RF attenuator power rating

[TheUnnamedNewbie]

I have wondered two things about attenuator power ratings. First, is the rating the maximum input power, or maximum dissipated power? For example, if I have a 3 dB attenuator rated at 3 W, am I allowed to insert 6 W of power (since the 3 dB attenuator will absorb half of the power, and half of 6 W is 3 W) or only 3 W?

And something I never understood: Why do all attenuators come rated in W and not dBm? Ofcourse they are equivalent but since any RF engineer I know 'thinks' in dBm, wouldn't it make more sense to have everything in dBm?

[Tony_G]

Answering the second part first, attenuators are rated in dB not dBm because they act on the signal they are given and not against a reference level. dBm is a ratio of power referenced to 1mW, so 0dBm is 1mW and 3dBm is ~2mW.

A 3dB attenuator will give you ~half the signal.

So for the first part, attenuators are rated for the input power - a 5W 3dB attenuator would take a 5W input and cut it in half (approx) and disapate 2.5W

TonyG

[chrisl]

I have wondered two things about attenuator power ratings. First, is the rating the maximum input power, or maximum dissipated power? For example, if I have a 3 dB attenuator rated at 3 W, am I allowed to insert 6 W of power (since the 3 dB attenuator will absorb half of the power, and half of 6 W is 3 W) or only 3 W?

And something I never understood: Why do all attenuators come rated in W and not dBm? Ofcourse they are equivalent but since any RF engineer I know 'thinks' in dBm, wouldn't it make more sense to have everything in dBm?

1) 3W.
2) I think in W, dBm and dBW not just in dBm.  I would hate thinking in dBm when I work on 100kW transmitters..

[TheUnnamedNewbie]

Answering the second part first, attenuators are rated in dB not dBm because they act on the signal they are given and not against a reference level. dBm is a ratio of power referenced to 1mW, so 0dBm is 1mW and 3dBm is ~2mW.

A 3dB attenuator will give you ~half the signal.

So for the first part, attenuators are rated for the input power - a 5W 3dB attenuator would take a 5W input and cut it in half (approx) and disapate 2.5W

TonyG

I think you missunderstood my second part. I was not referring to the attenuation in dBm, that would obviously not make sense. Instead, I was referring to the power rating, which is in W and not dBm.

2) I think in W, dBm and dBW not just in dBm.  I would hate thinking in dBm when I work on 100kW transmitters..

Good point, I didn't think of that. The main reason I was wondering is that most atteunators I own have the power rating on them only in W, and as such I have printed out a table converting between W and dBm since I'm used to "thinking" in dBm.

[Tony_G]

Apologies if I mixed this up.

Going back to the original post:

First, is the rating the maximum input power, or maximum dissipated power?

They are rated in terms of input power.

Why do all attenuators come rated in W and not dBm? Of course they are equivalent but since any RF engineer I know 'thinks' in dBm, wouldn't it make more sense to have everything in dBm?

These items are not equivalent - An attenuator may disapate X watts but it can't disapate X dBm (ignoring active attenuators etc) as they're not referenced to each other. This is why attenuators aren't described in dBm.

Again I'm sorry if I'm not being clear on this.

TonyG

[T3sl4co1l]

These items are not equivalent - An attenuator may disapate X watts but it can't disapate X dBm

0 dBm == 0.001 W.

Tim

[nike75]

But if attenuators are rated at own input power, why 10W 3dB and 20dB attenuators have the same type and the same heatsink is installed as one will dissipate almost 10W and the other only 5W? May be for standard manufacturing...

[Berni]

Its easier to give a 3dB attenuator an overised heatsink from a 10dB, 12dB, 16dB, 20dB, 30dB attenuator than make a special half sized heatsink just for the 3dB one.

Also these wattage ratings are not that precise. That 5W attenuator won't blow like a fuse at 6W. It most likely works just fine at 10W, maybe even 15W and more. Its just a general guide to tell you not to use it on the output of a big ass 100W transmitter.

