Hi.
This is sort of a How-to/FYI post. Over the last two years, 4 out of 12 MAX HD1 Dome Pro (HW1) devices have died suddenly, they won’t power on when injecting PoE via the attached splitter or directly. All devices are out of warranty of course, our supplier wasn’t able to help either. I am puzzled how this failure rate is even possible, I assumed some sort of power surge at first but the conditions of the devices were always different and we were always using protective equipment like surge protectors.
So the problem renders as follows:
- Plug in the device to PSE, a MirkoTik RB5009 router in this case
- PoE indicator LED on the MirkoTik turns on briefly, status of the used interface states “current too low” on the router’s side
- PoE indicator LED turns of
- Rinse and repeat. No indicator lights at all on the Peplink
I suspected something is broken at the PoE power stage on the PCB, like a TVS diode or something else. So I grabbed a good device and compared some measurements with a DMM but without a conclusion. After many hours of trying to understand the power stage of the PCB (and also later with the help of AI) and delving into the issue, I had some key aspects and a suspicion:
- the boards main DC voltage is 12V, delivered by flyback converter (marked as T3 on the board, LinkCom LDT6013)
- the signature resistance (25kΩ) for the PSE is present, explaining the MirkoTik trying to deliver power
- the component marked as U82 on the board’s back side (a KTA1137A CMOS for surge protection and PoE-PD controller) might be fried
While waiting for the delivery of some new KTA1137A chips, I thought trying to apply 12V manually might be worth a shot. No big surprise - the device turned on and was working just fine. What was even weirder: after applying power manually for a couple of minutes the device powered on through PoE! I tried this method with all broken devices, it was the exact same symptom and “fix” for all of them.
So the actual “fix” is to apply 12V limited at 1A manually with a lab power supply. I soldered a short tap to a pin of the flyback converter “T3”. I assume it’s pin 10, but there’s no real datasheet available for the component. Ground clamped to the screw next to it marked as “G1”. Let the device sit powered for around 5 minutes and it’s done.
So what’s the real issue here? Later I realized the broken devices were dormant/not powered for a while, so could this be some sort of capacitor leakage? I don’t know what Peplink might have changed with HW2 but hopefully this issue was addressed.
