POE OPTIONS BY MODEL

[sirhc]

Hmm, I don't think so (partially also because now the PoE setup has been proven to work (at least once, any way). But I will pull the logs for your assessment. Talking about logs, there's another queer issue with my switch logging, i.e. despite me setting a specific NTP server, the timestamp on the log is severely off.
Will post the log later.

Not sure why your timestamp is off other than you either have the NTP service or the switch IP address configured wrong or the switch does not have access to the internet maybe because of firewall rules or something. The NTP service is working fine for me and many others. see picture below

My guess is you are sitting behind a router and the switch can not get out to the internet to get the time.

I would suggest going to the CMD line of the switch and verifying that you can ping a public IP address and find out why the switch is not getting out to the real world internet.

NTP.jpg (126.78 KiB) Viewed 73982 times

which brings me to a PoE generic question:
Assuming the max power draw of the AP never exceeds 15 watts (even under load), is it better



for me to use 24V 0.75A instead of 48V 0.75A, i.e. use the lowest voltage and amperage to drive the PoE device?


update:
What I observed (today, i.e. attempt #3): I tried both 24V & 24VH (on port2) to power up the UAP-AC but no-go; in the end it's 48V (not 48VH) that succeeded (just like yesterday's attempt #2).
& what I interpret


as the switch "rebooting", was what occurred when I tried to use 24V & 24VH to power up the UAP today: the console page (of the Netonix) reported something along the line of, "
attempting to communicate with the device, no configuration can be loaded until this page is reloaded
".
I'm still perplexed though, i.e. right now I'm observing the power draw from the UAP-AC (as reported by the Netonix switch console page) to average around 10.3watts (even under load; yes, I took note today), & only hit 15watts once. By right, 24V 0.75A should be more than sufficient
to drive the AP; but apparently it can't. & I can't fathom the logic.

The fact that you seem to think the UAP-AC can even run on 24V or 24VH confuses and concerns me. Devices can not run on both 24V and 48V they are designed to run on one or the other. The UAP-AC is designed to run on 48V. It will never run on 24V or 24VH.....EVER! It will work fine with 48V or 48VH but 48V is all that is needed. As I said I do not have the AC unit but I called a friend who does and he is powering the device just fine with our switch on 48V so something has to be wrong with your wiring?

THIS DEVICE REQUIRES 48V IT WILL NOT RUN ON 24V EVER UNDER ANY CIRCUMSTANCES. IN FACT TRYING TO RUN IT ON 24V OR 24VH COULD DAMAGE IT AND OR THE SWITCH.

Your log you posted is not formatted, very hard to read. The Email log button is not broken it works fine, it does not ask for an email address to send the log to to it uses the email address configured on the Device/Config Tab. see picture below

If you do not have the SMTP properly setup under the Device/Configuration Tab with valid email credentials and or an SMTP server that will accept relaying from it will NOT email the log.

HOWEVER THAT IS A GOOD SUGGESTION FOR A FUTURE SOFTWARE CHANGE AS THE USER WANTING THE LOG MAYBE AT A DIFFERENT EMAIL ADDRESS THAN THE SWITCH SENDS ALERTS TOO.

EMAIL.jpg (134.64 KiB) Viewed 73982 times

Here is your log properly formatted:

