cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
1762
Views
0
Helpful
13
Replies

SG500-52p PoE Voltage fluctuating

gberg
Level 1
Level 1

Hi,

 

I wonder if it's normal for the PoE voltage be very fluctuating?

I have seen voltages from as low as 31V all the way up to 94V

Mainly I monitor the switch via SNMP using Observium, but I have also observed strange voltags using CLI command: 

 

show power inline consumption

13 Replies 13

marce1000
VIP
VIP

 

 - Make sure to use latest firmware

 M.



-- ' 'Good body every evening' ' this sentence was once spotted on a logo at the entrance of a Weight Watchers Club !

As far as I know I'm running latest version, both boot version and FW version. Here you can see what versions I'm running:

https://i.imgur.com/8ra0ZU4.png 

Hmmm perhaps an indication the power supply has seen some days and might need to be replaced? I wouldn't wanna jump to conclusions, but if its been installed and running constantly, that might be it. They wear down over time.

 

In my days as a data center tech power supplies and mechanical drives were the most often replaced FRUs. Not sure which was top dog... but those 2 parts waaay exceed any other component in replacements. There was a period of time that GBICs were getting replaced like crazy but I think there was a bad batch or something.

 

Anywasy, I'm only speculating. Unfortunately I know thatthe PSes on thes are not quick or hot-swap. That is unless the 52P is different that the 28Ps that I have. I mean yea the pull more power.. so maybe they are. I haven't seen on up close in ages.

 

I wanted to thank youfor your post becauses I had recently updated my fw on the 28P but I didn't realize there was a newer boot version. So I just looked through all the releases and found the 1.4.1.0.3 that gave me the same Boot version you have. I was so focused on all the incremental upgrading and was having issues with TFTP that I completely overlooked this one.

 

I'll keep an eye on this post. I'm curious as to what the answer to your power fluctuations is. Later! 

Zed969
Spotlight
Spotlight

So you're speculating that what you're seeing may be normal beahviour? Given that way that PoE is so variable that may be true. Some switches boast about how they can essentially control the voltage and current very tightly to conserve and provide power to more devices concurrently than it appears they should based on specs.

 

I honestly haven't paid close enough attention in practice to know. I do have a copy of the 802.3bt spec from IEEE. I have read through it but not to a critical level. Seems like a bit of research into that idea would be worthwhile. Either to confirm or rule that out as acceptable, within-spec, operating as designed beaviour. 

Zed969
Spotlight
Spotlight

Perhaps is has to do with the particular cabling. The grade, the type (stranded vs solid core), shielded vs non-shielded, category, length, wear or stretching, quality of terminations, etc. There are so many potential factors.

 

Are you seeing this remotely or do you have access to the switch physically? Just curious as to whether its happening on certain ports and if there's anything obviously different about the  cabling. 

 

It's not my intention to confuse the issue, I just started looking at the IEEE 802.3bt spec and this occurred to me.

Yes, of corse it could be the power supply.

But I also want to add that prior to SG500-52P I used a smaller cusin SG200-26P, and it showed the exact same fluctuating PoE voltage.

Also, the only thing connected to the switch that's using PoE is a Ubuquiti Unifi AP AC Lite, which draws about 4W, and I'm thinking that the readings aren't very accurate such small load?

 

I manly use Observium to monitor the switch via SNMP, but I have also seen strange voltages using the CLI command "show power inline consumption"

 

The AP connected to the  switch using a 3m Cat5e cable.

 

Btw, have you ever checked the PoE voltage of your SG500-28P switches?

 

 

Hello, I just got my first SG500-28P last week. I've had it up & running
for a few days. I upgraded the firmware and got it minimally configued.
It's powering only one thing currently at ~4W. I'll be receiving another
one in a few days. Once I get that one setup I will compare them to see if
they behave similarly. These are obviously used, so it may be a matter of
the amount of usage they have seen.

