Ok, that's a better explanation. Generally, the thing that fries when you plug PoE into a non-PoE device is the termination resistors. Whether they fry or not really depends on how they're wired....
Traditionally ethernet is terminated through a 75-ohm resistor from the center tap of each pair to a common point. This common point is then tied to ground through a capacitor. These are usually at best 1/4W resistors. When you plug PoE into a device, what happens is that you now have power in through one pair, across one of the 75-ohm resistors to the common point and then through a second 75-ohm resistor back to the return power wire. This gives you ~150 ohms across your PoE power source. Assuming a best case scenario where you only have a 24V power source, this 150 ohms results in only 0.16A of current, but 3.84W. Or 1.92W per resistor. This will cause those 1/4W resistors to get very hot, and one way or another will fairly quickly no longer pass any current at all. Assuming that nothing (other than the resistor) gets permanently damaged, often this even will just result in the resistor no longer being in the circuit - and the device will magically become PoE compatible. Note I'm NOT recommending this, just stating my experience. If you want to find this termination before melting/burning up a device, simply measure between all combinations of pairs with a DMM set to ohms (I'd probably twist the two wires in a pair together for this test). If you find a reading of <100K or so, and definitely if you find a reading of under 1000 ohms, then you have a non-compatible termination. I.E. measure from the blue pair to all other pairs (one at a time). Then from the green pair to all other pairs. And so on. Newer devices which are designed by people who understand PoE either don't have any termination resistors or they've added a capacitor to each resistor individually to block dc from flowing between the pairs. One other caveat is that I've seen designers just ground unused pairs or center taps, so it might make sense to also check connectivity to any connector shields or chassis grounds. On Thu, Jun 28, 2018 at 11:18 AM, Christopher Gray < [email protected]> wrote: > In this case, I'm actually checking to see what ports (that are not used > for PoE) can receive PoE voltages without being fried. > > It seems if there is a set of magnetics installed without connected center > taps (I think this is the isolation configuration), then you can run > standard PoE voltages into a port without damaging it. > > > > On Thu, Jun 28, 2018 at 1:02 PM Forrest Christian (List Account) < > [email protected]> wrote: > >> If that's the case, I totally agree with you... >> >> Electrically test each individual wire to the chassis/shield/ground >> lug/coax connector outsides, making sure there's no conductivity (with the >> caveat that there may be capacitors to ground which will show brief moments >> of connectivity as you initially hook it up). >> >> Then, power on the radio, using an isolated power supply. Repeat the >> above tests using a voltmeter instead - you shouldn't see any meaningful >> voltages. >> >> I'm in the process of trying to gather this type of information for each >> common radio/poe device. It's slow going and there are of course other >> things which always seem to be higher priority. >> >> On Thu, Jun 28, 2018 at 10:56 AM, <[email protected]> wrote: >> >>> The conversation originally started with –48 power sources and was it >>> safe to power an ethernet device off of such a supply. >>> Is the power galvanically isolated from any ground/shield type of >>> connections on the device. >>> >>> *From:* Forrest Christian (List Account) >>> *Sent:* Thursday, June 28, 2018 10:36 AM >>> *To:* AnimalFarm Microwave Users Group >>> *Subject:* Re: [AFMUG] How To Test Port For Magnetics? >>> >>> Could you clarify a bit what you mean by isolate? >>> >>> This could mean: >>> >>> The PoE and the lan sides are electrically separate (i.e. there's a set >>> of ethernet magnetics in it). >>> or >>> The power supply is electrically isolated on each port >>> or >>> There is no ground connection >>> or >>> ..... >>> >>> When you start talking PoE there are lots of ways to arrange the power >>> injection. For instance, on all gigabit power injectors that packetflux >>> currently makes, the LAN side of the injector is totally electrically >>> isolated from everything else. The PoE side has pins tied to either power >>> or return/neutral depending on how jumpers get set, with the note that by >>> 'tied' means that we're connecting the center tap on the magnetics to those >>> pins. >>> >>> Other injectors (and perhaps upcoming ones from packetflux) have fully >>> isolated DC power on the PoE, so there isn't any direct connection from the >>> PoE port to the source power supply. >>> >>> >>> >>> On Thu, Jun 28, 2018 at 10:00 AM, Christopher Gray < >>> [email protected]> wrote: >>> >>>> Chuck (Well, directed at Chuck, but interested to hear if anyone else >>>> has ideas), >>>> >>>> What would you consider a reliable method for checking a port for >>>> magnets / Ethernet transformers / isolation modules? Is it just a matter of >>>> testing the individual pairs for resistance or continuity? >>>> >>>> Every time I want to know if a port will isolate PoE on its own, I just >>>> crack the device open, and look. I'd like to be able to test externally if >>>> possible. >>>> >>>> Thanks for the help - Chris >>>> >>>> *Christopher R. Gray* >>>> Direct: 856-472-9733 >>>> [email protected] >>>> ------------------------------ >>>> *GrayTech Software, Inc* | 630-682-4030 | www. >>>> graytechsoftware.com >>>> ------------------------------ >>>> >>>> -- >>>> AF mailing list >>>> [email protected] >>>> http://af.afmug.com/mailman/listinfo/af_af.afmug.com >>>> >>>> >>> >>> >>> -- >>> *Forrest Christian* *CEO**, PacketFlux Technologies, Inc.* >>> Tel: 406-449-3345 | Address: 3577 Countryside Road, Helena, MT 59602 >>> [email protected] | http://www.packetflux.com >>> <http://www.linkedin.com/in/fwchristian> >>> <http://facebook.com/packetflux> <http://twitter.com/@packetflux> >>> >>> ------------------------------ >>> -- >>> AF mailing list >>> [email protected] >>> http://af.afmug.com/mailman/listinfo/af_af.afmug.com >>> >>> >>> -- >>> AF mailing list >>> [email protected] >>> http://af.afmug.com/mailman/listinfo/af_af.afmug.com >>> >>> >> >> >> -- >> *Forrest Christian* *CEO**, PacketFlux Technologies, Inc.* >> Tel: 406-449-3345 | Address: 3577 Countryside Road, Helena, MT 59602 >> [email protected] | http://www.packetflux.com >> <http://www.linkedin.com/in/fwchristian> >> <http://facebook.com/packetflux> <http://twitter.com/@packetflux> >> >> -- >> AF mailing list >> [email protected] >> http://af.afmug.com/mailman/listinfo/af_af.afmug.com >> > > -- > AF mailing list > [email protected] > http://af.afmug.com/mailman/listinfo/af_af.afmug.com > > -- *Forrest Christian* *CEO**, PacketFlux Technologies, Inc.* Tel: 406-449-3345 | Address: 3577 Countryside Road, Helena, MT 59602 [email protected] | http://www.packetflux.com <http://www.linkedin.com/in/fwchristian> <http://facebook.com/packetflux> <http://twitter.com/@packetflux>
-- AF mailing list [email protected] http://af.afmug.com/mailman/listinfo/af_af.afmug.com
