Netonix Boot Loop Issue When Powering ePMP 3000L Radios

[waldo.g@level-7.co.za]

Hi Netonix Support,
We are experiencing a recurring issue across multiple sites where our Netonix switches enter a boot loop when powering Cambium ePMP 3000L (3KL) radios on 48V passive PoE ports.
Here are the key points:
The issue occurs only when powering the ePMP 3KL radios.
The Netonix switches reboot continuously as soon as the PoE port powering the ePMP 3KL is enabled.
We've attempted to downgrade the firmware on the affected switches to rule out a software issue, but this has not resolved the problem.
Other devices, including Ubiquiti sectors and radios, are powering up fine on the same ports and voltage settings, which suggests the problem is isolated to powering the ePMP 3KL units.
We are using 24V passive PoE, matching the requirements of the ePMP 3KL.
Could you please advise if this is a known compatibility issue or if there are specific PoE settings or firmware versions you recommend for stable operation with Cambium radios?
We can provide serial numbers, logs, and further site details if needed.

[sirhc]

The ePMP 3000 series, including the 3000L, is polarity agnostic and can be powered with either standard PoE (4, 5, 7, 8) or alternative PoE (1, 2, 3, 6). The 3000L, in particular, requires 24V PoE and will not work with 48V. The ePMP 3000 AP (not the 3000L) uses 48V PoE.
Here's a more detailed breakdown:
ePMP 3000L:
Voltage: 24V.
Polarity: Agnostic (works with both standard and alternative PoE configurations).
Standard Pinout: + on 4/5, - on 7/8.
Important: Do not attempt to power with 48V.
ePMP 3000 AP:
Voltage: 48V.
Polarity: Standard pinout is expected.
Important Considerations:
Checking PoE Pinout: When connecting an ePMP 3000L, verify the PoE pinout of your switch and ensure it matches the 3000L's requirements.
Voltage Limits: The 3000L should not be powered by voltages higher than 30V.
Crossover Cables: If necessary, use a 5-pin to 7-pin cross-over cable adapter for compatibility with certain power supplies.

So unlike the emp 3000 which you can not power the 4500 you "should".

Firmware has nothing to do with it, use v1.5.25

Your details are lacking to say the least.

Model netonix switch.
Verify cable is crimped correctly.
Can you achieve a 1g connection on the cable or have you used a tester.

If DC switch you could be having inadequate power issues.

If AC switch and your below power supply rating you could have bad/damaged power supply or switch.

Have you benched tested this configuration before throwing out into production?

Could be bad radio?

The list goes on.

I would bench test same model switch with radio first make sure no issues.

Test cable.

If DC switch make sure you have enough power.

You could try turning everything else off then power up 4500.

Tons of things to diagnose. Since I'm not on site I'm just explaining things "I" would do to diagnose the problem.

Test that port with a jumper and a UBNT RADIO. Make sure you can turn on and off in UI.

Verify 1G link to laptop.

But firmware is NOT the issue, something else is wrong. Do some testing.

Test that port with

[gunther_01]

Having the same thing happen here.

Replaced radios, a day later. Same thing..

What would cause the netonix to reboot. Without any log entries off the serial port? A hardware only issue?? I see the output power reach 1.7 watts, and then reboot..

[gunther_01]

adding to this.

When we plug in a 3kl, sometimes the switch reboots. Twice it seems. Then it may or may not be accessible.
Other times. If we plug in a 3kl, the switch works just fine, powers up the 3kl, and the radio works fine.


What can cause this? It seems like it's a combination of the switch and or the radio. Not just one or the other at this point..

This is a bench test now. But, having this issue on a live tower daily. And it's causing a lot of issues...

[Julian]

I bet, friend Gunther.

Okay, so, we're using how many 3kls and on what switch? Typically, we ask for a (redacted if necessary) screen capture of the device status page with support requests, saves a lot of this back and forth telephony.

With the heater running, the 3kls each pull ~15W.
the 24v supply in a 12-250 maxes at 6A for 144w. 135 if you count the fan/system draw against it (which we should), 12 3kls chatting full bore mid-winter is about 192w, well into ws-26-* territory.

Overdrawing the 24v supply would reboot the switch since that feeds the rest of the power rails, but hardware faults can cause them to underperform; I load test them to 120w and observe ripple with an oscilloscope here, you should see like 30-40mV p-p on a properly performing supply at that load

[gunther_01]

on the bench.. 2 radios..

On the tower 4..

I have never seen them hit 15w. only 4 while testing on the bench..
12 port dc model. With a 10amp 48volt supply using 12 awg of about 150'.. Under normal operation, well under any port or total wattage maximums..

[Julian]

Hot plugging is not recommended at any voltage, under any circumstance, and may void your warranty. If you want the radios to draw that much you have to actually utilize them. And now that I've been thoroughly argued with, I have a little boy who wants help carving a jack-o-lantern.

I'm going to skip to the end and tell you rma.netonix.com/rma_create is that way if you'd like any further help.

[gunther_01]

These are now new out of the box... So, your just saying that if you hot plug a POE switch that you're out thousands of dollars now? WTH is that..?

[Julian]

if you blow the port or the switch doing so, yes, sir, you are responsible for the repair. - I will say, though, you're not out thousands of dollars if you have me fix the silly thing instead of tossing it; I average under $150 on a 12-port repair.

Say for example you buy a Rolex and go mountain biking with it on, during which time your watch meets some rocks and is dropped into a river, kicked into a ravine, and then finally fished back out with a magnet. You're saying Rolex should cover that?

[gunther_01]

The ports don't seem to be blown. Unless you are saying that sporadic rebooting is a port issue?

This has been on going for a week. We think it's the radio, replace it. Then a different one acts that way.. try a different port it might work, it might not. Try a new radio it works.

Take previous bad radio, and it works on a bench switch. Or it might not. There is no pattern. We are just replacing parts, for no rhyme or reason.

What can break in a 3kl radio that would cause the whole switch to reboot? And could something break in a 3kl that could cause random reboots of the switch??