Yes, of course, read the good thread by Plano. My question is not where in the system we amplify or attenuate. My question is different.
Speakers have different sensitivities, but I think, their amplitude is reasonably linear with respect to power.
Amps have gain, but as it is a ratio, and it varies amp to amp. The higher the power, the higher the gain.
So, I have a mains amp with 23 dB gain.
I have a sub amp with 27 dB gain.
I can use attenuators so at some point, say 1W, the amplitude to either side of the crossover is the same. But what happens at 5W? 50? Am I not driving the sub proportionally much louder than the mains?
Speakers have different sensitivities, but I think, their amplitude is reasonably linear with respect to power.
Amps have gain, but as it is a ratio, and it varies amp to amp. The higher the power, the higher the gain.
So, I have a mains amp with 23 dB gain.
I have a sub amp with 27 dB gain.
I can use attenuators so at some point, say 1W, the amplitude to either side of the crossover is the same. But what happens at 5W? 50? Am I not driving the sub proportionally much louder than the mains?
Um, imagine a power amp with a gain of 27db, it puts out X volts at full tilt. Now consider 16, 8, 4, and 2 ohm loads on that. Each of those draws its own amount of power. Gain and power not directly together.The higher the power, the higher the gain.
I've not been on the forum for a while. I was "passing by" and saw this.
I think you have to be far more specific when talking about speaker sensitivity
e.g. Max power handling 200 Watts, 8 Ohm Impedance, sensitivity 88 dB SPL at 1 metre at 1 Watt input
Only when you give that full spec, is it then possible do any calculations. In this case, if the speaker can truly handle 200 Watts, then the SPL out of the speaker at 1 metre at 200 Watts input = 88 dB + 23 dB = 111 dB SPL. The 23 dB comes from the fact that 200 Watts is 23 dB relative to the 1 Watt sensitivity of 88 dB.
Working backwards in the system, you then have to have a power amp with an output stage that's capable of delivering 200 Watts into an 8 Ohm impedance, which would be 40 volts measured across the terminals into the speaker.
But you can't really say that the power amp has a gain of 23 dB or any dB, all you can say is that it will deliver up to the max volts to give 200 Watts into the 8 Ohm impedance, without distortion, and that any other volts less than the max into the speaker, also have no distortion or extraneous noise on the volts - however the stages in the system are organised.
In the past, I have read the thread you refer to, and if I understand your point, you might be suggesting that the output stage itself might be "cleaned up" a bit by running it at full output voltage into something resembling a variac, and then using the variac to turn up and down the voltage that's going into the speaker ? (assuming that the amp is "cleaner" at full power than at lower powers and that it always sees the correct impedance on the input to the "variac", and that the "variac" will always be able to handle the speaker impedance)
I think you have to be far more specific when talking about speaker sensitivity
e.g. Max power handling 200 Watts, 8 Ohm Impedance, sensitivity 88 dB SPL at 1 metre at 1 Watt input
Only when you give that full spec, is it then possible do any calculations. In this case, if the speaker can truly handle 200 Watts, then the SPL out of the speaker at 1 metre at 200 Watts input = 88 dB + 23 dB = 111 dB SPL. The 23 dB comes from the fact that 200 Watts is 23 dB relative to the 1 Watt sensitivity of 88 dB.
Working backwards in the system, you then have to have a power amp with an output stage that's capable of delivering 200 Watts into an 8 Ohm impedance, which would be 40 volts measured across the terminals into the speaker.
But you can't really say that the power amp has a gain of 23 dB or any dB, all you can say is that it will deliver up to the max volts to give 200 Watts into the 8 Ohm impedance, without distortion, and that any other volts less than the max into the speaker, also have no distortion or extraneous noise on the volts - however the stages in the system are organised.
In the past, I have read the thread you refer to, and if I understand your point, you might be suggesting that the output stage itself might be "cleaned up" a bit by running it at full output voltage into something resembling a variac, and then using the variac to turn up and down the voltage that's going into the speaker ? (assuming that the amp is "cleaner" at full power than at lower powers and that it always sees the correct impedance on the input to the "variac", and that the "variac" will always be able to handle the speaker impedance)
27dB gain plus a 4dB attenuator is exactly 23dB gain. It's all linear (if it weren't it would be gross isdortion.I have a mains amp with 23 dB gain.
I have a sub amp with 27 dB gain.
I can use attenuators
Yes, different 'sensitivities' mean you won't get equal outputs. But the difference is linear until things break-up (hi-fi guys don't do that).
