Hi all. I'm considering doing some GPEN with Netpower 7R switches. Simplified description of the topology would be NP7R on poles along a run of aerial fiber and then a combo of aerial runs or vibe plowed runs to homes.
My question is, how does the NP7R handle varying PoE input voltages on the ports? Do it have any backfeed blocking? For example, first pole has 4 homes, all with 48v supplies but varying lengths so it would see 44,45,46,47v from those feeds.
electrically, if there is no protections or diodes, I'm going to try to draw 100% of the power from the 47v line until the volts sag to 46v, then asymmetrically load balance on the 46 and 47v feeds until it hits 45v and then again, asymmetrically load balance between those. Is that the design intent? I'm a bit worried about longer runs that might lose more volts causing a big enough delta that I'm over drawing amps off the 'high' volt supply? say the high supply arrives at 47v but the next one has sagged to 41v... I have to shed 6v in amp draw before I'm using the second feed. I'll also likely be pushing some volts back up that second feed line though I don't know that will be much of an issue.
I'm also considering lashing an ethernet along with the fiber and daisy chaining that with 48v on the line so I have power enough to program and light up the netpower before the first customer is on. again, this makes me wonder what protections are in the device if any.
Re: GPEN power handling ?
I wouldn't count on MT staff to give definite answer to your particular question here, you might be more lucky if asking them directly at support@mikrotik.com .
Forum members can only guess. And ny guess is tgat there are simple diodes on each PoE-in ports which means it will happen exactly the way you described. Which AFAIK is by design, Netpower switches are meant to run even off single live connection.
Netpower 7R specs tell that unit itself consumes max 7W (with attachments 29W but that includes 450mA on PoE-out port). 7W at 35V (to make calculation easier and to cover extreme case) is 0.2A ... which shouldn't be too much for any of supplying devices. Personally I'd avoid using PoE-out on production installations of Netpower devices though.
BTW, all supply voltages will be the same at Netpower as they're at PoE-out ports of "donor devices" when current draw is zero. Voltage drops only after current starts to flow. If all "donor devices" would output same voltage at their PoE-out ports, then Netpower would draw some power from all simultaneously, the amount of current through individual ports would be inversely proportional to voltage drop on UTP cables.
Forum members can only guess. And ny guess is tgat there are simple diodes on each PoE-in ports which means it will happen exactly the way you described. Which AFAIK is by design, Netpower switches are meant to run even off single live connection.
Netpower 7R specs tell that unit itself consumes max 7W (with attachments 29W but that includes 450mA on PoE-out port). 7W at 35V (to make calculation easier and to cover extreme case) is 0.2A ... which shouldn't be too much for any of supplying devices. Personally I'd avoid using PoE-out on production installations of Netpower devices though.
BTW, all supply voltages will be the same at Netpower as they're at PoE-out ports of "donor devices" when current draw is zero. Voltage drops only after current starts to flow. If all "donor devices" would output same voltage at their PoE-out ports, then Netpower would draw some power from all simultaneously, the amount of current through individual ports would be inversely proportional to voltage drop on UTP cables.