Axis Q6034-E Speed Dome cam No Power (teach me about POE)
Good day folks. A friend of mine scored a couple of these Axis speed dome cams which no longer work and asked me to have a look at them, see if I can get it going. He also grabbed 2 POE injectors that used to power them. He said he tried both of them with a functional camera and they work, so the problem is most likely the cameras.
The problem: no power and no response from the cameras at all. The "activity" LED on the injector also blinks amber and according to the injector's manual, this means overload.
To cut the story short, I opened up one of the cameras and extracted the board from the bottom. The same behaviour occurs even with all other connectors unplugged, so I THINK the problem is the POE part on this board. Measuring with a multimeter, I DO see power spiking to 50v for a very brief moment in some places, including directly on the RJ45 connector itself, but this quickly drops out. I checked for shorts but found none, at least not in "big" components.
Sadly, it's also here I realised I have no idea how POE actually works and proceeded to doing some research, but came pretty stuck because it quickly went a bit over my head. I was able to identify two "building-block" DC-DC converters on the board there (large caps+"coilcraft" coils), but these too have no power on the output - just those bursts of 50v on those big input caps. If you're wondering, yes, out of curiosity, I pulled out one of the original caps, but it read perfectly fine on my ESR meter, even after all these years, yet I tried replacing it to see if it makes the least bit of difference - it didn't, so I didn't bother replacing the other one too until I figure out how it actually WORKS, which is the main point of this discussion really !
From the RJ45 connector, I can clearly see 4 traces going to that black "poe+ETH" box thing, which I presume is a transformer like on all NICs/laptops, whatever. From there, the 2 pairs continue to the connector that goes to the moving head, so I assume those are for data and that's where the actual "NIC" of this thing is in the head - not down here on this board. What about power then ? It's not very clear on the photos I've taken, but I see 2 pins directly shorted together (the first ones, depending on your orientation). I assume they are pins 7-8, so I'm expecting 4-5 to ALSO be shorted together, just not visible from underneath (forgot if I actually measured this).
The two ICs that seem to run the whole show are those NCP1081. All the application schematics there show pins 4-5 and 7-8 shorted together, which would confirm what I found above. I doubt this camera uses Gigabit Ethernet, so it leaves the 2 "data" pairs alone (green and orange) and directly shorts the remaining 2 (blue and brown) to carry power.....ALTHOUGH, it could be the black box has center taps on those data lines, so those TOO could "extract" power from the data pins themselves, just so the camera is compatible with both A and B standards, which I read is a requirement (???) There's probably something signaling the injector to stop, not necessarily a short....again, I'm not entirely sure what conditions must be met for it to keep supplying power, besides the transmission itself...Any ideas what I should check ? Should I just replace the NCP chips or try a different injector, despite the guy claiming both of these work ???
The injectors output 55v (label says so), so I was hesitant to connect them to a known good 48v Hikvision camera to it to see if that "port" LED behaves differently or if it's unable to sustain the load and the problem has been on the injector side and not the camera side this whole time !
The problem: no power and no response from the cameras at all. The "activity" LED on the injector also blinks amber and according to the injector's manual, this means overload.
To cut the story short, I opened up one of the cameras and extracted the board from the bottom. The same behaviour occurs even with all other connectors unplugged, so I THINK the problem is the POE part on this board. Measuring with a multimeter, I DO see power spiking to 50v for a very brief moment in some places, including directly on the RJ45 connector itself, but this quickly drops out. I checked for shorts but found none, at least not in "big" components.
Sadly, it's also here I realised I have no idea how POE actually works and proceeded to doing some research, but came pretty stuck because it quickly went a bit over my head. I was able to identify two "building-block" DC-DC converters on the board there (large caps+"coilcraft" coils), but these too have no power on the output - just those bursts of 50v on those big input caps. If you're wondering, yes, out of curiosity, I pulled out one of the original caps, but it read perfectly fine on my ESR meter, even after all these years, yet I tried replacing it to see if it makes the least bit of difference - it didn't, so I didn't bother replacing the other one too until I figure out how it actually WORKS, which is the main point of this discussion really !
From the RJ45 connector, I can clearly see 4 traces going to that black "poe+ETH" box thing, which I presume is a transformer like on all NICs/laptops, whatever. From there, the 2 pairs continue to the connector that goes to the moving head, so I assume those are for data and that's where the actual "NIC" of this thing is in the head - not down here on this board. What about power then ? It's not very clear on the photos I've taken, but I see 2 pins directly shorted together (the first ones, depending on your orientation). I assume they are pins 7-8, so I'm expecting 4-5 to ALSO be shorted together, just not visible from underneath (forgot if I actually measured this).
The two ICs that seem to run the whole show are those NCP1081. All the application schematics there show pins 4-5 and 7-8 shorted together, which would confirm what I found above. I doubt this camera uses Gigabit Ethernet, so it leaves the 2 "data" pairs alone (green and orange) and directly shorts the remaining 2 (blue and brown) to carry power.....ALTHOUGH, it could be the black box has center taps on those data lines, so those TOO could "extract" power from the data pins themselves, just so the camera is compatible with both A and B standards, which I read is a requirement (???) There's probably something signaling the injector to stop, not necessarily a short....again, I'm not entirely sure what conditions must be met for it to keep supplying power, besides the transmission itself...Any ideas what I should check ? Should I just replace the NCP chips or try a different injector, despite the guy claiming both of these work ???
The injectors output 55v (label says so), so I was hesitant to connect them to a known good 48v Hikvision camera to it to see if that "port" LED behaves differently or if it's unable to sustain the load and the problem has been on the injector side and not the camera side this whole time !
Attached Files
Comment