The attenuator is also a linear ratio, but I keep getting wrapped around a picture of different ratios. Guess I better do a Spice simulation to un-confuse myself.
There is different ratio involved yes: for power this is 10Log ( +3db) but for voltage it is 20Log (+6db).
An expressed in Db value gives a way to compare between both units ( or other ones): a doubling of voltage equal +6db, if you want same gain in db from amplifier ( power) you'll need 4 time the initial power ( 3db +3db).
Things change when a definition is added to db: eg db SPL, or dbu.
In these case this is not a 'floating ratio' anymore as there is point of reference ( 94dbspl: 1pa, +4dbu: 1,23v rms or 0dbu: 0,7746v rms).
An expressed in Db value gives a way to compare between both units ( or other ones): a doubling of voltage equal +6db, if you want same gain in db from amplifier ( power) you'll need 4 time the initial power ( 3db +3db).
Things change when a definition is added to db: eg db SPL, or dbu.
In these case this is not a 'floating ratio' anymore as there is point of reference ( 94dbspl: 1pa, +4dbu: 1,23v rms or 0dbu: 0,7746v rms).
The key is here. It's linear. If it weren't - audio would be terrible. But I understand where your confusion comes from.It's all linear (if it weren't it would be gross isdortion.
Or if you prefer to think in decibels, remember that 23 or 26 dB gain is the overall gain of the amp.
If the input signal to either amp increases by 3dB, the amp output increases by... 3dB
The overall level is shifted up by 23 or 26 dB depending on the amp, but the output change is still 3dB.
If the input signal to either amp increases by 3dB, the amp output increases by... 3dB
The overall level is shifted up by 23 or 26 dB depending on the amp, but the output change is still 3dB.
the way i see it, adjust the input level to the sub amp to enable blending of the sub output to the main output (by ear or by f/r measurement)
then see if there is clipping/noise/etc issue.
then see if there is clipping/noise/etc issue.
Just a bit more input
Two pairs of speakers
1. Max power handling 200 Watts, 8 Ohm Impedance, sensitivity 88 dB SPL at 1 metre at 1 Watt input
2. Max power handling 400 Watts, 4 Ohm Impedance, sensitivity 88 dB SPL at 1 metre at 1 Watt input
A power amplifier capable of handling both pairs.
With the 8 Ohm speakers, at 200 Watts into them, that would be 40 Volts and 5 Amps, and an output SPL of 111 dB.
Let's assume that the power amp has a "gain" of 23 dB. That would mean that the input is 1 Volt.
With the same 1 Volt into the power amp, but with the 4 Ohm speakers connected, presumably the output is still 40 Volts but the power into the 4 Ohm speakers is now 400 Watts i.e. 40 Volts and 10 Amps, and an output SPL of 114 dB.
So the same "gain" of 23 dB, and same input signal, yet a gain of an extra 3 dB SPL. Is this the problem of trying to use "gain" when speaking about a power amplifier ?
Two pairs of speakers
1. Max power handling 200 Watts, 8 Ohm Impedance, sensitivity 88 dB SPL at 1 metre at 1 Watt input
2. Max power handling 400 Watts, 4 Ohm Impedance, sensitivity 88 dB SPL at 1 metre at 1 Watt input
A power amplifier capable of handling both pairs.
With the 8 Ohm speakers, at 200 Watts into them, that would be 40 Volts and 5 Amps, and an output SPL of 111 dB.
Let's assume that the power amp has a "gain" of 23 dB. That would mean that the input is 1 Volt.
With the same 1 Volt into the power amp, but with the 4 Ohm speakers connected, presumably the output is still 40 Volts but the power into the 4 Ohm speakers is now 400 Watts i.e. 40 Volts and 10 Amps, and an output SPL of 114 dB.
So the same "gain" of 23 dB, and same input signal, yet a gain of an extra 3 dB SPL. Is this the problem of trying to use "gain" when speaking about a power amplifier ?
Of course it's not really the amp gain that's to blame. It's current flow into the lower impedance that means more SPL. The amp voltage gain hasn't changed at all.
But the problem is that specifying the voltage gain of a Power Amplifier is meaningless as it gives no indication of the SPL out of the speaker. You can only derive the SPL by having the speaker sensitivity and the Power out of the Power Amplifier. Obviously you won't get max power without sufficient input voltage into the Power Amplifier, but there's no real issue about input voltage because you have a volume control in the pre-amp.
- Home
- General Interest
- Everything Else
- System gain structure