WS-12-250DC not switching on anymore

[alexishh]

Hello,

We had a WS-12-250DC up for almost one year when a few days ago it turns off for no reasons : all ports off, PWR led off, fan off. Reboot it = no reactions.
We replace it and took it to our lab for some deeper inspection.
We opened it and see some led blinking on the power board but nothing on the switch board (powering with stabilized 24VDC) : D2 is fix on, D3 is fix off, D6 blink 10 times then off for a few seconds, then on for another few sec, then off.
The 48V soket (HDR1) delivers nothing (0V).

When we unplug the 48V soket (so it does not try to power the switch board), the power board seems to "boot" correctly : D2 is fix on, D6 blink and then became fix on and D3 became fix on at the same time.
The 48V soket delivers 51V in that configuration.

We inspect the switch board the best we can without seen anything suspicious, all components seems to be good (visually), no dark/burn marks...
Do you have any advice or ideas of what we can test/repare ?

Thank you.

[sirhc]

You can use a meter to test the board voltages at the test points.

24V
3.3V<= SOC SHORTED/FRIED
2.5V<= SOC SHORTED/FRIED
1.8V<= SOC SHORTED/FRIED
1.0V <= SOC SHORTED/FRIED

If the SOC is shorted/fried then that occurs most often from a couple things:
Over current on a port fries the Ethernet Transformer in such a way to allow POE current back into the PHY/SOC

If you want to pin-point which port is the offender then make the test rig with 2 volt meters as explained in the bench test thread:
viewtopic.php?f=6&t=2780#p19221

But for that to work the 24V test point must read 24V and switch board green led in the front must be on.

It could still be the power supply which you could attempt to power the switch with a 48-53V power source.

Damage to the Ethernet Transformers can not "usually" be seen until you de-solder them and flip them over as seen in example below:

[alexishh]

Hello,

Thanks for the fast answer.

A few tests later, with no surprises :

24V (TP1) = 0V
3.3V (TP3) = 0V
2.5V (TP8) = 0V
1.8V (TP6) = 0V
1.0V (TP4) = 0V

We tested the switch board with an other 48VDC power supply that immediately fell into protection mode (shortcut).

No way to get the power led back on.
Just to be sure, the SoC you are talking about is the one under the heatshink (U15 i think) ?

It will cost us about 40USD to send the switch to RMA, can you tell us a rough estimated cost for the replacement of the SoC and (probably) an Ethernet transformer ?

Thank you.

[sirhc]

If the SOC is fried we do not replace that, it cost more in labor then it is worth, we replace the whole board.

However your SOC may not be fried possibly your 24V power supply is the issue which "sometimes" can be fixed.

You can purchase a new board here for $199.95 but make sure to follow instructions as we NEED the MAC address to pre-program the board.
https://www.netonix.com/replacemet-parts/ws-12-rb.html

If we replace a board in RMA we charge $175.00 (Cheaper than buying board but you have to weigh in shipping I guess)

Most common cause for damage to our switches and all WISP gear is ground current.

99% of all WISP equipment damage is from ground current (I own a WISP)
Another common cause for damage is a shorted cable or incorrect POE option applied to device.

Here are some good posts on grounding:
viewtopic.php?f=30&t=2786#p19279
viewtopic.php?f=30&t=1816
viewtopic.php?f=30&t=188
viewtopic.php?f=17&t=1786&start=30#p13447
viewtopic.php?f=30&t=1429

Read the posts above but the basics are as follow:

Tower ground rods must be bonded to electrical service ground rods HEAVY #2 wire.
I always add 1 or 2 “new” ground rods to older existing electrical services.
In the event bonding service rods to tower rods is not possible then definitely add additional NEW ground rods to electrical service to help prevent service ground current from ever seeing tower ground rods as a better path to ground.

[alexishh]

Thanks.
I suppose the $175 for the RMA does not include the shipping back costs, so we will order you a new card.

Last question : what component makes the 24V on the board ? Maybe we can check it and replace it if it is kind of "standard". Nothing to loose...

[alexishh]

For information : we receive the new board 3 or 4 days after ordering it (nice !), replace it, and everything start up correctly.

The all replacement cost us about 260 euros (294USD) because of the customs taxes (58EUR / 65USD) and bank charges.
(75EUR (84USD) less than a new WS-12-250DC.)

Thanks again for support and fast shipping.

[sirhc]

Well be warned "if" you issues was with ground current and you do nothing to improve your grounding it will simply happen again.

The nice thing about RMA is we tear it apart and find out what the issues is, the daf part s often people do not listen or even read our report and pictures we send them.

Unless you know the failure was something you did add 2 NEW ground rods to electrical service.
Make sure you bond service ground rods with tower ground rods with #2 wire.
No not run copper from switch to another ethernet device on a different service panel or service or a different grounding rod that is not bonded to your grounding system.

[CHTechServ]

Hi, A week ago, we had a freak electrical storm about 2 miles away from the off-grid WISP site. 6 days later, our WS-12-250-DC decided to stop working.
No green front panel light. Tried resetting it per instructions twice, still no front panel green light. I can see several green lights blinking inside the unit.
Happy to do internal test measurements if you give blessing to open and test. The firmware was updated a month ago.
Otherwise, I need to get this R&R'd ASAP or get a replacement ASAP.

Thank you,
Chris

[CHTechServ]

Greetings:

Since my WS-12-250DC was out of warranty, I decided to look under the hood and try some basic troubleshooting. Other than a dead spider carcass under the board, no signs or scent of burn marks anywhere.

Going a bit further, I found the main DC-DC converter green LED's flashing but unable to put out 50.2 VDC due to a shorted 1A 1kV Fast Recovery SM Diode aka "D2" on the WS-12 board.

Geek Tip: I was able to find the shorted D2 using a constant current power supply delivering 2A at 0.1VDC to the WS-12 board and observing the board with a Flir Thermal imager. Took a while, but sure enough D2 was warmer than anything else on the board.

After lifting one end of D2, the short was cleared and after re-assembling, the switch powered up again. It even has the original POE configuration and IP address still loaded as attempts to reset 3 times failed due to lack of power.

Not sure why D2 decided to short out but might have been latent damage from a nearby lightning strike a week prior. DigiKey has replacement part for 46 cents so I will order that next week and replace. I am also adding large snap-on ferrite chokes to each wire or CAT cable that connects to the switch.

Meanwhile, I have the 26 port 400W big brother Netonix switch inbound which will be here shortly along with two more grounding rods as suggested.