dB

[dr.power]

Hello guys.

I indeed know what dB is and how it does work.
I of course know two dB formulas for Power and Voltage too.

But most of times I see somebody or a Paper just talk about dB, yet do not refer if they mean dB calculated by VOLTAGE or by POWER. For instance somebody just told me "is using a 40dB filter sufficient for one of my modulator circuits or not". or somewhere I saw that "the gain of an antenna or a booster is 110dB" and so on.
The question which comes out is how to realize If guys or the papers and etc refer to dB voltage or dB power when They talk about dB values??
Is this problem just for mine or you guys have trouble to realize such situations??

Thanks for any input

 

[Externet]

Goes with the context.
If someone talks about decibels on a power subject, dB is about power.
The same for voltage levels. As in a 100mV audio signal, attenuating it a number of decibels corresponds to a ratio in its voltage.

Expressing the power itself, not a ratio of amplification or attenuation, is written as dBm; or dBW, or whatever the reference is taken in Watts.
Expressing the voltage itself, not a ratio of amplification or attenuation, is written dBV; or dBu, or dBuV, or whatever the reference is taken in volts.

 

[Mr RB]

dB should be relegated to the trashcan, along with Furlongs, Fathoms and Fahrenheit.

And feet and inches should follow shortly afterward, when people get used to the idea that we are in a new millenium now...

 

[Externet]

Do not forget the fluid ounces, grains, acres, bushels, gallons of both kinds, BTUs...
I would add changing the time to decimal, as say 1000 units per earth rotation and forget about hours, seconds...sexagesimal latitude, longitude, degrees of angles, and other like the 12 month, 7 days weeks; etc. But too late for such changes.

 

[crutschow]

dB should be relegated to the trashcan, along with Furlongs, Fathoms and Fahrenheit.

And feet and inches should follow shortly afterward, when people get used to the idea that we are in a new millenium now...

And why do you consider dB obsolete? It is a convenient, logarithmic way to refer to attenuation or gain values. It thus allows simple addition of values to calculate the gain or attenuation of amplifiers or attenuators in series. Unlike the other terms you mention then is no alternate substitute for dB.

 

[carbonzit]

Yes, I'm quite puzzled hearing this from you, Roman: perhaps you're joking, and not using smileys? or are you serious (about getting rid of dBs)? (I might ask the same question of the poster above with the "abolish" sig.)

Regarding those other values you seem to think have no value (heh), I certainly don't want to start a long flame war (well, at least not here, but perhaps we could take it to a more appropriate forum?), but I happen to very much like our non-MKS units for everyday usage (foot, inch, pint, gallon, mile, etc.), and I can confidently tell you that we ('Merkins) are probably never going to move away from them. That was tried, and failed miserably, in the 1970s. Thanks, but no thanks.

For some work, like carpentry, a system like ours that divides into powers of 2 (half/quarter/eighth) is actually much easier to use than a decimal system.

So there!

 

[crutschow]

Strictly speaking dB always refers to power ratios. Thus you don't have to worry about whether it was originally calculated from voltage or power, they are equivalent.

So if you want the power ratio to be expressed in power dB you use 10 log(power ratio). If you want to express the power ratio in voltage, you use 20 log(voltage ratio) since the power is proportional to volts squared and multiplying the log of a number by 2 is the same as squaring the number. But both numbers are equivalent and signify a ratio of power.

 

[Mikebits]

Strictly speaking dB always refers to power ratios. Thus you don't have to worry about whether it was originally calculated from voltage or power, they are equivalent.

So if you want the power ratio to be expressed in power dB you use 10 log(power ratio). If you want to express the power ratio in voltage, you use 20 log(voltage ratio) since the power is proportional to volts squared and multiplying the log of a number by 2 is the same as squaring the number. But both numbers are equivalent and signify a ratio of power.

Nuf said...

 

[carbonzit]

Strictly speaking dB always refers to power ratios. Thus you don't have to worry about whether it was originally calculated from voltage or power, they are equivalent.

Now I'm confused.

I don't want to muck up this thread too much, but if one calculates the voltage gain, say, of an amplifier, in dB, how is that a power ratio?

And I thought dBs could be used to express any ratio that lends itself to logarithmicity (to coin an unwieldy word).

Anyhow, it's the power thing that's throwing me now.

 

[Mikebits]

P= V^2/R Thus power gain in Db = 10log(Pout/Pin) Voltage gain in Db = 10log(Vout/Vin)^2 or 20log(Vout/Vin)
I think this only works when impedance are equal.

 

[Externet]

Hello.

"....are you serious (about getting rid of dBs)? (I might ask the same question of the poster above with the "abolish" sig.)

Now I'm confused.

I don't want to muck up this thread too much, but if one calculates the voltage gain, say, of an amplifier, in dB, how is that a power ratio?

And I thought dBs could be used to express any ratio that lends itself to logarithmicity (to coin an unwieldy word).

Anyhow, it's the power thing that's throwing me now.

-You just answered yourself.-

 

[crutschow]

Now I'm confused.

I don't want to muck up this thread too much, but if one calculates the voltage gain, say, of an amplifier, in dB, how is that a power ratio?

And I thought dBs could be used to express any ratio that lends itself to logarithmicity (to coin an unwieldy word).

Anyhow, it's the power thing that's throwing me now.

The dB refers to power ratios but common practice is to use it also for voltage gain. It equals the power gain if the input and output impedances are equal, such as for an RF amp. dBs are also used for other measurements such as acoustic intensity, but again it refers to the power level of the sound.

Wikipedia tells all.

 

[carbonzit]

Moderator: please delete this post.

