...
    (comp.sys.apple2)
  • Re: terminal strips versus tagboard versus pcb - construction questions
    ... If I might open a discussion on guitar amplifiers in this newsgroup, ... Sadly the old power amp ... chassis from an antique TV that I used, ... 12ax7 with a gain stage in half and the other half a split-load phase ...
    (alt.guitar.amps)
Loading