3dB 20W and 20dB 20W attenuators in series. What is the max power?

[LM21]

I hope the subject says it, but I try to cheat and get 40W (23dB) attenuator,  or at least higher power. For example,  if I have a 3dB 20W attenuator, does it mean that it can dissipate 20W and pass another 20W. Is it even  possible to say this with various manufacturers.

[jonpaul]

each attn is rated 20W MAX thru.

You WILL NOT get higher power thru with series.

Got to mfg specs.

We use Minicircuits, see spec.

Jon

[pdenisowski]

There's a reason why I don't try to make one of the big ones by connecting together a lot of the small ones

[LM21]

"You WILL NOT get higher power thru with series. "
I  checked  Minicircuits data. Minicircuit like you said,  is a reliable supplier. But they did not answer my question. i don't doubt your word but I would like to read more about it. Or is it just ohms law in attenuators.

[ahbushnell]

To minimize stress on attenuators you use the lowest first.
Assume all matched terminations

3dB  => .5 in power
So half the power will go ot the 3db and the other half down the line.  So 3db gests 10 watts
10 watts goes into the second 20 db attenuator. 
20db => .01 gain
10 watts goes to 0.1 watts.
The 10 db attenuator gets 9.9 watts.

If you start with the 20 watts it will get a large chunk of the power.  (20-.2)=19.8 watts.
Very little into the second 3 dB. 

Andy

[radiolistener]

I hope the subject says it, but I try to cheat and get 40W (23dB) attenuator,  or at least higher power. For example,  if I have a 3dB 20W attenuator, does it mean that it can dissipate 20W and pass another 20W. Is it even  possible to say this with various manufacturers.

No. The input power is limited with your first attenuator rating. If your first attenuator is rated for 20W, then you can put max 20W on the input.

There is no way to cheat with low power attenuators. The only way to put 40W is to put first attenuator rated for at least 40W.

You can use more powerful attenuators. For example, you can use 100W attenuator for 40W signal.That is okay.

[rfdave#gmail.com]

think about it this way. a 20W 3dB attenuator has a maximum input of 20W. If you put 40W into that, then you'll damage it, that's too much power.
However, at the output of that 3 dB attenuator, there's now 10 W. If you have a 10W attenuator, then you can use it safely there, and stack up attenuation that way.

[fourfathom]

But remember when stacking attenuators that a fair amount of heat can be transferred by the connectors.  The first attenuator will be heating the second beyond ambient, etc.  So you may want to de-rate the attenuators a bit.  But really, you always want to de-rate an attenuator -- they get hot!

[jonpaul]

get passive R pad calcul, T pad or Pi pad.

https://www.pasternack.com/t-calculator-pi-attn.aspx
https://www.rfcafe.com/references/calculators/attenuator-calculator.htm


Setup Zin=Zout = 50

ATN = 3db, 20 db.

Calc resistors

Use Ohms law, draw/sim ckt with Vin = eqv 20W on 50 Ohms.

Current (I) = 0.632 e (A)

Voltage (V) = 31.62  (V)
OR calc https://www.calculator.net/ohms-law-calculator.html?v=&vunit=volt&c=&cunit=ampere&r=50&runit=ohm&p=20&punit=watt&x=Calculate

Get every internal R rating ..

Should be 10 min on an envelope by hand faster if you sim

j

[LM21]

So max input power of 20W attenuator is 20W. 10W becomes heat and the other 10W comes out of it. I should do the calculations, for example with Minicircuits PI configuration. It is not hard.

Other problem is that are those attenuators symmetrical, what is high power input, and what is output.

[vk6zgo]

I hope the subject says it, but I try to cheat and get 40W (23dB) attenuator,  or at least higher power. For example,  if I have a 3dB 20W attenuator, does it mean that it can dissipate 20W and pass another 20W. Is it even  possible to say this with various manufacturers.

Ignoring the second attenuator for a moment & just using the 3dB one, if you are starting with 40 Watts, & you put a 3dB attenuator in line, a matched load will see 20W, & 20W will be dissipated in the attenuator, which is its maximum rating.

If you now replace the load with a 20dB attenuator of the same power rating & place the load at the output of that second attenuator, this will yield 200mW into a matched load.

The 20dB attenuator is dissipating 19.8 W which is very close to its maximum rating.

In this case, both attenuators are very close to being "cooked". (don't ask how I know!! )

[ejeffrey]

think about it this way. a 20W 3dB attenuator has a maximum input of 20W. If you put 40W into that, then you'll damage it, that's too much power.

