If the 750UP is powered through PoE with the 24V 2.5A adapter that comes with it how much can it output per port? Is it still the 24V@500ma per port?
I am wanting to put it on a tower to power 3 Ubiquiti Rockets. If power is run through about 150ft of ethernet cable will there be enough power leftover to power up the Rockets? Basically, how many Rockets could you power from the 750UP when it is powered through PoE? 1? 2? 3? 4?
Ahh, looks like I was wrong (and I am most happy to stand corrected on this!). See JanisK’s post below.
I had written:I think the answer was given that each port has a limit of 500ma - and this includes the port (if any) that you are powering the 750UP from. So if you are delivering power to an ethernet port, you are limited to a total draw of 500ma (i.e. the 750UP plus its attached powered peripherals must draw no more than 500ma total).
limit on power out port is 500mA regardless of output voltage. so at 12V you will be able to power only MAX 6W device. while at 30V 15W.
If you power RB570UP with PoE 24V adapter over 100m cable, then resulting output voltage at full load (all 4 ports giving out 500mA totaling to 2A) then voltage to powered ports at RB750UP end will be around 21V then losses on cable and protections you end up with around 19V at the end device. and cables stretching for 200m from PSU brick.
PoE in port has no input limitation only output is limited to 500mA and with 4 prots it is 2A
Is the limit software controlled? Is there any way to disable it? I assume it is set well below the actual hardware safe output maximum.
Having this limit renders the RB750UP completely useless for probably 50% of solar wisp applications.
Is the output 802.3af? If not then thats a very big risk and something that should be put in big bold writing on the unit
why would it be? it’s not 48V for starters.
You dont exactly make that clear on rb.com, This means you are stuck powering only MT devices and you better be sure PoE is turned off before plugging anything non-MT into it. A warning label would be suitable!
I don’t see any specification that said 802.11af. This standard is also only for the 48V range.
Why should MT make a low cost device for a POE-standard not implemented in other routerboards. ubiquiti is also not compatible with 802.11af.
No need for a warning label, because the routerboard will detect if the other device can accept power, before powering it. This has been discussed in other threads, it measures resistance between certain pins on the ethernet port. You can plug in your laptop and nothing will happen.
I personally think that there is no need for a label, as nobody ever has advertised it as a 48V 802.3af device. Otherwise we would need a label that it’s not WIMAX compliant as well.
as normis told, it is not 802.11af since it does not have the power requirements of 48V. So it is not. but a lot of engineers have chosen to go for 12 - 28 V range for a reason and there are a lot of devices available from different vendors.
there are safety features in place that are borrowed from 802.3af, like detection if power can be turned on the port. Using PoE i have tested a lot of devices (not only MT ones) will they turn on, will they burn if poe-out=auto or on is given. What we saw was - 750UP will not burn gigabit ports (laptops, cards, switches etc.) it will not burn 100Mbit ports if they are not PoE ports. Older RouterBOARDS also worked perfectly with UP board.
we are thinking to add 4th option for poe-out setting that will allow devices with lower resistances to be added and turned automatically on, but that is going to be dangerous feature and you will have to use at your own risk.
So have you implemented that change… the change that will allow us to turn PoE on without the protection? I see that 5.12 was released, but no notes about a change here.
Thanks,
Dave
Heres a way to make a 750UP not completely worthless for us that run 12v:
http://forum.ubnt.com/showthread.php?t=45985&highlight=750UP
5.13 will have these changes. you can apply for test package by writing to support. Manual on wiki has been upgraded to reflect the introduced changes
also, about some comments for the thread that was linked to:
newer boards use different voltage/temperature measurement system. All newer RouterBOARD models use that. RB750UP uses ATtiny to measure these values and is very accurate. RB450G uses older measurement schematic that was quite inaccurate.
Also, my advice would be to go as high voltage as devices you try to power can allow and have short/over-current protection and control over port state. Later on probably some tasty features will come, so you will be able to use them. Or you can think about feature that port is used to operate relay, that in turn control power to your device (electric saw?). So you still get warranty if there are some problems discovered.
EDIT: Doh I should have checked the other thread. NEVERMIND! 
Thread necromacy. I see that 5.2 introduced a 1A limit per port. This would allow the powering of ubiquiti gear on a 12v site. Anyone test this yet?
new firmware where brought at 5.20 (five dot twenty) do not confuse RouterOS release numbering.
Yeah yeah, you know what I meant though. PS that is a stupid way to number releases.
Remit, this is industry standard to number releases. Apple does it, Linux does it (kernel numbers), Cisco does it.
wikipedia quote:
When a period is used to separate sequences, it does not represent a decimal point, and the sequences do not have positional significance. An identifier of 2.5, for instance, is not “two and a half” or “half way to version three”, it is the fifth second-level revision of the second first-level revision.
Still dumb, due to the exact reason above. 5.2.1 would basically eliminate any possible misunderstandings. As most industry leaders have adopted 