what are the differences in using the following voltages?
12v – ?
18v – ?
24v – ?
help is highly appreciated…
12v would not recommend it because the voltage drop in long cables could cause you problems.
You can use 18v but stp cable that are not too long.
24V is the voltage works best for me.
do you suggest 12v for system trials or demo? where the unit is very near to the battery… say the distance of is not 1 to 3 meters max??
-sid
In that case, there is no difference which voltage to use.
will 12v 7AH battery do the work? and for how long will that work??
Hi …
Well, the last 2 RB433 I tested was unstable when battery drops to 12V. I mean, I had 411ARs that run ok from 13.6 (battery floating voltage) to almost 11V (drawn) then stops.
I replaced 411ARs by RB433 keeping the same wiring (7m UTP cable). The first time mains power went out the existing battery lasts 2 hours (a 45Ah one) then AP stopped. The 411ARs lasts 24 … 36h.
Note that the single difference was an extra R52nM card (1.6 watts rated). I only move the existing cards from 411AR to 433 (e.g. a XR9 and changed the built in by a R52nM).
At the bench, the 433 delayed to boot from a 12.0V 2A power supply (using power jack). No problem at all with 15V on POE from a UBNT power supply.
Now my APs are 433, 433AH & 433UAH based.
So I’m changing all existing 12.6V solution (battery + floating charger) which I use to power 411ARs, WRAPs & etc for two 12.6V batteries in series + floating charger (27.4V) + linear 5A voltage regulator (22V).
Regards;
Thz for that, mramos
If you have only 3m of cable, then there is no problem to use 12V. If you run your device from battery, then it is better to use 12V, because then Routerboard takes less power than if you use 24V. Dont know why, just measured this behaviour. If you need longer cable, then use CYKY cable (e.g. 1.5mm) and end it up with a jack.
how much less power?
Are you refering to current draw from the battery @12volts and also current draw from 24v supply and is there a radio card attached.
We used a laboratory power supply with adjustable voltage output. As device we had RB433 with one R52 card. I dont remember the values, but I guess it was something like as few as 3W at 12V and 6W at 24V.
The only problem with battery is that its voltage varies from 13.8V to as low as 9V. Mikrotik says that Routerboard supports 10V supply. Under 10V it wont work I think. Anyway you should not go under 10V at your battery, it’s not good for it’s health. Usually 10.5V is the low limit. There is not much more capacity under 10.5V anyway. Maybe this is good for you to have 12V instead of 24V because if you use batteries in series and get 24V without any protection, then the routerboard will suck your batteries down to 10V (if that is even possible), but totally drop their voltage, which can be final for them.
Plus if you want more capacity, just plug the batteries in paralel and get more Ah. You can use any amount of batteries 
like 3, which you cant do when using 24V.
And yes, I remembered one more reason why to use 12V - if you use 2batteries in series, and you recharge them using 24+V (or higher), then the current will be the same in both batteries. But batteries sometimes differ in their capacity. So one battery can be already at 13.8V, but the second one will be at 12.5V. And if you want to get 13.8*2V = 27.4V (theoretical full capacity), then you will keep charging and charging and you will be overcharging the one that first reached 13.8V. Some people made a “battery balancer” or whatever they call it, but its extra money.
I used 12V batteries with approx. 5m cables for measurements done in environment where was no electricity. One 12V 7.2Ah battery can power up RB433 with a R52 card approx. 10 hours without any problem (that was the time I spent in terrain, many times I used the battery twice without recharging, so I guess 15hours and more…)
Ok just to rephase your first comment about
because then Routerboard takes less power than if you use 24V. Dont know why, just measured this behaviour
is correct about using less power but also it should be noted less power for the radio card so instead of giving out say 25dbm is will be much less and lower signal level from your AP than running at 24volts.
this to me needs more testing… if the routerboard is taking twice as many watts just using a different voltage then something is wrong. Watts are watts …
The tx of the radio will not be affected. Thats just not true.
Can you retest and post exact values of observed draw at certain voltages?
Using Ohm’s law it’s simple for power if you drop the voltage current must increase to compensate if not the power levels drop,
http://www.angelfire.com/pa/baconbacon/page2.html remember watts are not watts reduce one the other must increase?
Another example is in a audio PA amplifier with say DC rail voltages of +/- 90volts @230v AC input reduce the 230v to 170v and rail voltage could drop to +/- 70volts and output watts will also be reduced.
is correct about using less power but also it should be noted less power for the radio card so instead of giving out say 25dbm is will be much less and lower signal level from your AP than running at 24volts.
That is a nonsence. The output signal level is driven by programmable setting.
Can you retest and post exact values of observed draw at certain voltages?
Ok, I’m going to test this for you once again
Just think for a minute why do most WISP’s use a 24V power supply @ 1.6Amps (or 800mA) to supply AP’s when if as you say “signal level is driven by programmable setting” and just use a 12volt at 800mA, the only way this going to answered is when someone puts a RF dummy load onto the radio card and check with a RF power meter it’s output dbm’s when the DC supply voltage and current is varied.
Just think for a minute why do most WISP’s use a 24V power supply @ 1.6Amps (or 800mA) to supply AP’s when if as you say “signal level is driven by programmable setting” and just use a 12volt at 800mA, the only way this going to answered is when someone puts a RF dummy load onto the radio card and check with a RF power meter it’s output dbm’s when the DC supply voltage and current is varied.
The reason for using 24V instead of 12V is that you can have long cable from the power supply to the device. Using 12V and long thin UTP cable wires will cause significant voltage drop and losses on the cable. When using 12V on distance longer than few meters it is clever to use cable like CYKY or CYSY 1.5mm. But of course these are more expensive and you have to end them with jack… which is much more work than simply use UTP cable.
As you write “24V@1.6A or 12V@0.8A” - this depends on the power supply. You can find 24V@0.5A and also 12V@5A - that is no problem. It just depends on the construction of the power supply.
If you think more voltage = more rf power, or more Amps at power supply = more rf power, just do simple tests to prove you wrong. It does not matter if you use this or that power supply. You can still have your wireless card in Default tx power mode and you will get the very same output signal. Try it.
And now the testing results. Tested 3.11.2011, using laboratory power supply. RB433 was powered via jack, it had one CM9 card using 5GHz band (By the way CM9 card is a really bad card - it does not regulate under 10dBm at all. If you set 2dbm output power, it still transmits 10dBm… what a piece of trash). Unfortunatelly there is no ethernet connected (it also takes some watts - something tells me it was +0.6W if any ethernet interface was connected), but still the results are nice.
By the way, I think this behaviour is caused by some stabilizator (stabilized voltage) on the board which transferes higher “unstable” voltage to stabilized 5V and 3.3V. The more voltage is on the input, the more loss has to occur on the stabilizer and therefore the device consumes more power.
Voltage Amps Watts
10.5 ------- 0.25 ------ 2.63
12.5 ------- 0.22 ----- 2.75
15 --------- 0.19 ----- 2.85
18 --------- 0.17 ----- 3.06
20 --------- 0.16 ------ 3.20
24 --------- 0.1516 ---- 3.64
27.5 ------- 0.1420 ---- 3.90
Here are some photos of the measurement. The power supply shows Output voltage on its top LCD and Current on its bottom LCD. Multimeter shows precize value and proves it.
To prove to myself if the output of the radio card varies or not I must setup a variable test bench power supply and vary the voltage and current to a test AP and monitor with a CPE the db signal levels which should not drop when the voltage is reduced and current limiting is used on the power supply.