That is definitely correct with respect to what the data sheet says.  However the construction of the 20W 3 dB attenuator (10 watts dissipation) sure looks a lot like the 20 W / 20 dB attenuator.  (19.8 watt dissipation). It's certainly possible that the internals are different enough that the ratings are correct, but its also possible (even likely?) that the entire range including the 3 dB attenuator are designed for 20 watts thermal.  I think this is likely in order to account for increased dissipation when the load is mismatched or open circuit.

If all that is true, technically you could run 40 watts input power to the 3 dB attenuator as long as you always have a matched load such as another attenuator on the output.  But this is definitely operating outside the specifications.

[PA0PBZ]

Ignoring the second attenuator for a moment & just using the 3dB one, if you are starting with 40 Watts, & you put a 3dB attenuator in line, a matched load will see 20W, & 20W will be dissipated in the attenuator, which is its maximum rating.

If you now replace the load with a 20dB attenuator of the same power rating & place the load at the output of that second attenuator, this will yield 200mW into a matched load.

The 20dB attenuator is dissipating 19.8 W which is very close to its maximum rating.

In this case, both attenuators are very close to being "cooked". (don't ask how I know!! )

I disagree, the power rating of the attenuator is the maximum input power, not the dissipated power. Maybe that's why you cooked yours?

[fourfathom]

I disagree, the power rating of the attenuator is the maximum input power, not the dissipated power. Maybe that's why you cooked yours?

Yes!  Don't think of the attenuator rating as if it's a resistor. 
And depending on who is selling them, those small aliexpress attenuators have ratings all over the place for the exact same part.  Plan on derating those by half, and still expect them to be uncomfortably hot at that power level.

[PA0PBZ]

Plan on derating those by half, and still expect them to be uncomfortably hot at that power level.

Also the rating decreases with increasing temperature, did someone say runaway?

[Hamelec]

I would like to drill an 8mm hole - is it possible to use a 5mm and a 3mm drill bit for this? 

[radiolistener]

Ignoring the second attenuator for a moment & just using the 3dB one, if you are starting with 40 Watts, & you put a 3dB attenuator in line, a matched load will see 20W, & 20W will be dissipated in the attenuator, which is its maximum rating.

As I know, attenuators are rated with max input power, not dissipation power. So if you push 40W to a 20W/3 dB attenuator, it can be burned out despite the fact that dissipation power is just 20W.

It is very easy to burn out attenuator, I have several burned attenuators. So, be careful.

[CaptDon]

Read the specs for the exact attenuator you have. It may be 'Max Input Power' rated or it may be 'Max Dissipation' rated. You need to know!!!! I have some of each type. Kay Electronics, EMCO, HP etc.

[LM21]

And depending on who is selling them, those small aliexpress attenuators have ratings all over the place for the exact same part.  Plan on derating those by half, and still expect them to be uncomfortably hot at that power level.

Easy to beliewe, but they are so cheap.

[radiolistener]

but they are so cheap.

As the saying goes, "Cheap fish - crappy yushka"

[vk6zgo]

Ignoring the second attenuator for a moment & just using the 3dB one, if you are starting with 40 Watts, & you put a 3dB attenuator in line, a matched load will see 20W, & 20W will be dissipated in the attenuator, which is its maximum rating.

If you now replace the load with a 20dB attenuator of the same power rating & place the load at the output of that second attenuator, this will yield 200mW into a matched load.

The 20dB attenuator is dissipating 19.8 W which is very close to its maximum rating.

In this case, both attenuators are very close to being "cooked". (don't ask how I know!! )

I disagree, the power rating of the attenuator is the maximum input power, not the dissipated power. Maybe that's why you cooked yours?

No, it was because of putting the 20w attenuator first & the 40W attenuator second---what happens when your load has the wrong sex connector on it, & in a moment of "brain fog", I reversed them instead of using a second "N" adaptor.
You are right, of course, as the attenuator is a network, so the first shunt/series resistor combination is hanging directly across the higher power input.

I normally used much higher rated attenuators than the power I was dealing with.

[Marsupilami]

Everything that needed to be said has been said, but it's better with pics.
If you think about max input power in dBm that makes chain derating much much easier.

