Coupler/attenuator - how many watts?

Hi all,

I have been looking for the answer to my question but no success so far. Hopefully you can help me out.

Let's assume we have 3 mobile operators each having GSM900, GSM1800, UMTS2100 and LTE2600, 35dBm on each of these output. These signals go to the input of 15.1/0.1 coupler (let's assume there's no combiner between the basestations and coupler). If we now want to use attenuator on 15.1 output, how many watts minimum does this attenuator have to be?
35dBm - 15.1dBm = 19.1dBm = 0.098W. We have 12 signals (GSM900, GSM1800, UMTS2100 and LTE2600, 3 operators) coming from basestations, so we have 12*0.098W = 1.18W on coupler output, so it means the attenuator has to be >=1.18W. Is this correct?


Phil

Actually GSM mobile max power is +33dBm, but basestation power is greater than +35dBm, I guess you mixed up here.
Say basestation power is +50dBm in peak, and GSM has some pilot channel, other system are same, so the attenuator max input is +50dBm+10log(12*2)-15.1+3dB(redundancy)=51.7dBm=21.7dBw. This is huge.
In reality, when multi-system, some TX may be back off from max power, but it is really hard to evaluate it.
The cost way is using a standalone attenuator and coupler for each channel, then is +50dBm+10log2-15.1+3db=40.9dBm.

JuniorRF said:

Hi all,
35dBm - 15.1dBm = 19.1dBm = 0.098W. We have 12 signals (GSM900, GSM1800, UMTS2100 and LTE2600, 3 operators) coming from basestations, so we have 12*0.098W = 1.18W on coupler output, so it means the attenuator has to be >=1.18W. Is this correct?
Phil

Click to expand...

Yes the calculation of the RMS value is correct. Be careful with peak to average value that for UMTS and LTE is quite high (about 10 dB). This means that for UMTS and LTE you have to conider a peak value of roughly 45 dBm. So the peak power will be 3*(3.2W+3.2W+32W+32W)=211W that at the output of the 15.1 dB coupler will be 6.5 W peak.
So the attenuator must be able to handle an RMS power of 1.2W with peaks up to 6.5W

tony_lth said:

Actually GSM mobile max power is +33dBm, but basestation power is greater than +35dBm, I guess you mixed up here.
Say basestation power is +50dBm in peak, and GSM has some pilot channel, other system are same, so the attenuator max input is +50dBm+10log(12*2)-15.1+3dB(redundancy)=51.7dBm=21.7dBw. This is huge.
In reality, when multi-system, some TX may be back off from max power, but it is really hard to evaluate it.
The cost way is using a standalone attenuator and coupler for each channel, then is +50dBm+10log2-15.1+3db=40.9dBm.

Click to expand...

Thank you tony. I forgot to add that this is an indoor solution, so probably that's why the basestation output signal value was only 35dBm.

albbg said:

Yes the calculation of the RMS value is correct. Be careful with peak to average value that for UMTS and LTE is quite high (about 10 dB). This means that for UMTS and LTE you have to conider a peak value of roughly 45 dBm. So the peak power will be 3*(3.2W+3.2W+32W+32W)=211W that at the output of the 15.1 dB coupler will be 6.5 W peak.
So the attenuator must be able to handle an RMS power of 1.2W with peaks up to 6.5W

Click to expand...

Thank you albbg. This was an exhaustive answer. It answered my question completely.
But what if a attenuator with a value of smaller than 1.2W is used (besides it won't attenuate the signal anymore)? Will it get hot and damage the cable it is connected to, for instance?

you can use two coupler seriesing, so that you can get 15.1dB+15.1dB=30.2dB attenuator, then you can use that for attenuator. but you need a load for the 2nd coupler.