Made in error, don't know how (keyboard switch bounce?)

Thank you.

 

[carbonzit]

OK, let me see if I've got this right (and forget about Wikipedia, the "encyclopedia" that any pimple-faced junior-high-school student or crank can edit and make like they're an "expert"):

In the case of a RF amplifier, input and output Z are assumed to be equal, so dB here is a power ratio.

For a small-signal AF amplifier, with inputs and outputs probably not equal (high input Z, low output), dB is not a power ratio. (Although it would be possible to express the amp's gain as a power ratio if input and output impedances were normalized, correct?)

I guess it's as another poster above put it: it all depends on the context.

 

[carbonzit]

"....are you serious (about getting rid of dBs)? (I might ask the same question of the poster above with the "abolish" sig.)

-You just answered yourself.

How did I just answer myself?

Could you please elaborate your argument against this measurement unit? I'm curious.

 

[crutschow]

OK, let me see if I've got this right (and forget about Wikipedia, the "encyclopedia" that any pimple-faced junior-high-school student or crank can edit and make like they're an "expert"):....

I will try to remember to refer you only to references that meet your high standards from here on.

 

[carbonzit]

I will try to remember to refer you only to references that meet your high standards from here on.

Well, seriously, your own explanation would suffice.

 

[dr.power]

Thanks a lot for all inputs,

Actually I have no idea if it is better to abolish the DB scale or not. I just see it WIDELY being used in electronics/ acoustics fields, So maybe there is a reason for increasing usage of it. But because it is some times vague when talking about dB so maybe thats why some guys (including me hating it), I myself think it was much better to use the ratios as thy originally are (for instance saying P2/P1 =10^-15 or V2/V1 is equal to something). I myself am not so comfortable with dB and always have to change it to its original ratio (P2/P1 or V2/V1) and see how many of gain it means in non dB scale or give a normalized value to the input (P1 or V1) and then see how much of incrasement I'll get in the output (ie in P2 or V2) by the said non dB gain, I just can treat dB as such a way I said, So I do not know why to use It while have to convert it to its original gain and ratio! I will be very glad if somebody experted in the field let me know how they treat with dB and if they always convert it to its original??!

Ok, Anyway I must say That I did not get my response to my first question in post #1??!
Today I came across The phrase dB in the datasheet of a microphone yet they did not explained if they were talking about dB power or dB voltage!

What to do?

 

[Mr RB]

And why do you consider dB obsolete? It is a convenient, logarithmic way to refer to attenuation or gain values. It thus allows simple addition of values to calculate the gain or attenuation of amplifiers or attenuators in series. Unlike the other terms you mention then is no alternate substitute for dB.
...

To answer your (and Carbonzit's question) I think it is largely obsolete way to express a number. It has some convenience for gain but is it really necessary?

Would you ask your boss for a 2dB pay rise? Would you express an economy as 1.2dB inflation?

There are plenty of things with some logarithmic properties in real life that we still express perfectly well in real numbers, and I see nothing wrong with saying something has a gain of 100 or a gain of 0.01. If the gain or ANY number is really large or small we already have a system for that.

As someone who dabbles in computer programming I have become more of a minimalist over the years. How many ways to we need to express a number? If var A is 50 in binary and var B is 7dB in binary hows does the microcontroller get a result? That's nasty. If A is 50 and B is 103 the microcontroller can crunch 50*103 or even use the inbuilt 8bit*8bit multiplier module and do it in one instruction.

I see log scales as useful for charting, for sure. And they have some convenience for really large/small numbers ie 120dB. But as far as expressing simple gains in 90% of the time I would prefer a simple decimal number; If my 'scope shows 1v in the amp and 7.3v out the amp, it's showing me a gain of 7.3. It's not showing dB.

Of if my Micro has a sine lookup table with values 0-50 and the user enters "display wave at 2.47dB" that's not nice. What the micro must do is FIRST convert the silly dB into a real, usable gain value then just multiply that gain value for every sample. Now if they had entered "display wave with gain 2.43" all would have been good.

... Regarding those other values you seem to think have no value (heh), I certainly don't want to start a long flame war (well, at least not here, but perhaps we could take it to a more appropriate forum?), but I happen to very much like our non-MKS units for everyday usage (foot, inch, pint, gallon, mile, etc.), and I can confidently tell you that we ('Merkins) are probably never going to move away from them. That was tried, and failed miserably, in the 1970s. Thanks, but no thanks. ...

Yes I'd like to avoid any issues of national pride etc but even Britain that was using "imperial" measurements long before you is slowly changing to metric, and even in the USA your military uses a lot of metric, as does much of the aerospace industry and electronics is now getting metricised (see most fo the new SMD parts!).

The short answer is; metric is better, the world slowly gets better. There's no need for two measurement units when one is better. And that pretty much ALSO applies to my opinion of dB.

 

[crutschow]

Mr RB, If you don't want to use dB, don't. But just because in your experience you don't see much reason to use dB doesn't mean that it's not useful or is obsolete. Numbers, when working with computers, are expressed as binary, octal, decimal, hexadecimal. Are all those really "necessary"? Probably not, but each has its uses. In you are a true minimalist then I assume you only use one of those when working with computers, right?

dB does have specific uses in electronics where it is a very handy way to designate gain and attenuation. I have worked on microwave systems, and the gain of amplifiers and the attenuation of attenuators and coaxial lines as well as amplifier noise figures are always stated in dB. It makes it easy to determine the total gain of a string of amplifiers and interconnect lines, since all the dBs simply add. I would assume that the design of lower frequency RF systems also us dB for the same reasons.