Netonix Forums

This must be something simple but I need the hive mind to help... Brand new WS-12-DC switch. Going half way up our tower and being fed by fiber along with a 12 gauge DC cable. Total cable distance from bottom switch (a Netonix AC model) and power supply is 60 feet.

When I test the WS-12-DC in the tower shack using the DC power supply and a short fiber, everything is great. I then move the WS-12-DC into its metal enclosure on the tower - ground it and power it. It powers up and my laptop with a short Ethernet cable can connect to it locally just fine.

However, the fiber never syncs up - no matter what I do. So, I then use a 65' Ethernet cable that is running up the tower (spare for radios) to connect the switches together to trouble shoot the fiber without having to hang in the air. That won't link up either. I take that very same 65' Ethernet and plug it into my laptop - links right up...

So, I get two separate SFP ports that won't link fiber and a long Ethernet cable that won't link up when the switch is up on the tower. I take the switch down to the bottom, use the same power supply (with a short cable) and the SFPs both work fine (on short fiber) and Ethernet works fine.

So, not a bad switch since I can make it work somewhere. I did not haul a volt meter up the tower (left it at home) but when I did log into it with my laptop up on the tower, all status -> voltages were green.

Any ideas? Fiber tests good and the 65' Ethernet tests good and will power radios on the tower but I can't get anything other than my short laptop cable to talk to the switch as soon as it is mounted on the tower... New one for me... Same type of power supply is powering a WS-8-150-DC at the very top of our tower (125' up) and it works great. I need the four 48VH ports of the WS-12-DC at this location so I can't swap this switch with a WS-8-150-DC.

So switch works fine on ground not on tower.

Using an existing Ethernet cable on the tower to the ground you can not link up.
Although you can use laptop and short ethernet cable and link right up with it on the tower.
My guess here is your Ethernet cable or ends going up the tower is bad?

Fibers link with short jumper on the ground but not with long jumper on the tower.
The fiber may test out good but if the cable splicer did not FLIP Tx to Rx when putting the Ends on it will not link.
Fiber needs to be cross over from Tx to Rx and Rx to Tx

What are the ODDs of both being at fault........who knows.
Could also be a Fiber module that has weak laser or weak sensor?

All just guesses.

I did try flipping polarity on the fiber. We have 4 strands running up to this switch and it's a brand new armored cable protected in a conduit. I have also changed out SFP's. I have a pile of them that work with Netonix. Been through 6 in this switch - and they work in others...

My only thought was power. Am I loosing too much power for some reason and this switch will turn on but won't fire up the lasers in the fiber modules?? Can't really explain the Ethernet cable issue unless it takes slightly more juice to sync up an Ethernet port on a longer cable run than a short one.

Heading back out next week with a volt meter to see exactly what is hitting the switch.

cwachs wrote:I did try flipping polarity on the fiber. We have 4 strands running up to this switch and it's a brand new armored cable protected in a conduit. I have also changed out SFP's. I have a pile of them that work with Netonix. Been through 6 in this switch - and they work in others...

My only thought was power. Am I loosing too much power for some reason and this switch will turn on but won't fire up the lasers in the fiber modules?? Can't really explain the Ethernet cable issue unless it takes slightly more juice to sync up an Ethernet port on a longer cable run than a short one.

Heading back out next week with a voltmeter to see exactly what is hitting the switch.

You had your laptop hooked up to the switch on the tower right?
Simply go to the Device/Status Tab and look at 48V Board voltage that is the voltage in.

How many volts are you putting into the bottom?

The WS-12-DC wants 48-54V input, will turn on at less than 48V, down to 26V but below 43V the 48V POE options are locked to OFF.

The WS-12-250-DC will take 9V to 72V but to be honest 12AWG 60 feet is not really heavy enough for lower voltages, remember amps goes up as voltage goes down watts stay the same.

You said you wanted (4) 48VH ports, if powering (4) AF24 that is 200 watts + 12W for switch and SFPs so 212 watts / say 48V reaching the switch means your carrying 212 / 48 = 4.42 Amps which is a LOT of watts for 12 AWG
Now if you were using say a WS-12-250-DC and putting in 24V at switch then 212W / 24V = 8.83A which is really way too much for 12AWG
People really need to do their math when transporting DC over Distance or melted cables and fires could result

I'm 3 x 48VH (2x DragonWave and 1 x AF24) and nothing else powered. That's why I went with 48v powered switch. Should be OK with 12g cable. I have an adjustable power supply so I can go up to the switch maximum of 54v. But, I need to get the switch talking first....

The WS-12-250 at the top is powering 3 ePMP APs which take very little power and I'm hitting that switch with 48v as well. Still fine with 12g cable.

Do keep in mind that DC voltage drop on long runs affect both polarities, so the ground side will not be at 0V. At that point the protection ground bonding also becomes a factor. How is the switch chassis bonded to protection ground and how is tower ground bonded to shack ground?

To answer your ground questions, we have a dedicated ground going up the tower and the tower itself is grounded to the same ground as the shack (and our dedicated ground).

Is the switch chassis bonded to protection ground?

Yes, it is. It is bonded to the ground wire we run up the tower for the radios.