It's more covenant to have the max input power written on it since it gives you a quick at a glance ballpark figure of how much you can shove in there. If it was dissipated watts you would have to actually calculate it. Also writing "6dB 30dBm max" on the side of an attenuator is more prone to confusion than writing "6dB 1W max". Yes i know you are a educated engineer that knows the difference between dB and dBm, but we are all still only human and could quickly make the mistake of taking this one out of the box when you are actually looking for a 30dB attenuator.

[Joel_Dunsmore]

At least when I worked on power ratings (for bridges and power splitters) the rating was the withstand power at 24 hours at 55 deg C that would not cause a change in value such that the unit was out of spec.   So, a 1W (+30 dBm) attenuator would be able to be used at a 1W input power (power applied from a 50 ohm source, not from a voltage source of zero ohms, for example) for 24 hours at 55C, and still be in spec when returned to ambient and no power applied.  And, yes, we derate 6 dB for the guys that drive it from a voltage source.  So we normally would need to see >+36 dBm survival for a +30 dBm (1W) part.  But that's just HP.., uh.. Agilent, .. uh Keysight.

Oh, and shout out to Amplifier Research, never had one fail when doing these tests.

[radiolistener]

I have wondered two things about attenuator power ratings. First, is the rating the maximum input power, or maximum dissipated power? For example, if I have a 3 dB attenuator rated at 3 W, am I allowed to insert 6 W of power (since the 3 dB attenuator will absorb half of the power, and half of 6 W is 3 W) or only 3 W?

It is rated for input power, so if it has rating 3 W, then input power should not exceed 3 W (34.7 dBm for 50 Ω).

But for Chinese attenuators I would suggest to divide it's rating at least by 2 or even by 3.

I have several Chinese burned out attenuators which was burned at power which is a little bellow it's rating. For example 5 W attenuator was burned out from 4 W.

For some unknown reason Chinese attenuators cannot keep even rated power. But I have Chinese 10W dummy load which survive at 20 W power for several seconds a lot of times. I don't know why there is such difference between dummy load and attenuators, but Chinese attenuators are move vulnerable for input power than a dummy load.

Also be careful, attenuators are not always bidirectional, usually it has max power connector and low power connector. Do not mix them up, otherwise the attenuator will burn out. I have such experience, because Chinese seller didn't mark where is max power side (he even don't know that  ). For example I burned out 35W attenuator with 5 W power in such way.

I found that max power side usually marked with SU label on Chinese attenuators. Usually this is a male connector.

And something I never understood: Why do all attenuators come rated in W and not dBm? Ofcourse they are equivalent but since any RF engineer I know 'thinks' in dBm, wouldn't it make more sense to have everything in dBm?

Watts is more useful, because it allows to see power more precise and allows to estimate heat dissipation. The difference between 2 W and 100 W is just 17 dB, so it's better to see 100 W. With dB values it will be hard to estimate it's heating.

[ejeffrey]

IME, above about 1W / 30 dBm, power levels are often referred to in watts.  That is because that is when you generally start worry about the power supply, heat sinking, and thermal requirements where the log scale is not helpful.  So you might see an LNA with a output power of 17 dBm, but a 30 W power amp.  Of course if you look at the RF performance on the datasheet, it will have electronics performance attributes like P1dB and IP3 as dBm, but 30 W gives you a good idea how big a power supply you need and how big a heat sink you need.  In the case of an attenuator, the top line power rating is also an average power -- pulsed systems can exceed that by quite a bit for a short period of time if the duty cycle is low.

A side benefit of this convention is that you are familiar with it you are unlikely to get the power and attenuation ratings switched.  A 1 W attenuator that is labeled "30 dBm" might get misread as a 30 dB.

[Bud]

At least when I worked on power ratings (for bridges and power splitters) the rating was the withstand power at 24 hours at 55 deg C that would not cause a change in value such that the unit was out of spec.

Great educational piece of information, many thanks