3 dB bandwidth

showed

than

TTS (

it and

At that kind of SPL - I'm sure you're mainly right !

Try 90-100 and it gets interesting.

Graham

It's excellent at relative nevertheless.

I once performed a blind test to 'line up' a multitrack audio tape recorder by ear though - lol !.

I managed to match the track one calibrated test tape playback levels to 0.3dB. This potentially explains a lot of the golden ears stuff. You have to *educate* your ear to work that well.

Graham

Actually, ( basically you're right ) 1dB is the volume change discernable by the supposed 'average' listener in average situations.

A 'trained ear' has little trouble trebling that accuracy.

Graham

You're right and few ppl know it or understand it.

I guess few ppl care to imagine why we call them *deci* Bels.

Simple answer.

It's 1/10 th of a Bel ! A Bel being the ratio that sounds to the ear like a doubling of intensity. Named after Alexander Graham Bell of telephony fame.

Graham

In valve amps ? Erk !

Essentially you're right. Pro-audio has always tried to make the passband as flat as possible. You won't find me arguing over that one for sure !

Graham

Well it would certainly do by back a service by not being such a nitwit. My Fender Twin Combo is just about impossible to carry single handed up the stairs.

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.

Dont forget the valve telephone SSB amps. Flat from 60Khz to 120Khz at

0.01% thd. They needed to be as they had 100s in a chain.

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.

You havent the slightest idea of what your talking about. You have learnt a phrases from somewhere with no idea as to how and when they are used.

The maximum power transfer theorem is useless and irrelevant for many/most power amp situations.

Loads for most power amps, if not all are determined from the maximum voltage and current ratings of the devices used. Period. For example, if a tube has a max voltage of 400V and a max current of 100ma, it wants to see a load of 4k, other wise one or the other aspect of the tube ratings will be wasted, hence a bigger more expensive device is being used. This is also true for RF power amplifiers. A transistor with an 15V max V rating and 1A max current will want to see a load of 15 ohms. That's what one does when one "matches".

Matched power means that one will waste 1/2 the power in the driving source. This is about the daftest thing one can do for a power amp.

Hands up those that actually know when impedance matching actually has a use?

Hint: "Singing" return loss on a 4 to 2 wire converter. Hint: Switching fast signals.

{snip more drivel}

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.

No it isn't ! See Ohms Law !

You really mean transistors ?

I studied them when I was 12 !

That's why toobs are only used by nitwits these days.........

< plonk >

Graham

I forgot to mention. Not in audio. RF I expect.

You can voltage match toobs in audio perfectly well. I have done so myself. I can design discrete vacuum tube, bipolar or fet ( junction or MOS ) circuitry and any amount of of IC circuitry inclusing DSP and its micrcontroller host requirements..

I trust that explains !

Graham

Where do you obtain the 0.5 value and what is the "potential" divider?

When an amplifier's input and output impedances are the same (for example

600 Ohms), and it's gain is adjusted with a 600 Ohm 0 dBm input signal, to provide a 0 dBm output into a 600 Ohm meter, There is neither gain nor loss. Only an impedance mismatch or an uncalibrated oscillator or meter will provide a different "view."

Don

Sigh. Another ineducable newsgroup participant.

The formula for calculating power decibels is:

Po G = 10 log10 ---- Pi

Where G = Gain in dB Po = Power output from the device Pi = Power input to the device

The formula for calculating voltage decibels is:

Vo G = 20 log10 ---- Vi

Where G = Gain in dB Vo = Voltage output from the device Vi = Voltage input to the device

Or. if you don't believe me, you can read this tutorial:

[ ]

or this one:

[ ]

You can join Aylward in my killfile. Assuming that he still posts to sci.electronics.design, the two of you can have your own little flame war without bothering me with the personal attacks that you think are an acceptable alternative to having a civil discussion on the topic at hand. Maybe later in life, after you have learned to read, write, spell, and count, you will have more success. True, these are rudimentary skills that many of us "normal" people take for granted that everyone has an easy time of mastering. But we sometimes forget that there are "challenged" persons in this world who find these things to be difficult. If I had known that this was true in your case then I would have never have exposed myself to what you wrote. It just wouldn't have been "right." Sort of like parking in a handicap space. I wish you the best of luck in the emotional, and social struggles that seem to be placing such a demand on you.

*plonk*

Sorry mate, they sound *great* on guitar !

I had the audiophool fraternity more in mind when I said that.

A bit of colouration / soft-compression and THD makes guitar sound wicked ! You pay the price in iron though.

Graham

I note that you redirected my reply to alt.dev.null !

I think that speaks for itself ! You're a usenet troll and clearly know nothing about electronics whatever.

Here's my original reply that you attempted to censor !

............................................................................................................

Guy Mac> The formula for calculating voltage decibels is:

Yes.

So when Zin = Zload ; Vo/Vi = 0.5

Log10 ( 0.5 ) = -0.301.....

G = -6.02.....

Shame you didn't bother to check the math wasn't it ?

Killfile ? I'd be pleased to be in the killfile of someone who doesn't even understand basic dB calcs !

Sorry to anyone else I have may offended. I loathe incompetence when it's stuffed up your nose by a half-arse who erroneously and elaborately wants to make a cheap point of correcting your supposed 'error' without even bothering to check his own sums first !

Graham

Once again you misdirected my reply to alt.dev.null.................................................

My reply is below

..............................................................................................................................

Don't waste my time on formulas I already know.

In audio 600 ohm working the source Z is 600 ohms and the load Z is also 600 ohms.

That forms a potential divider with a value of 0.5.

Thats -6dB !

Don't play semantics with your 10 log, 20 log stuff. I eat ppl like you for breakfast ! After which I'm employed to sort out their f*ck-ups.

Graham

Ro/(Ro + Rl)...dah...

Sure, if you "adjust the gain", i.e. increase it. That is beside the point.

The point is that given that one has an output from a *low* impedance source (i.e *the* typical case) that has an output matching resister, there will be a loss of 0.5 when there is that same valued resister connected to ground at the tap point, when referenced to the raw low impedance output.

People rarely design an amp with an inherent non resister resistance to match some standard load impedance. However, Marshal do that on some of their transistor guitar amps to simulate tube amp characteristics.

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.

I did realise you were not referring to guitar amps graham.

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.

Well, I will quible on "a lot".

Yes. The ear is quite sensitive under the right conditions, especially when trained. There is no doubt that pretty much all non musical minded individuals simply do not recognise what parts belong to what instruments in a finished piece of music. Ask them if the bass is too loud or soft, and they will have no idea. One has to learn to correlate prior knowledge of expected sounds with new sounds.

I do find it quite fascinating how I can pick out the individual intrument parts of music, and then sequence them. Most of the information is in the same bandwidth at similar levels, so how does the brain do it? Try and get a computer to do that and your nacked. It clearly uses come sort of correlation technique, but the details are still pretty amazing.

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.

My most-hated electronic calculator is also my best one. I bought it about 15 years ago; it is programmable and whatnot.

Ten years ago, when I turned it on, it told me, in clean black-on-white alphanumeric characters, that its battery was low and I had to change it. Nothing else worked any more.

The way an electronic pocket calculator is supposed to work when the battery goes out is to get progressively fainter until, after at least a decade of usage, you have to squint at the LCD from a certain angle to make out the numbers. Then you buy a new one.

Having lost interest in the fancy calculator, I quickly misplaced and forgot all about it. Recently, I found it again at the bottom of a junk drawer and turned on. "Battery low", still in full contrast.

All my LCD calculators now have PV cells.

robert