ADAR1000 - Receive section GAIN in RevB datasheet

Single Channel and Coherent Gains are correct.  Electronic Gain as stated is not correct. 

You can define Electronic gain as the output power/sum of the input powers.   So it is basically measured the same way as coherent gain, but instead of being referenced to a single inputs' power, it is the sum of the input powers.   So if you have 0 dBm (1 mW) on each input, across four inputs, that is 4 mW sum.  If the output is at 19 dBm (79.5 mW), then the electronic gain is:

• 10*log10(79.5 / 4) = 13 dB

Coherent gain would be:

• 10*log10(79.5 / 1) = 19 dB

I am assuming the signals are the same amplitude when applied to the combiner and in-phase with each other. 

By the way, these input powers would push the receive well into its compression, so they are not realistic.  The above analysis assumes that the compression isn't happening.  

Can you explain this a little bit more in detail to understand the logic why the coherent gain is 12 dB higher and not 6 dB?

I want to simulate the scenario of combining 4 coherent receive signals with e.g. 0 dBm. Then each of the signal travels through the internal signal chain of the ADAR1000 with a gain of e.g. 7 dB (@14 GHz) and afterwards the 4 Signals will be added to each others with the combiner. Therefore, the power level will increase by 6 dB. But according to your statement i would measure in this scenario a gain of 19 dB.

Why do i measure 19 dB?

Can you give me a logical explanation where the additional 6 dB come from?

For a 4:1 power combiner like on the ADAR1000, coherent gain is 12 dB larger relative to a single input driven, with the other three inputs terminated with 50 Ohms.  We test the part on the bench quite often with a single input driven, thus the reason we specify this gain and then relate it to coherent gain, where four input are driven with equal magnitude and in phase. 

You are correct, if you have 4 signals and each are 0 dBm and in phase, and you feed them into a 4:1 ideal combiner, you get +6 dBm on the output. 

But, if you drive a single input of the combiner with a 0 dBm signal and terminate the other three inputs, you will get - 6 dBm on the output.  The power combiner is acting as an attenuator when only a single input is driven.

There is a 12 dB difference between the two scenarios.

For a 4:1 power combiner like on the ADAR1000, coherent gain is 12 dB larger relative to a single input driven, with the other three inputs terminated with 50 Ohms.  We test the part on the bench quite often with a single input driven, thus the reason we specify this gain and then relate it to coherent gain, where four input are driven with equal magnitude and in phase. 

You are correct, if you have 4 signals and each are 0 dBm and in phase, and you feed them into a 4:1 ideal combiner, you get +6 dBm on the output. 

But, if you drive a single input of the combiner with a 0 dBm signal and terminate the other three inputs, you will get - 6 dBm on the output.  The power combiner is acting as an attenuator when only a single input is driven.

There is a 12 dB difference between the two scenarios.