erronious wattage reporting and then

[LRL]

Here the history. I was running this ws-12-250A on 1.0.8 at our core for distribution. One day I noticed the wattage was way off. The switch showed both of the AF24's as only drawing 11watts and it showed the AF5 as drawing 9 watts. My rockets were showing as 2.4 watts. The total wattage was 56watts. This switch has 2-AF24, 1-AF5, 1-AF5X, 4-rocket M5s.

I figured a reboot was in order, but if I was scheduling downtime I wanted to upgrade. I upgraded to 1.2.2 and rebooted the switch several times (via software, and did a "reload cold"), but still the low wattage was being reported. So I drove to the head-end and preformed a power cycle. Now the switch is showing what I believe to be correct wattage on all of the ports except port2, an AF24. On port2 it reports 0 and occasionally .5watt.

Hardware, or software issue here?

7-12-2015 12-00-01 AM.jpg (221.19 KiB) Viewed 6247 times

[LRL]

I forgot. The errors on both AF24's are CRC and I cannot explain them. We've never had an issue with the cables and it seems odd to have it with both at the same time. Cable diag doesn't show anything on port3, but port2 show an abnormal termination on pair 2.

Code: Select all

 Pair 1 Status  Ok
Pair 2 Status  Abnormal termination
Pair 3 Status  Ok
Pair 4 Status  Ok

Pair 1 Length
226.38ft /69m

Pair 2 Length
3.28ft / 1m

Pair 3 Length
226.38ft / 69m

Pair 4 Length
226.38 ft / 69m

I have a spare ws-12-250a and I'm considering changing it out.

[sirhc]

Hardware issue, if it was software everyone would be seeing this.

That is odd, did you try moving the unit from port 2 to port 1 and see if the problem follows the wire?

I have only seen the current reported wrong once and that was my own WISP's WISP Switch but only after it took a lightening hit on a tower.

The other night we had a storm that took out the first radio in 4 years on one of my towers which was our MAIN airFIBER24 and sadly it also took out the WS-24-400A, well sort of.

The airFIBER 24 was completely fried, dead, non responsive.

The switch was still functioning and all other radios including 2 more airFIBERS were still running but the switch was reporting incorrect current draw on the remaining 2 airFIBER links.

viewtopic.php?f=17&t=662&p=4777&hilit=lightening#p4777

[LRL]

I've ran a cable check on all radios and only one appears strange. But strange is subjective because I'm not 100% sure what an af5X test should look like.

Pair 1 Status Abnormal termination
Pair 2 Status Abnormal cross-pair coupling, pair 1
Pair 3 Status Ok
Pair 4 Status Ok

Pair 1 Length 236.22 ft / 72 m
Pair 2 Length 236.22 ft / 72 m
Pair 3 Length 236.22 ft / 72 m
Pair 4 Length 236.22 ft / 72 m

[sirhc]

I've ran a cable check on all radios and only one appears strange. But strange is subjective because I'm not 100% sure what an af5X test should look like.

Pair 1 Status Abnormal termination
Pair 2 Status Abnormal cross-pair coupling, pair 1
Pair 3 Status Ok
Pair 4 Status Ok

Pair 1 Length 236.22 ft / 72 m
Pair 2 Length 236.22 ft / 72 m
Pair 3 Length 236.22 ft / 72 m
Pair 4 Length 236.22 ft / 72 m

Assuming this device is powered up (Cable diagnostics report differently when not powered up) these are the things that stand out as bad and WORSE.

BAD:
Pair 1 Status Abnormal termination
This is normally from a bad cable or poorly crimped end?

WORSE:
Pair 2 Status Abnormal cross-pair coupling, pair 1
This is BAD because pair 1 is POS and pair 2 is NEG!

I do have some AF5X but they are not yet deployed so I will check one tomorrow to see what the cable diagnostics report but something does not look good there?

[sirhc]

Another thing that concerns me is your powering 3 AF24 radios at 50 Watts each

up to 20 Watts base line for switch board and fans

3 AF24 at 50 each = 150 watts

AF5X which is on port 9, if this is an older rev board port 9 only has a .5A Polyfuse which means you are living on the edge (limit) but 12+ watt average

4 other 24V radios average 6+ watts each so 18+ watts average

So in conclusion this switch is powering 20 + 150 + 12 + 18 =200 watts

You should not push a switch power supply more then 80% of capacity all day every day especially in hot environments so 250 watts x .8 = 200 watts

So I would not power more devices with this switch.

What type of temperature environment is this switch in?

[LRL]

This guy is in our data center, ambient air is kept at 68f.

Now that the switch is reporting again, the switch shows a total of 180w, not sure if that includes the baseline of 20 for the switch. It has been my intention to split the radio's up between two switches, just for redundancy purposes.

I'm concerned about the af5x on port 9 and to be honest that is why I noticed the 1/5 wattage reports as soon as I did. But since I'm planning to deploy quite a few af5x's I wanted to see how well things went on a .5 port. Over an hour I show the AF5X's average watt draw to be 9.7 and I don't see any peaks over 11.2. I checked the cable with our fluke and it appears fine. Pair 1 shows a 1dB drop compared to all the other pairs but that's it.

The diag doesn't stay consistent on the af5x. Sometimes it reads as ok across the board and other times abnormal across the board or a mix. Since the af5x is intended to be powered using all 4 pairs could this be causing some of the findings? There would be some resistance even across the + and - pairs.

I switched the AF24 to port1 and the cable tests fine with both with our fluke and now with the WS. Any device I plug into port 2 shows abnormal on pair 2.

We've definitely had some really bad storms through here lately so I'm certainly suspect of weather related issues and certainly don't expect you to warranty something like that. If I send this guy back do you think you guys can repair the port and could I trust the switch afterwords? I'm definitely going to replace it, but if it's fixable I'd rather do that than trash it.

[sirhc]

Well the older Rev A,B,C WS-12-250A and WS-24-400A boards have .75A Poly Fuses on ports 1,2,3,4 and .5 on the rest.

All Rev D board on all models have .75A on all ports.


.5A standard 24V is fine for AF5X with cable lengths up to 60 feet but if the switch is in a cool room like yours you could probably push 100 feet or more because remember the polyfuse trips on temperature?

.75A standard 24V is fine for AF5X with cable lengths up to 160+/- feet more in a cool room.

The WS-12-250-AC offers 24VH on ports 1,2,3,4 which will drive an AF5X up to 300 feet of cable but if you use a port in any switch that has .75A for standard 24V you are fine up to 160 feet. All switches even RevA boards offer some .75A 24V standard POE.

With that said we do offer board repair $50 minimum and yes repaired boards are tested fully.

The switch reports TOTAL usage including board and FAN so if you are reporting 200 or below you are fine especially if it is in a cool room like you say. If the room is that cool you could push it close to 250 with confidence. 80% is suggested MAX for non conditioned boxes that get HOT.