WS-12-250-DC Going offline for several hours

[Dawizman]

I have a WS-12-250-DC at a remote solar site. The site is running at 48V. The switch has been going offline for the last few days whenever the voltage is getting above 56V, and then comes back online after the sun goes down. Now I was under the impression that these switched were good for 60V? Anyway, I'm getting nothing showing in the logs whatsoever, other than the normal bootup stuff. Nothing that indicates why it might be shutting down. I have the switch setup to send everything to a syslog server. Running firmware version 1.3.7.

The other strange thing that happens, it that when the switch comes back up, port 5 has the POE disabled instead of being set to 24V.

Any Netonix folks. have some suggestions?

Thanks!

[sirhc]

They are good to 60V, maybe go on site and check voltage with a multimeter as the unit will shut down if the input voltage goes above 60V.

I would upgrade to v1.3.8 as there were fixes for DC switches.

Please post a screen grab of your Power Tab and Device/Status Tab here in this post.

Also a formatted copy of the Device Log after the upgrade would be nice.

[Dawizman]

We monitor voltage at both the switch and our solar controller. Both are showing voltage right around 56v when the switch goes down. I will give 1.3.8 a try and report back.

[Dawizman]

I'll post the logs and voltage readings next time (if there is a next time) the switch goes down. For now, here's the log output after the upgrade (Nothing really to show:

Code: Select all

Dec 31 17:00:33 syslogd started: BusyBox v1.19.4
Dec 31 17:00:34 dropbear[1366]: Running in background
Dec 31 17:00:35 STP: set port 10 to learning
Dec 31 17:00:35 STP: set port 10 to forwarding
Dec 31 17:00:35 STP: set port 9 to learning
Dec 31 17:00:35 STP: set port 9 to forwarding
Dec 31 17:00:35 STP: set port 8 to learning
Dec 31 17:00:35 STP: set port 8 to forwarding
Dec 31 17:00:36 STP: set port 7 to learning
Dec 31 17:00:36 STP: set port 7 to forwarding
Dec 31 17:00:36 STP: set port 6 to learning
Dec 31 17:00:36 STP: set port 6 to forwarding
Dec 31 17:00:36 STP: set port 5 to learning
Dec 31 17:00:36 STP: set port 5 to forwarding
Dec 31 17:00:37 sysinit: 1037
Dec 31 17:00:38 system: starting ntpclient
Nov 21 20:16:53 switch[1397]: Detected warm boot


[sirhc]

Dave doubled checked (just now), our switches will not shut down until it detects more than 60V.

Is it possible your charger is spiking to above 60V even for a msec when it gets STRONG sunlight because it will shut down at even a smidge over 60V.

[Dawizman]

I suppose anything is possible. We are running a Morningstar Tristar TS-MPPT-60. I will setup live monitoring on it which will record parameters every second, and see if it spikes up at all. With the size of battery back we have (8x 6V 460Ah Rolls Batteries) I figure any voltage rise should be pretty controlled. Either way, I'll look in to it a little bit further.

Thanks for the info.

Cheers!

[sirhc]

I suppose anything is possible. We are running a Morningstar Tristar TS-MPPT-60. I will setup live monitoring on it which will record parameters every second, and see if it spikes up at all. With the size of battery back we have (8x 6V 460Ah Rolls Batteries) I figure any voltage rise should be pretty controlled. Either way, I'll look in to it a little bit further.

Thanks for the info.

Cheers!

If we determine it is hitting the 60V PS limit we "might" be able to bump that a smidge in software.

[bradg]

Or, maybe send out a final logging "gasp" before shutdown, which could then get captured by a central syslog server? At least then it would indicate that it was an "intended" shutdown.

[Dave]

bradg

Or, maybe send out a final logging "gasp" before shutdown, which could then get captured by a central syslog server? At least then it would indicate that it was an "intended" shutdown.

I like that idea also. We will look into it.

Dave

[Dawizman]

So after some digging, our solar controller was indeed over-volting the switch. We weren't polling at a quick enough interval to to see it happen though. The controller was set to an absorption voltage of 57.6V, but what I failed to notice was that is had a temperature correction as well. So that 57.6V was at 25C, and it went up 0.12V every degree below that. So it was actually getting up to 60V at anything below 5C. I took care of that by setting a hard limit of 58.5V in the controller. It wasn't very sunny today though, so I never got to see if it worked.

One thing that still has me scratching me head though, is why is my port 5 poe being disabled every time the switch goes down in this manner? Any thoughts?

Thanks