WS-8-150-DC 48v rails dying

[mrprimo]

Hi Guys,

I have dozens of netonix switches out in the field but I have been having big issues with the new WS-8-150-DC and the 48V bus dying.

3 different sites this has happened too now, no storms, no power surges, all other netonix models are fine.

Each time this has happened i just load the config onto a new one and replace it, I have sent two back to our supplier already thinking it might have just been a coincidence, but now a 3rd makes me think there is a hardware issue?

[sirhc]

That's not how the switch works, if the 48V was burned out the switch would be dead.

It takes whatever voltage you feed it and creates 48V internally then steps it down 24V if needed and also steps the 48V down to 3V and 1.5V to run the switch core.

If your burning out 48V ports this is because something is wrong with your tower grounding or something else causing this damage to the switch and it is obvious something is wrong with your setup as you are already on your third switch.

What your doing is burning out the port MOSFET circuit or something.

The chance if you getting 3 defective switches is ridiculously high, something is wrong with your site setup and you need to address is.

We tell dealers to refer customers to us for RMA as most RMAs are from poor grounding or what have you and we are able to diagnose the switch and help the customer resolve the issue that is frying equipment.

The fact that your dealer took the units back is up to them but if we determine the unit is damaged not a failure we charge to repair it.

What do you think the odds are that you got 3 defective units in a row or does it make more sense something your doing or not doing is burning through switches?

I suggest you post up some pictures of this site and detailed information on your power source and grounding systems and let us help you find out what is damaging these units.

[mrprimo]

It was 3 different and separate sites, no ground issues, etc
turns out this is not hardware issue at all and more a software issue, the 48v ports were able to be turned back on once we got the units back to the workshop and I sat down and factory reset them and went over everything with a fine tooth comb and simulated our setups.

Traditionally we have always used the ws-12-250-dc switches on our sites which i now release run different background scripting or software when it comes to the power settings compared to the new ws-8-150-dc?

In the 250 models there is a Warning Voltage, Hibernate Switch Voltage, and a Wakeup Switch Voltage which I set to be 24.5, 23.4 and 23.5.

Running a 24V system with a float voltage of 27.3V means the netonix always has 27.3v but when there is a power cut and it runs off the batteries this will slowly drop, and when it hits 24.5 it sends us an email to warn of low voltage, then at 23.4v it will turn off, and once it goes back above 23.5v it will turn back on - this works awesomely, we have enough time to get to site with a generator, etc etc

But I was assuming the ws-8-150-dc worked the same as above I was setting the warning voltage on those to be 24.5v, and the other settings are not available so that was all I set. As soon as the switch hit this voltage it was turning off all the 48v ports and saying it was low voltage.

So why do the 2 different switches act differently in this situation?

[sirhc]

Because the WS-12-250-DC is a more expensive 2 board setup

Since it is 2 boards we put a separate LOW POWER CONSUMPTION CPU on the DC power supply board and the switch board talks to the power supply via I2C so the WS-12-250-DC can actually shut down the switch board and then wake it up.

People Wanted a smaller less expensive version so we created the WS-8-150-DC

We are thinking of putting back in the hibernate and wake up function but it will be simulated compared to the WS-12-250-DC.

The WS-12 completely shuts down the switch if told too and just monitors the input voltage for wake up consuming about 0.1 watts. With the WS-8 we will turn off as many parts of the switch as we can such as fan, PHYs, and LEDs but we will be around 1+/- watt of power consumption.

This comes down to people wanting a Porsche but only wanting to buy a Volkswagen then asking why they can not corner at 120 Kph

[mrprimo]

Ah that explains it all, I think having those two extra modes simulated for us will be fine as we just want an email to let us know the mains power has gone out, and when it is running low so we can go onsite and changed batteries or put in a generator.

With it turning off the 48v ports, we loose the router and our uplinks (mimosa) and they dont turn back on when voltage is up again, it requires a truck roll, thats all im worried about is my time, some of my sites take over an hour to drive too, sounds like I need to stick to the ws-12-250-dc models as there is not much cost difference between them.

[mrprimo]

Hi again, when will this likely be available is it a time or money limitation, can I help in any way to speed it up?

[sirhc]

Hi again, when will this likely be available is it a time or money limitation, can I help in any way to speed it up?

Could be a few weeks.

In the mean time do you not have an open port? If you have an open port that is not used then turn the POE on with a low voltage shut off as it will also alert you the same way that the input voltage has dropped to your warning level.