I have an ap that seems to be rebooting when under load. (411ah with xr5 card)

The cat5e run is approx 70meters and being fed by 24v poe injector. To be fair there is a splice in the wire.

Can anyone give me an idea of what can be done distance wise with cat5e and 24v poe to power a 24v mikrotik board?

I've seen how to's that tell me how to build a tester to plug into the far end of the cat5 but that won't help me as this only happens when the radio isn't idle.

depends on the cabling. i had one going a few hundred feet and it seemed to work fine, although I wouldnt leave it that way. bring up the voltage monitor and see what it says when its running. then start maxing out the card and keep reading the voltage. if the voltage drops while your testing then its probably not enough current. an option would be to run 2 cables, one just for power and one for ethernet. double up the pins for power to get further.

What you see in /log print after reboot? All Mikrotik hardware designed to work with up to 100m POE

I wouldn't be so concerned with the voltage (your voltage should still be up around 20v at the RB even when you max out the ethernet standard at 100m, unless that splice is really causing some major issues). What's the amp rating of your PoE power supply? An RB411AH plus an XR5 is going to pull, when peaking, around 9 watts of power. That is about .375 amps at 24v. If your power supply is only rated at .375 amps or lower, that could be your issue. Using a power supply with a higher amp rating should fix your issue.

Also, since you mentioned the ethernet cable is spliced: is your power supply powering anything else? Don't forget to add in its power usage on top of the RB411AH's usage.

Scott

I'm not sure what I'm looking for in the log but I'll have a look through it. Interesting point about doubling up the pins if ever necessary. I'll double check the amps but I can tell you this radio is at a 24v solar facility with 420amp hours of back up and two 130watt panels charging the system. The cable run is 70 meters. I went up yesterday and pulled a new cable without the splice. My fear is with heavy rain here water might have made it into the splice. The bh is down again today so I have to go back to take a look but I have to back pack in and it's raining so...

I believe the power source should be fine as it powered two motorola canopy units for more than a year before being redeployed to power mikrotik equip.

the amp rating of the poe is .8amps

There's really no other reason for the RB to be crashing then. Replacing the cable was a good idea, though I'm surprised it didn't fix it.

How does the RB fail? Can you ping the unit through the ethernet connection when it's hanged? How soon does it happen after a reboot? Can you define the traffic load necessary to cause the router to hang?

It may not have failed in that sense. The only connection to the rig is via the bh as it is 100% off grid in mountains. We are experimenting with going vpol or hpol. It is now set to vpol and it seems that ground absorption is probably keeping the link down. (It was up and showing 75dbi when I was up there the other day).

As a side note we decided to try vpol as we could a get a 5.8 omni to receive from the bh and also broadcast to cpe's. Looks like we are going to go back to 5.8 hpol arc panels and put a second radio, probably 900mhz to feed the clients at this point. I'll post back after I make the changes once the rain stops in a few days.

Since changing the cabling, and while I was on site, the rb411ah stayed on. In the end I suspect my root issue was the splice and water.

We have done 24V poe runs of over 400 feet. The key to success is using high quality cable (no CCA or CCS crap) and no splices. Using a 24V 2A POE injector also helps. If all else fails, use a 50V POE injector with the UBNT instant-poe adapter for dropping the voltage to 16V at the CPE.

http://www.calculator.net/voltage-drop-calculator.html

There are many of these available on the net - google a bit...

Excellent. Many thanks. That helped out.