Hello to all,
I have Dlinks POE SWITCH and LINKSYS POE SWITCH but cannot give the alimentation.
I must make a different color of rj45 cable?
Please help me.
best regards
good luck, PoE switches are generally only 48v, where the 333 is max 24v
I wish all the boards were 48v as that is the voltage for true 802.3af PoE devices, but for whatever reason, they felt like going 24v instead…
It’s a pretty big headache that it seems like every board they make has different power specs. I don’t understand why they don’t make them all 9-48 volts and support true poe. Or at least with all the higher end boards. I can understand not wanting to on the 133c because of the low cost, but how much does it really raise the price?
In my experience, an isolated dc-dc power supply is more complex and expensive than a non-isolated one.
Making a 9-48V dc-dc non-isolated is not expensive or complicated
Making 9-24V dc-dc non-isolated is marginally cheaper, because lower V-rated capacitors required, and easier to design also.
Isolated dc-dc converters require transformers and feedback circuits which increase cost and complexity.
Also, “telecom spec” 48V is usually 36-72V range, to allow for the voltage range of the telecom battery load/charge conditions. 802.3af is nominally 48V but rated with a much tighter voltage range, beacause it’s assumed that the power is coming from a switch mode supply, not a battery. Most 802.3af equipment can’t withstand +72V or anywhere like that.
AFAIK the 802.3af POE spec does not require ground isolation (i.e. DC 0V floating relative to mains earth); but many real-world deployments do, for example if the radio is grounded on a tower. A proprietary POE with separate power “brick” can be isolated (floating 0V relative to earth) - but note: not all are - which can solve this problem.
Another point, an external POE “injector” can (potentially) have surge suppression and save the Ethernet switch ports in case of lightning strike. I don’t think the common injectors do have suppression though, so you have to add surge protectors to the CAT5 cable anyway, or put your switch or customer at risk.
Also note, 802.3af is only valid up to ~13W load. Above that, it is proprietary anyway (until the new PoE-Plus standard - a.k.a 802.3at - is finalised). High power radio cards and fast CPUs are great users of power, which can exceed the 13W limit and hence today, proprietary POE is required.
At 48V, longer cable runs can be used than at lower voltages - especially important for high power equipment because of the V=IR “drop” across the cable, which is generally worst-case at “power up” with most devices.
http://en.wikipedia.org/wiki/Power_over_Ethernet
is a good read.
Hope some of that helps
Regards
24v produces less interference in some instances.
http://forum.mikrotik.com/t/interference-problems-with-police-band-radios/12585/1
that is an inaccurate description. The transformer that MikroTik uses on the 112’s and the 532’s causes a ~50mhz harmonic in the lower bands, even at 24v, however the higher the input voltage, the stonger the interfering harmonic. this has been documented by numerous people on the forums before.
So you are partially correct, but…
The “interference” from 48v is actually a poorly designed / low quality transformer on the board, if MikroTik used a higher quality transformer, that problem would go away (so I’m told, I’m no EE). There are also some posts I’ve seen before about different types of PoE injectors that help greatly reduce the harmonic, even at 48v.
That makes a lot of sense. Thank you very much for the correction.
any time… 
Also note that there are POE converters which will convert standard 802.11af to 12 and 24 volts.
These can be used on the tower with the radio and fed by 802.11af from the ground.
There are also poe splitters for powering multiple boards.