Thu Dec 24, 2009 1:32 am
Hrm, I think that's a bit ambitious as far as i'm concerned (re: 150w). I've had a number of problems that are a direct result of POE limited power supplies. I've gone trough a number of POE supplies for a 433AH, using a brand new 2a 24v regulated active POE injector I kept getting power outages and actually the POE burnt out within 12hrs of use, with a 1m cable run. Though it may just be a bad POE injector. I ended up having to use an old laptop power supply at 18.4v at 4a and a passive POE inserter (30cm POE cable run!), with that I've had no problems. In this setting i've ended up trying 24v 1.4a power supplies and 48v 0.8a with passive POE's and they've not worked either. I.e, in my setting a regulated active POE at 50w was not sufficient to power a single RB433AH with 2 relatively high powered radios (1x R52h and 1x kingston 1000mw radio).
150w at 24v is only 6a! What power supplies are you using at the minute? If you're using single power supplies at 24v and 1a and that's working for you, I suggest you test combining them with the calculated maximum rated output of your individual power supplies, so 10x individual power supplies at 1a = 10a power supply for 10 RB's, but you'll likely to experience more voltage drop with this type of setup and that has a massive implication on the amps you'll need.
Additionally assuming you're not putting the RB's in an identical position the cable lengths will vary, if this occurs with POE you are likely to see a varying voltage drop based on varying cable length, this will increase the current draw in amps for devices further away. Example;
30m cat5 cable for RB1 sees a voltage drop of about 2v. A 24vdc POE will be 22v at RB1 and if RB2 is at 60m, 20v. If the load (w) is the same for both, assume 15w with inefficiency, RB1 will draw (15w/22v)=0.68a, RB2 will draw (15w/20v)=0.75a. You can test your voltage drop using a volt meter, voltage drop will vary depending on quality of cable used. Also note that on the 433's system health shows the system voltage. When I put up my RB433AH towers I run mains cable (2.5mm 2 core) with the DC power going up this thick cable, voltage drop is significantly reduced, thus you can use a smaller power supply, this work great in combination with a passive POE inserter and separate DC power supply.
POE seems to be more unreliable with RB's and you may do better to power via the jack, though I don't have enough experience to talk personally, only passing on advise from others . . .
The additional questions that need to be considered are;
1. What distance does the DC current need to travel from AC power supply to RB's?
2. Are you running DC cable to the RB's or POE?
3. Is there significant distance between the RBs themselves? (i.e. RB 1 is 5m from the power supply and RB 10 is 100m from the power supply?
4. What are you finding your current power supply needs per RB?
5. What is the consumption of 'the worst' RB? (i.e. with the longest cable run / most active time / clients connected)?
5. What is your current voltage drop for each RB? (Use a volt meter to test the POE at source then at the RB end!)
Theoretically this should be a case of simple maths, and the big assumption being that the big power supply is giving what it says it will, my experience is that it will not. I.e. I'm sure my regulated 24vdc 2a active POE wasn't giving me 48w to the RB.
Hope that helps.
Josh