Power over ethernet

Quoted:

The 802.3af PoE specification details all of the requirements for designing PoE equipment. Two types of devices are specified in the standard: Power-Sourcing Equipment (PSE) and a Powered Device (PD). The PSE provides 48-V DC power, with a current limit of 350 mA, to the PD?be it a VoIP phone or wireless access point. The PSE is limited to a continuous maximum power output of 15.4 W.

End Quote

If it can provide 48 V at 350MA, you should surely be able to get your done. All of the routers I have seen were normally sized, so unless you are making a device that is not a 1U rack form factor, I cannot see how you would be so concerned.

It's an access control terminal, pcb is about 3 in by 3 in.

The design in the link below is similar to the ones I have been looking at on the LTC and Nat Semi websites

On a sunny day (Thu, 9 Apr 2009 18:53:03 +0100) it happened "IanM" wrote in :

I just put 12V on one end, and a 7805 on the other, worked for more then a year, until I no longer needed it. Not any official standard though.

That alone looks like it would take up more than 3 x 3.

An "access control terminal" sounds more like a device, than a source.

A PSE is a source equipment device. A PD is a powered device in PoE nomenclature.

Perhaps you could simply use a PoE router so the power is generated before your device, and then passes said power downstream.

There IS a standard. 802.3af

Yes it is a PD. The goal is to take whatever it is a PoE router passes down the liine and convert it to 9V @ 400mA.

Put the power conversion inline. There are likely even off the shelf solutions for DC-to-DC power sources that are inline.

Yes, I like that idea. Thanks.

Only applies to 10/100BASE T.

Cheers

There is one for GbE then as well, because we have GbE Poe Cisco routers in our buildings. They are fully addressable ports, which only work when activates via administrative hooks.

John Larkin had found a nice line of little switchers of the "crack out of the package, solder, done" kind. Could be from National. Pricey but if low qty could be ok. TI has some, too.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.

I've designed a PSE (Power Sourcing Equipment) Gb Ethernet interface. From memory, the powered device must correctly identify itself as a POE load (by means of a correct start-up load sequence), if the PSE chipset does not see the correct load then power is switched off. So you can't just connect a DC-DC converter to the end of the line, you need a specifc PoE load chip to take care of the smart power-up mode.

There are simpler non-stand solutions that simply couple 48V down the line, but they are not part of the PoE standard.

The chip solutions generally don't need a large PCB footprint, and a 48V to

9V DC-DC converter can be made really small as well.

Dave.

That won't work. You need one of these PD chipsets:

Plus the 48V to 9V DC-DC converter. This looks like the smallest PD chip solution from Maxim: Cheap and easy to implement.

Dave.

Thanks Dave,

thats the lowest parts count solution I have seen so far. The application examples in the datasheet with the wall wart adapter as alternative power source are exactly what I want to do.

Ian

IIRC the standard is 42V(ish) at upto 400mA between two pairs.

If you don't care about standards and are prepared to just use unused pairs in a 100Mb/s CAT5 run the you can do anything you want.

where they are quoted the connectors seem to be specced for conducting upto 2A current.

if you want to go the standards way you need centre-tapped ethernet transformers that can handle the common-mode current. and apropriate regulation and signalling circuits to make it all work.

Ever heard of google? It would save your foot from teeth marks.

48V @ 350mA.

You better be concerned. The power isn't applied unless you enable it using some kind of protocol. You'll need switching controller with POE negotiation. Freescale has some pretty simple, reasonable priced controllers for this purpose.

--
Failure does not prove something is impossible, failure simply
indicates you are not using the right tools...
                     "If it doesn\'t fit, use a bigger hammer!"
--------------------------------------------------------------

Maybe you should try reading the entire thread next time, dipshit.

I'm sorry that I didn't read all the episodes of your answer. Maybe you should try to squeeze it all in one -not so boring- episode :-)

At least I point to a working solution for the OP's problem.

--
Failure does not prove something is impossible, failure simply
indicates you are not using the right tools...
                     "If it doesn\'t fit, use a bigger hammer!"
--------------------------------------------------------------