Jan 1 08:00:08 kernel: vtss_port: Loaded port module on board Luton26, type 5
Jan 1 08:00:09 system: Setting MAC address from flash configuration: EC:13:B3:61:06:4A
Jan 1 08:00:11 admin: adding lan (eth0) to firewall zone lan
Jan 1 08:00:17 pinger: Ignoring port bounces because date is wrong
Jan 1 08:00:20 dropbear[729]: Running in background
Jan 1 08:00:22 switch[764]: Port 2 unplugged, PoE Smart is disabling PoE
Jan 1 08:01:30 UI: Configuration changed by 192.168.1.103
Jan 1 08:01:30 UI: Port 2 PoE: changed from 'Off' to '24V'
Jan 1 08:01:33 switch[764]: PoE enabled on port 2, PoE Smart is starting cable check
Jan 1 08:01:38 switch[764]: Port 2 cable check results: Abnormal termination, Open, Abnormal termination, Abnormal termination
Jan 1 08:01:38 switch[764]: Port 2 passed PoE Smart cable check, turning on power
Jan 1 08:01:41 switch[764]: Port 2 unplugged, PoE Smart is disabling PoE
Jan 1 08:03:30 UI: Configuration changed by 192.168.1.103
Jan 1 08:03:30 UI: Port 2 PoE: changed from 'Off' to '24V'
Jan 1 08:03:32 switch[764]: PoE enabled on port 2, PoE Smart is starting cable check
Jan 1 08:03:37 switch[764]: Port 2 cable check results: Abnormal termination, Abnormal termination, Open, Open
Jan 1 08:03:37 switch[764]: Port 2 passed PoE Smart cable check, turning on power
Jan 1 08:03:40 switch[764]: Port 2 unplugged, PoE Smart is disabling PoE
Jan 1 08:04:07 UI: Configuration changed by 192.168.1.103
Jan 1 08:04:07 UI: Port 2 PoE: changed from 'Off' to '24VH'
Jan 1 08:04:09 switch[764]: PoE enabled on port 2, PoE Smart is starting cable check
Jan 1 08:04:14 switch[764]: Port 2 cable check results: Abnormal termination, Abnormal termination, Open, Open
Jan 1 08:04:14 switch[764]: Port 2 passed PoE Smart cable check, turning on power
Jan 1 08:04:17 switch[764]: Port 2 unplugged, PoE Smart is disabling PoE
Jan 1 08:04:40 UI: Configuration changed by 192.168.1.103
Jan 1 08:04:40 UI: Port 2 PoE: changed from 'Off' to '48V'
Jan 1 08:04:43 switch[764]: PoE enabled on port 2, PoE Smart is starting cable check
Jan 1 08:04:48 switch[764]: Port 2 cable check results: Abnormal termination, Abnormal termination, Open, Open
Jan 1 08:04:48 switch[764]: Port 2 passed PoE Smart cable check, turning on power
Jan 1 08:06:42 UI: Configuration changed by 192.168.1.103
Jan 1 08:06:42 UI: Port 2 PoE: changed from '48V' to 'Off'
Jan 1 08:07:17 UI: Configuration changed by 192.168.1.103
Jan 1 08:07:17 UI: Port 2 PoE: changed from 'Off' to '48V'
Jan 1 08:07:20 switch[764]: PoE enabled on port 2, PoE Smart is starting cable check
Jan 1 08:07:25 switch[764]: Port 2 cable check results: Abnormal termination, Open, Abnormal termination, Abnormal termination
Jan 1 08:07:25 switch[764]: Port 2 passed PoE Smart cable check, turning on power
Jan 1 08:25:14 UI: Configuration changed by 192.168.1.106
Jan 1 08:25:14 UI: NTP_Enable: changed from 'Disabled' to 'Enabled'
Jan 1 08:25:14 UI: NTP_Host: changed from 'time.windows.com' to '0.ch.pool.ntp.org'
Jan 1 08:25:19 system: starting ntpclient Jan 1 08:25:20 system: starting ntpclient

Btw, if indeed I tripped the poly-fuse on some of the ports,
assuming I don't do it often
, due to the nature of polyfuses, there
ain't no permanent
damage done right?

The Polyfuses are there for safety to prevent fires from bad wiring, they will not protect the switch from a dead short. They do NOT make Ethernet transformers with a high enough rating to also be able to deliver 1 AMP per pair at 55C. Since Polyfuses are rated at "so much current at such and such temperature" if they are rated at .75A per pair at 55C then at room temperature they will push far more current which will exceed the Ethernet Transformer capability. This is is why I say DO NOT DEAD SHORT TEST THIS SWITCH unless you want to fry a port.

If an Ethernet transformer is blown it will not achieve a 1G link or sometimes any link at all. The nice thing about our switch is that if you do DEAD SHORT a port you lose that port not the whole switch. Every time I ever DEAD SHORTED a ToughSwitch PRO 8 on 48V I lost the whole switch. They may have since improved the design as I have not recently tested one?

MY SUGGESTION IS YOU LOOK CLOSELY AT YOUR WIRING. I SEE IN THE PICTURES YOU ARE USING A PATCH PANEL AND INTERNAL WIRING.

Look there are 3,000+ switches in service around the world, do you not think that if they did not have the ability to even power devices correctly that people would not have noticed this already as this is their main function? The POE ability is hardware not software, it is not like we can change some program code and deliver different voltages and current to different pins. Yes a software function turns it on and off but what voltage and polarity goes to each pin is predetermined by the hardware and can not be changed with software.

ALL Ubiquiti products use the same polarity pin outs which is the 802.3af/at STANDARD. All WISP Switches use the same 802.3af/at pinout standard.

The fact that you are having a problem powering a device is more than likely is do to the fact that you are trying to power it with 24V and it is a 48V device and or your internal wiring is wrong.

