3 dB bandwidth

Hi,

A voltage gain of -3dB is a voltage gain of 1/sqrt(2). Since power is relative to the square of the voltage, this implies a power transfer ratio of 1/2. So basically the -3dB point is the frequency where the output power of the circuit is half of the power in the passband.

greetings, Tom

Why half power? Why not 1/4th power? why do we choose half power?

thanks

Tom (at tomsweb.net) wrote:

For a single pole low-pass filter, that drop (3.01 dB, actually) occurs at the frequency where the straight-line projection of the passband intersects the asymptote of the stopband. For a series R/shunt C filter, it also where the magnitude of drop across the R equals the magnitude on the C. So it is readily calculated by hand, (as things were when that standard arose).

--
--Larry Brasfield
email: donotspam_larry_brasfield@hotmail.com
Above views may belong only to me.

3dB ~ half-power

--
  Keith

Why not?

That would be 6bB. Choose it if you so desire.

Two is such a nice number. Besides, two is the loneliest number since the number one and the 0db "point" doesn't tell us much.

--
  Keith

And besides, if you're going to pick a single frequency as the boundary between the passband and the stopband, it's natural to pick the point at which half the power gets through.

Cheers,

Phil Hobbs

And the issue isn't the measuring point, but that the point be consistent. One can come up with really good filter specs, but it means nothing if it turns out they were measuring at the 1.5dB point rather than the 3dB point. Use an odd reference point, and at the very least you'd better be sure to specify it. And even then, it can make for complicated comparisons with other such units, and might even be considered outright misleading.

Michael

Hello Keith,

Actually that is done for many filters when they characterize the roll-off at the 6dB and the 60dB points.

13 is more lonely...

Regards, Joerg

Sure, if those parameters are of particular interest. That was sorta my point.

Not according to Nilsson.

--
  Keith

Hello Larry,

Hey, they still are. I have one calculator in my drawer (HP11C) but several slide rule calculators. No abacus though.

Thing is, if the power goes out I can keep going. A long time ago they pestered me about it at a client's lab and one of the engineers bragged about a new calculator with "continuous memory" that didn't fail when the battery died. I told him my slide rule had continuous memory as well. The slider...

Then one day a company came out with a slide rule calculator that had an LCD and buttons. That almost made me sick.

Regards, Joerg

I have a slip-stic in my office and and HP45 around somewhere. The batteries in the '45 have long gone to their maker though.

Putting the slip-stic in its scabbard was tended to clear the memory.

I see it as someone with a sense of humor. I hope the PHB that funded the development didn't expect to make much money on it though.

--
  Keith

Addiators work better for addition and subtraction, and are still

100% mechanical. There are usually a few on eBay cheap because many were made and few people collect them. (Always contact the seller and confirm that the slides move freely and that the numbers are still readable).

I remember that one. The calc only added and subtracted. There was a sliderule with an addiator on the back too.

Back to the topic, 3DB is also the amount of drop you get when you have the output impedence and the input impedence matched, which was common in the days of transformer-balanced audio lines.

3dB down is half power. It's kinda 'traditional' to measure this.

In audio ( for one ) it's no longer the standard it was. Not least since if you put say 10 devices in series, the signal will then be 30dB down at the frequency in question !

With the higher performance available from modern circuit design and components bandwidth in pro-audio is often now specced at -1dB or even

-0.5dB.

Graham

Nicely put.

Reactive and real component of the filter are equal. It's as good an explanation as I need. ;-)

Graham

Nooooo ! That's a 6dB drop.

Thank God we got rid of so-called 'matched impedance working' and went for 'voltage matching' in audio a long long time ago.

I can recall seeing 600 ohm variable attenuators though in an antique studio when I was a young 'un. I expect they'd cost around £1000 to make today. Just to control the 'volume'. Wow !

Graham

Lissajoooze figures on a scope

martin

I have carefully 'lined up' audio tape recorders to an accuracy of better than

0.5dB 'by ear'.

3 dB is a dumbfuck 'standard'.

Graham

Even though I'm no expert, it seems feasible that 1 db would be "just detectable"; 3 db is, after all, _twice_ as much power!

I've also heard or read somewhere that it takes a full 10 db change to sound "twice as loud".

Cheers! Rich

Depends what you mean by "modern". -0.5dB performance was trivially and cheaply achievable 25 years ago.

Kevin Aylward snipped-for-privacy@anasoft.co.uk

SuperSpice, a very affordable Mixed-Mode Windows Simulator with Schematic Capture, Waveform Display, FFT's and Filter Design.