Forgive me if I am asking redundant questions, but I am involved with
several projects and have been studying various aspects of networking and
other related technology deeply for the last month or so, so my questions
may spring from many sources and I'm digesting and correlating a lot of
info. Sometimes things have to go to background processing and then get
regurgitated.

Are you seeing problems with your switch or devices as a result of the
fluctuating voltages? Like things behaving erratically or reporting errors?
Or is this something you've noticed and doesn't seem right, so you're
monitoring the situation.

I know from reading the release notes of the lastest firmware, as well as
the incremental upgrades that I had to do since my switch was back-revved
pretty far, that the PoE controller gets updated with a few versions along
the way to the most recent one. I'd be interested in behaviour before and
after any updates that involved the PoE controller. If you've collected
logs, it would be interesting to compare them.

And hey, this is kind of an academic, abstract thing for me. If I see
issues on mine, I'll take a voltmeter and take some real-world
measurements. That brings me back to the concrete. I tend to get a little
theoretical and propose several, what about this, what about that scenario
when I don't have tangible things to check.

Please be assured that I am actually trying to help pinpoint what's
happening and discover a solution but this problem is also useful to me to
validate my understanding of the elements involved.

If you have any logs, or other data to analyze, please upload or email them
so I can see. If I discover anything useful from working with my switch(es)
I will do the same.

Oh, have you disabled the EEE settings? and tried toggling the cabled
detection, short reach and other settings under the energy management
section? seems like those things would likely affect your voltages.
.

I am not seeing any real issues with this, both the switch and the connected AP runs ust fine, I just feel that it can't be good for the AP to get such wide variation in the input voltage.

 

I have also thought about measuring the actual voltage, but it's also a bit hard to do when the AP is connected using a patch cable connected directly to the switch, I'll try to connect it via a wall jack and measure the voltage on the krone connector.

 

Here are some log from Observium:

https://i.imgur.com/rMQd9wg.png

 

You're not kidding! From a quick glance, its all over the place. I got a
bounce notification for my last reply, but I sent the output of my
green-ethernet and EEE settings as well as the output of 'show power
inline output' or whatever that command was that you ran. I ran it a buncha
times iteratively with a few minutes in between and mine's pretty solid.
it's 52V -54V with only a single V increment or decrement each time i ran,
so 52, 53 & 54 were my only values shown compared to your logs there . just
Sheesh!

Have you checked the EEE power saving and other green-ethernet settings? I
shut all that stuff off because it messes with my critically timed Dante
audio and Video over IP. Every configuration recommendation and guide I
have for the stuff I do say SHUT IT OFF!

Like I said, the clocking signals PTP stuff is the most critical traffic on
my switches because they have to be tightly aligned to ensure the lowest
latency >10ms for audio and video. If you're dropping packets due to timing
isses that results in distorted audio or dropouts, either of which are
unacceptable. When setting up QoS they get set to the highest possible
priority, with even the actual content being set to second priority. That
is how important the distributed clock signals are. Anyway, I digress.

I have to think that EEE power saving could be affecting it. May also the
AP that you're powering is inconsistent in its demands. That's a direction
we haven't explored. do you have any other devices to plug in and power,
even temporarily just to see how they behave?

I've run that 'show power inline consumption' iteratively every few minutes
for a while and the voltage only varies from 54V -52V on mine. Seems to
only change in 1V increments and appears consistent. I do have all that
802.3az EEE stuff turned off though.