[CaptDon]

I can say for a fact that with the larger 50 watt and up attenuators they are dissipation rated!!! A 500 watt 3dB unit can and will max out at 1KW input with a properly matched load on the output. That very same physical size attenuator when purchased in the 20dB or 40dB model can and will max out at around 480 watts input and the 40dB model can be used as a dummy load with very little V.S.W.R. reflected to the input side with nothing connected to the output side although I keep a 10 watt 50 ohm 'N' connector dummy load to use as a terminator in those cases. Handy rule of thumb is if you are approaching 80% of the rated dissipation of your attenuator based on dissipation and NOT input power you probably need the next higher wattage rated model. I have never ruined an attenuator and we ran up to 50KW water cooled units and dummy loads with water cooling, air cooling or vapor phase cooling up to 4 megawatts. Cooling systems were based on customer specs, duty cycle and available finances vs. 'dependability'. As always, go cheap and it will surely burn up!!!

[vk6zgo]

I can say for a fact that with the larger 50 watt and up attenuators they are dissipation rated!!! A 500 watt 3dB unit can and will max out at 1KW input with a properly matched load on the output. That very same physical size attenuator when purchased in the 20dB or 40dB model can and will max out at around 480 watts input and the 40dB model can be used as a dummy load with very little V.S.W.R. reflected to the input side with nothing connected to the output side although I keep a 10 watt 50 ohm 'N' connector dummy load to use as a terminator in those cases. Handy rule of thumb is if you are approaching 80% of the rated dissipation of your attenuator based on dissipation and NOT input power you probably need the next higher wattage rated model. I have never ruined an attenuator and we ran up to 50KW water cooled units and dummy loads with water cooling, air cooling or vapor phase cooling up to 4 megawatts. Cooling systems were based on customer specs, duty cycle and available finances vs. 'dependability'. As always, go cheap and it will surely burn up!!!

That is the sort of thing I was used to, & used the dissipation rating as that was how I remember using them, but it does seem intuitive that smaller ones are rated by the input power they can handle due to the fact that they are networks, so, assumed I remembered incorrectly.

[fourfathom]

I suppose it depends on the manufacturer.  Here is a Mini-Circuits 100W 20dB attenuator, rated at 100W input on the male port, and 20W input power on the female port:
https://www.minicircuits.com/pdfs/BW-20N100W+.pdf

Lesson: check the specs!

[CaptDon]

Excellent that they gave both specs!! MiniCircuits is a great company although they have 'military' pricing!! For my own experimentation I can only afford their products at used Ebay prices 'with return guarantee' since a lot of that stuff on Ebay comes 'Pre-Destroyed' saving you the effort of blowing it up yourself!!

[mag_therm]

I think LM21's question is not trivial and can't properly be answered by the anecdotes and vendor ratings etc.
I would agree with jonpaul reply 8:
The Tee and Pi impedances per network are cast in stone and element values can be measured (o/c , s/c, thru)  or calculated, for example
https://www.everythingrf.com/rf-calculators/attenuator-calculator

Nest step is to connect networks in series as LM21 requests, and solve using Kirchoff.

That will give the current, voltage and power in each element of each network.

Above is how old guy like jonpaul,  myself would have done it on slide rule or calculator  ( and now easier in sim) ,
 but young guy would probably use S-parameters and get the same results.

[fourfathom]

I think LM21's question is not trivial and can't properly be answered by the anecdotes and vendor ratings etc.
I would agree with jonpaul reply 8:
The Tee and Pi impedances per network are cast in stone and element values can be measured (o/c , s/c, thru)  or calculated, for example
https://www.everythingrf.com/rf-calculators/attenuator-calculator

Nest step is to connect networks in series as LM21 requests, and solve using Kirchoff.

That will give the current, voltage and power in each element of each network.

Above is how old guy like jonpaul,  myself would have done it on slide rule or calculator  ( and now easier in sim) ,
 but young guy would probably use S-parameters and get the same results.

Fine, but those resistance measurements tell us nothing about power dissipation ability, or if the attenuator is symmetric.  This is why you want vendor ratings.
An yeah, I still have my slide rules, but what's wrong with using the attenuation values to figure out the power input at each of the series-connected attenuators?  Sounds simple and accurate to me.

[mag_therm]

Hi fourfathom,
" This is why you want vendor ratings."
https://app.box.com/s/sxy8ctepb6ay4ijfk3c69gcbdcsmiu6v
( just for fun)
regards

[CaptDon]

I swear, that crazy lash up looks like something my college grad engineers would do!!! Then after the destruction they would put all the attenuators back in the drawers and do a Sgt. Shultz "I know nothing.....nothing"!!! There are times when I wonder why smoke isn't rolling out of the storage bins. It gives you a sad feeling when you walk into the test equipment storage room and smell burned electronics!!! Hence, the now implemented signout sheets!!