More questions about output

Well... I was talking to a tech about output for my project and he mentioned some things which got me to thinking. I should've asked him but they popped in my head after we were done and he's really busy so I figured you guys may be able to answer.

1. This ones hard to word but I will try... For any db increase that is noticeable you need to double the output (wattage). I know this is true but I guess I'm looking for why this is true? If it's too deep in physics just let me know.

2. He also mentioned that a 50 watt tube amp is like running 150 watts solid state. If true, how?

3. I know matching amp speaker impedance is crucial but why is it that you can't hook an 8 ohm speaker rated for 100 watts to an amp that puts out 100 @ 4 ohms?

I guess I just want to understand it all better.

Once again, many thanks.

 

Not entirely true Ian. I'm running my 4-6 ohm speakers at 8 ohms on my power amp. It just sounds better than when hooked up in 4 ohm. Tubes however may be a different animal. Sorry I can't be of more help.

 

Thanks Dave. Any info helps.

 

Because the dB scale is logarithmic. Any time you increase the dB by 3, you double the level. So if 1 watt of power output makes 90dB at 1 meter, it will take 2 watts to make 93, 4 watts to make 96, etc.

It's just the nature of the dB scale...

 

Decibel (dB) is a measure of relative size.

For power ratio:
dB value = 10 log (P1/P2)

For amplitude ratio such as voltage (or current) gain:
dB value = 20 log (V1/V2)

The 3 dB value MITBeta mentioned means twice the power:
P1 = 2 P2
10 log (2P2/P2) = 10 * log(2) ~= 3

I was told that human ears sense the loudness in logarithm.

 

That's pretty much it. Specifically, each 3dB increase requires a doubling of wattage. The explanation you were given only differs in semantics as it's generally accepted most folks don't notice an increase below about 3dB. Here's a handy-dandy formula for calculating the increase in decibels for a given increase in wattage:

dB= 10 x log (P1/P2)

where:

P1 is the upgraded level of wattage
P2 is the wattage you started with

Obviously, doubling the wattage gets pricey very quickly, so this formula will allow you to calculate the dB increase for any increase in wattage.

Although watts are watts (quality aside), tube watts tend to *subjectively* sound like double (triple is a bit high) the equivalent solid-state watts. Tube clipping is gentler than solid-state clipping, which creates the illusion of more power. Good tube amplifiers seem to have a way of portraying dynamic contrasts better, too, which can also create the illusion of more power.

You *can*, but there are some considerations. As Dave alluded, there are differences between tube and solid-state amplifiers, as well as the occasional qualifier within each type.

Tube amplifiers use output transformers (unless it's an OTL design, which is rare) to couple the output of the amplifier to the speaker load. Tapping the transformer at different points in the wiring (4 ohm tap, 8 ohm tap, 16 ohm tap, etc) allows for maximum output when the tap chosen matches the nominal impedance of the speaker.

Solid-state amplifiers don't use output transformers (unless it's a McIntosh ), so the speaker load is directly coupled to the amplifier's output. This direct coupling results in the amplifier/speaker becoming a circuit, which accounts for the wide variance in output into different speaker loads.

If the amplifier in your example is solid-state, you can only get 100 watts if the impedance dips to 4 ohms...unless it uses output transformers, which is limited to McIntosh AFAIK. That wattage is rated into a tougher 4 ohm load. You'll get less than 100 watts into an easier 8 ohm nominal load. This is the effect of the amplifier/speaker behaving as a circuit.

If the amplifier used with the speaker you described employs output transformers, using the 4 ohm tap will result in less than 100 watts output except when the load dips to 4 ohms, although the variation in output at different loads will be much less than with amplifiers without output transformers. The highest average output from the amplifier across it's operational bandwidth will result from using the tap that corresponds to the nominal load of the speaker. However, using a lower tap will result in a lower output impedance, which will mean less variation in frequency response as the load presented by the speaker varies. That *could* result in better overall performance, but is system-dependent.

So, to sum it up, if you're using a solid-state design (provided it doesn't employ output transformers) select speakers with a wattage requirement that match what your amplifier can deliver into the nominal load. If you're using a tube design (provided it isn't an OTL design), selecting the output tap that matches the nominal impedance of your speakers will *probably* work best. In some cases, however, you *might* prefer the greater linearity of using a lower tap than that normally dictated by the nominal impedance of the speaker...provided you have the wattage to spare.

These are the basic concepts behind your questions. It goes deeper, but my schooling was in business, so you'd have to outsource further detail.

 

Thank you very much gentlemen
I'll make sure to mention you guys in my presentation tonight

 

Another thing to be careful about in speaker impedance and amplifiers is that some amps do not like to see too low of an impedance (no speaker has a perfectly flat impedance curve, it varies with frequency). Some amps can tolerate a larger variation in speaker impedance than others. If the amp is not stable driving low speaker impedance loads it will distort or in extreme cases start to act like an oscillator.

 

Hey guys.

Just to let you know, you may have just helped me snag an A on the presentation tonight.

Once again, thank you so very much for your help with this. I couldn't have done it without you guys.

I gave you all a shout out at the end too

 

Ah...I thought you were just kidding about the presentation! I figured your "project" was an upgrade or something. Apologies if I missed something from another thread.

That said, great news! Glad it went well.

 

Good job, Ian! Are our names and this thread going to appear in the reference section of your report?

 

Of course. It goes without saying.