This is what I would do:
1) DO NOT USE THE INTERNAL WIRING FOR TESTING
2) Use a short pre-made jumper and plug the device directly into the switch with just the pre-made jumper.
3) Verify if the device is 24V or 48V, in this case it is 48V so turn 48V POE on that port with the short pre-made jumper.
4) determine if the switch can indeed get to the internet to update the time. You can do this by going to the console and typing CMD [Enter] which will drop you to the linux shell and ping 8.8.8.8

[lligetfa]

Thanks for the pics but it doesn't change the 100 meter limit set by the EIA/TIA 568 standard.

Please note that termination workmanship can affect the cable diagnostics on the Netonix. Testing through patch cables and patch panels may produce odd results due to TDR reflections. Use a proper cable certifier if in doubt.

The spec sheet on the UAP-AC states 22W max @ 48V.

[ralliart12]

Not sure why your timestamp is off other than you either have the NTP service or the switch IP address configured wrong or the switch does not have access to the internet maybe because of firewall rules or something. The NTP service is working fine for me and many others. see picture below

My guess is you are sitting behind a router and the switch can not get out to the internet to get the time.

I would suggest going to the CMD line of the switch and verifying that you can ping a public IP address and find out why the switch is not getting out to the real world internet...

...4) determine if the switch can indeed get to the internet to update the time. You can do this by going to the console and typing CMD [Enter] which will drop you to the linux shell and ping 8.8.8.8

Yes, indeed my switch is sitting behind a router. Will try what you said.

...The fact that you seem to think the UAP-AC can even run on 24V or 24VH confuses and concerns me. Devices can not run on both 24V and 48V they are designed to run on one or the other. The UAP-AC is designed to run on 48V. It will never run on 24V or 24VH.....EVER! It will work fine with 48V or 48VH but 48V is all that is needed...so something has to be wrong with your wiring?

THIS DEVICE REQUIRES 48V IT WILL NOT RUN ON 24V EVER UNDER ANY CIRCUMSTANCES. IN FACT TRYING TO RUN IT ON 24V OR 24VH COULD DAMAGE IT AND OR THE SWITCH...The fact that you are having a problem powering a device is more than likely is do to the fact that you are trying to power it with 24V and it is a 48V device and or your internal wiring ins wrong...

Al right, let's bottomline this:
1a) it's extremely likely what I perceived to be the switch rebooting (due to certain status messages in the switch's console page) is due to [u]me trying[/u] to power the UAP-AC using 24V & 24VH
1b) I have a highly-flawed assumption that as long as the wattage is sufficient, any combination of voltage X amperage is fine with any PoE device.
2) there's nothing (drastically, if at all) wrong with my internal wiring (at least some of them, any way) since I'm able to get the entire setup functioning

...As I said I do not have the AC unit but I called a friend who does and he is powering the device just fine with our switch on 48V...

...Look there are 3,000+ switches in service around the world, do you not think that if they did not have the ability to even power devices correctly that people would not have noticed this already as this is their main function?...

Just so that we are clear: I did not, at any time, outright claim that your switch is unable to power any standard PoE device. I'm just asking why is something happening, & I will like to know how to make it work.

If you do not have the SMTP properly setup under the Device/Configuration Tab it will NOT email the log...

It also conveniently, did not tell me there actually ain't no email going anywhere (since I indeed, not have any SMTP server configured in the device's configuration). An error prompt may be nice?

...Your log you posted is not formatted, very hard to read...

Here is your log properly formatted:...

I was trying to email the log to myself but apparently I didn't catch that the email button requires a pre-configured address. Hence I just perform a "Select All" on the log's text & pasted it in the forum's text editor. I tried to use the code-tags but that didn't do the necessary line-breaks for the log. Next I tried pasting it in the Notes app of a Mac & then re-copying from that back into the forum's text editor & somehow the lack-of-line-breaks manifest itself again. I tried pasting it in an online Yahoo mail composition & the log gets formatted just fine so I assume somehow the forum's text editor will "return" the line breaks when I submit the post.

Essentially, since I couldn't figure out why the log-emailing isn't working & I pasted it in all the various text-editing forms I can get hold on & I still end up with logs without line breaks, well, at least you know I'd tried.

I do agree, that posting the log in the format I ended up doing, looks crappy for debugging

...The Polyfuses are there for safety to prevent fires from bad wiring, they will not protect the switch from a dead short. They do NOT make Ethernet transformers with a high enough rating to also be able to deliver 1 AMP per pair at 55C. Since Polyfuses are rated at "so much current at such and such temperature" if they are rated at .75A per pair at 55C then at room temperature they will push far more current which will exceed the Ethernet Transformer capability. This is is why I say DO NOT DEAD SHORT TEST THIS SWITCH unless you want to fry a port.

