# Power Ratings of RF Attenuators In Series

If I connect RF attenuators in series, does the power rating of the attenuator closest to the antenna have to be high enough to handle the expected RF power on its own? Or does the placing attenuators in series limit the amount of power each attenuator must dissipate by spreading it out amongst attenuators?

For example, if my expected RF power is 2W, do I need to use a 2W attenuator nearest to the antenna even if there is going to be a 1W attenuator right behind it? Or can I get away with a 1W?

A 2W attenuator is not designed to dissipate 2W, it's designed to receive 2W.

The amount of power at its output will be determined by its attenuation. A 3dB attenuator would output 1W. A 10dB attenuator would output 200mW.

If you cascade attenuators, each attenuator only needs to be rated for the expected output power of the previous one.

Yes, the initial attenuator still must handle the power level present.

Attenuators are not just series resistors, they are dividers with shunt resistors as well, so they do not combine like resistors in series.

It is convenient that you can add the logs of the power ratios when combining attenuators - but the actual power converted to heat depends on the absolute power level input, not just the ratio of the input to output.

To the degree that each component is ideal, the power the first attenuator converts to heat is the same regardless of what follows it, as long as what follows matches the impedance in a purely real way. In a mismatch case it could actually be slightly higher.

• With that in mind, is there any difference in performance if you place the attenuator with the higher or lower attenuation rating closest to the antenna? Assuming all attenuators have the same power rating, I'm thinking the attenuator with the lowest attenuation should closest to the antenna since that is where the highest power is encountered so if it attenuates the least there, it uses less of its power budget compared to if another attenuator in the cascade was placed there which might improve performance (like linearity or something else)? – DKNguyen Apr 13 at 15:46
• And if there are shunts inside attenuators, that would mean that a DC block can only protect attenuators on the side of the DC block opposite of the antenna right? – DKNguyen Apr 13 at 15:54
• A nasty trap is that some attenuators are asymmetric WRT power handling. For example I have a 250W, 30dB unit where only one end can actually tolerate 250W, going the other way it is a 0.5W attenuator. This can be an expensive mistake as both ends have N types fitted! – Dan Mills Apr 13 at 19:29