BigPowerSwitch01#show green-ethernet

Energy-Detect mode: Disabled
Short-Reach mode: Disabled
Power Savings: 67% (3.88W out of maximum 11.57W)
Cumulative Energy Saved: 0 [Watt*Hour]
Short-Reach cable length threshold: 50m

Port Energy-Detect Short-Reach VCT Cable
Admin Oper Reason Admin Force Oper Reason Length
-------- ----- ---- ------- ----- ----- ---- ------- ----------
gi1/1/1 on off off off off
gi1/1/2 on off off off off
gi1/1/3 on off off off off
gi1/1/4 on off off off off
gi1/1/5 on off off off off
gi1/1/6 on off off off off
gi1/1/7 on off off off off
gi1/1/8 on off off off off
gi1/1/9 on off off off off
gi1/1/10 on off off off off
gi1/1/11 on off off off off
gi1/1/12 on off off off off
gi1/1/13 on off off off off
gi1/1/14 on off off off off
gi1/1/15 on off off off off
gi1/1/16 on off off off off
gi1/1/17 on off off off off
gi1/1/18 on off off off off
gi1/1/19 on off off off off
gi1/1/20 on off off off off
gi1/1/21 on off off off off
gi1/1/22 on off off off off
gi1/1/23 on off off off off
gi1/1/24 on off off off off
gi1/1/25 on off off off off
gi1/1/26 on off off off off
BigPowerSwitch01#show eee

EEE globally disabled

EEE Administrate status is enabled on ports:
fa1/2/1-48,gi1/1/1-52,gi1/2/1-4,fa2/2/1-48,gi2/1/1-52,gi2/2/1-4,fa3/2/1-48,gi3/1/1-52,gi3/2/1-4,fa4/2/1-48,gi4/1/1-52,gi4/2/1-4,fa5/2/1-48,gi5/1/1-52,gi5/2/1-4,fa6/2/1-48,gi6/1/1-52,gi6/2/1-4,fa7/2/1-48,gi7/1/1-52,gi7/2/1-4,fa8/2/1-48,gi8/1/1-52,gi8/2/1-4
EEE Operational status is enabled on ports:
EEE LLDP Administrate status is enabled on ports:
fa1/2/1-48,gi1/1/1-52,gi1/2/1-4,fa2/2/1-48,gi2/1/1-52,gi2/2/1-4,fa3/2/1-48,gi3/1/1-52,gi3/2/1-4,fa4/2/1-48,gi4/1/1-52,gi4/2/1-4,fa5/2/1-48,gi5/1/1-52,gi5/2/1-4,fa6/2/1-48,gi6/1/1-52,gi6/2/1-4,fa7/2/1-48,gi7/1/1-52,gi7/2/1-4,fa8/2/1-48,gi8/1/1-52,gi8/2/1-4
EEE LLDP Operational status is enabled on ports:

BigPowerSwitch01#

This is what the output of "show green-ethernet" looks like:

https://pastebin.com/raw/bTcq33Wd

 

And this is what the output of "show eee" looks like:

https://pastebin.com/raw/14YUdTir

Yep. Looks like you have all the various energy saving things on &
available. Looks like the switch is choosing how to apply it to each port
including by cable length.

I'd suggest backing up your current config and just for diagnostics,
turning all of it off. Then see if the voltage stabilizes at all. If not,
you can just re-apply the original config. I have never had that stuff on
past the initial config of my switches for the reasons I mentioned earlier,
so I don't know if it will make a difference. Just an idea. If you
decide to try it, please let me know the results.

Zed969
Spotlight
Spotlight

Perhaps is has to do with the particular cabling. The grade, the type (stranded vs solid core), shielded vs non-shielded, category, length, wear or stretching, quality of terminations, etc. There are so many potential factors.

 

Are you seeing this remotely or do you have access to the switch physically? Just curious as to whether its happening on certain ports and if there's anything obviously different about the  cabling. 

 

It's not my intention to confuse the issue, I just started looking at the IEEE 802.3bt spec and this occurred to me.

Getting Started

Find answers to your questions by entering keywords or phrases in the Search bar above. New here? Use these resources to familiarize yourself with the community:

Switch products supported in this community
Cisco Business Product Family
  • CBS110
  • CBS220
  • CBS250
  • CBS350
Cisco Switching Product Family
  • 110
  • 200
  • 220
  • 250
  • 300
  • 350
  • 350X
  • 550X