If an Ethernet transformer is blown it will not achieve a 1G link or sometimes any link at all. The nice thing about our switch is that if you do DEAD SHORT a port you lose that port not the whole switch. Every time I ever DEAD SHORTED a ToughSwitch PRO 8 on 48V I lost the whole switch. They may have since improved the design as I have not recently tested one?...

Am learning something new every day; I didn't even know things like a self-healing (poly-fuse) exists.

...MY SUGGESTION IS YOU LOOK CLOSELY AT YOUR WIRING. I SEE IN THE PICTURES YOU ARE USING A PATCH PANEL AND INTERNAL WIRING...

I will; had ordered a cable tester & will put it to use when it arrives.

...
This is what I would do:
1) DO NOT USE THE INTERNAL WIRING FOR TESTING
2) Use a short pre-made jumper and plug the device directly into the switch with just the pre-made jumper...

Eventually I will have to deploy the AP & connect it to the Netonix switch via internal wiring (& the patch panel), don't I?

Thanks for the pics but it doesn't change the 100 meter limit set by the EIA/TIA 568 standard...

No problem; just wish to give a clearer perspective.

...Please note that termination workmanship can affect the cable diagnostics on the Netonix. Testing through patch cables and patch panels may produce odd results...

Yes, I understand this.

...The spec sheet on the UAP-AC states 22W max @ 48V.

Come to think about it: does it means I have to explicitly limit the wattage output from the PoE switch, i.e. ensure it does not exceed 22watts?

[Dave]

ralliart12

Your posting is good, it helps educate us all...thanks...

dave

[lligetfa]

...The spec sheet on the UAP-AC states 22W max @ 48V.

Come to think about it: does it means I have to explicitly limit the wattage output from the PoE switch, i.e. ensure it does not exceed 22watts?

In the very basic Ohm's Law, Wattage (Current) is the result of Voltage and Resistance. You do not limit Wattage, the Voltage and the Resistance do that. The device is responsible for the Resistance and you are responsible for the Voltage.

http://www.allaboutcircuits.com/textboo ... ce-relate/

[sirhc]

Come to think about it: does it means I have to explicitly limit the wattage output from the PoE switch, i.e. ensure it does not exceed 22 watts?

No, you do not have to limit watts, the device spec sheet is just telling you how many it needs to operate normally.

Think of it this way, a standard bedroom Breaker in your house is 20 Amps, no ONE device in that room should need 20 AMPs there is just 20 AMPs available in that bedroom.

If you select 48V on port 12 to power this device then there is .75A or 36 watts available, the fact that the device only uses 22 watts is not a problem there is just head room.

Your car has over 100 horse power but you rarely actually use the full amount of power available to you.....well I hope you do not anyway or a lot of people are in danger!

[ralliart12]

Just curious:
The switch will not be left powered on 24/7; it's a home environment after all, i.e. I will switch it on every evening & off within the same night.
Will I have to hit the switch's console page to manually turn on the relevant PoE option for the relevant port every time I power on the switch, or will the switch remember that, e.g. PoE (48V) on port 2, was activated during the last "session" & (after executing the necessary PoE-smart test) automatically turn on PoE on that port?

ralliart12
Your posting is good, it helps educate us all...thanks...
dave

Thank you for your kind words; luckily I've a non-production environment to make the mistakes I made. Hopefully my erroneous approach can serve as useful information for others.

[sirhc]

It will remember, that is why it writes the config to flash

[ralliart12]

Not sure why your timestamp is off other than you either have the NTP service or the switch IP address configured wrong or the switch does not have access to the internet maybe because of firewall rules or something. The NTP service is working fine for me and many others. see picture below

My guess is you are sitting behind a router and the switch can not get out to the internet to get the time.

I would suggest going to the CMD line of the switch and verifying that you can ping a public IP address and find out why the switch is not getting out to the real world internet...
4) determine if the switch can indeed get to the internet to update the time. You can do this by going to the console and typing CMD [Enter] which will drop you to the linux shell and ping 8.8.8.8

I am able to ping Google's DNS server via the switch's cmd-line:

[sirhc]

OK then you know you have access to the world.

Now you need to determine why you can not get to the NTP server you configure.

The NTP service work or else a lot more people would be complaining about it as would I because I deploy about 20 of these switch at my WISP.

So with that being said you need to try and figure out why your not getting time.

Maybe try a different server maybe look at your router to see if there are any stateful firewalls or access control lists.