What's the maths

[Ben Lawrance]

Evening all,

 

I've been scouting about on the net for a bit this evening trying to see if there is a formula for working out gain attenuation vs output power with an amplifier.

 

What I am trying to work out is thus:

 

I have a QSC PLX3002 (900w p/c @ 4ohms) running at -4dB on my DCX. What sort of output should I be expecting?

 

The problem I have is that I'm trying to make up another rack (exactly like the one I have, to have a matching pair), but I can't get hold of a 3002 for love not money. However, I can get a PLX2402 (700w p/c @ 4ohms)

 

I just wondered if the -4dB dropped the output of the 3002 far enough to be able to use a 2402 at full in it's place.

 

 

I apologise if I've not worded this very well, I know what I mean, however explaining it seems a bit difficult.

 

Cheers

Ben

[Shaggy]

Most modern amps work with a constant gain, rather than an input sensitivity. This means that rather than the output on an amp changing for the same input with different loads, or different amps in the same range giving different output with the same input, the voltage output is constant accross the range in most cases. This is the case with the PLX series, so up until you hit the limits of the amp both a PLX3002 and a PLX2402 will give the same output.

 

To put it another way, turning the amp down 4db doesn't decrease the power it puts out, just increases the voltage at the input required to get the same power, so you can do a straight swap until you start running the amps too hard and clipping them. Just before then is where you want to put your limiter (or earlier depending on the load) so it is the limiter you need to adjust not the gain for different amps.

[paulears]

Shaggy - I've read this a few times and am really confused? Modern amps all give the same output at the same input? Until their limit is reached? Turning down 4db doesn't decrease the output, just the input?

 

Sorry - but every time I think I have a handle on this, it runs away again.

 

You can do a straight swap of different power amps and it works? Lastly - you use a limiter to adjust the output, not the input control?

[weatherhead]

To put it another way, turning the amp down 4db doesn't decrease the power it puts out, just increases the voltage at the input required to get the same power,

 

surely this is splitting hairs isn't it? I mean, the effect is identical, in that for a given input signal when you turn the knob down we get less volume.

 

Also, isn't this what one might describe as an "input sensitivity" method of gain structure? In which case, I don't understand your point at the start that "modern amps don't work by input sensitivity".

Please could you clarify this for us?

 

In reply to the OP, I would say a few things: when you say "what kind of output should I be expecting?", do you mean in RMS power output, dBV, dB(A), or "loud", or what? In general, one of the characteristics of an ideal amplifier would be to produce as far as possible a linear relationship between input signal level and output signal level. This is closest to being realised AFAIK in class A amplifier circuitry. However, your amps are very unlikely to be class A circuitry. I'm not familiar with the particular model, however. Your amp is likely to be class AB which still is fairly linear, but not quite as good as class A.

 

I would hazard a guess that the response curves of your two amplifiers are likely to be similar, but scaled relative to each other.

 

Is your 4dB attenuation meaning 4dBu at your DCX? Or does your DCX "know" the dBO scale of your amp (I'm not familiar with these devices either!) In the first case,a -4dBu signal (i.e. a signal that was 0dB before you attenuated) means that your input voltage to the amp will be 10^(-4/20)*0.775 , in other words 0.49V. In general, power is proportional to the square of the amplitude, but in audio and ac circuits it can get complicated. Anyway, as a rough figure -4dBu means losing nearly half your signal amplitude. You would expect that since there is only 200w less power per side in one of the amps, you would *cough cough* expect that 4dBu attenuation is "enough" to allow you to use the smaller amp without clipping.

 

The only way to really "work it out", would be to have the plotted response curves of the amplifiers, and check this 0.49V input figure against the output voltage.

 

EDIT: got a few things a*** about face there, just corrected it after having read it back

[Shez]

Whenever I'm working out amp / speaker/ LMS levels, I convert everything to dBu as it makes life so much easier. Trying to work in volts or powers makes life unnecessarily complicated.

 

E.g. Speaker can handle +35dBu; amp gain is 32dB, max amp input is +5dBu.

If you're in the amp-input-attenuators-turned-all-the-way-up camp, you'd set the LMS limiter at a couple of dB below +3dBu (to allow for headroom). If not, perhaps turn amp attenuators down by 12dB and set the limiter to a couple of dB below +15dBu.

 

As Shaggy said, the PLX series all have the same fixed gain (32dB); it's the maximum input levels that differ. There's actually only 6dB between the smallest and largest of that range; there's only about 1dB difference between the input sensitivities of the 2402 and the 3002.

 

So my answer would be to set the amps up identically and just drop the limiter by about 1dB on the output that's driving the 2402. You'll get the same levels out of the system, it's just the protection that'll kick in slightly sooner on the smaller amp.

[David A]

If the output voltage across the range is the same and the speaker impedance is the same, I think you are breaking Ohms law.

[mackerr]

If the output voltage across the range is the same and the speaker impedance is the same, I think you are breaking Ohms law.

In what way? If the output voltage is the same, and the speaker impedance is the same, it just means that the power is the same. How does that violate Ohm's Law? The point in this case is that the 2 PLX series amps have the same gain, which is all that matters. At any given input level they will be at the same output level, until you surpass the maximum input level of the smaller amp at which point it will clip before the bigger amp. Up until that point they will be identical. Maximum output power is just that, MAXIMUM output power. It is not GAIN, nor power under normal operating conditions.

 

Mac

[weatherhead]

Speaker and amplification systems cannot in general be regarded as ohmic resistors, precisely because a large portion of their response curve is not linear. Ohm's law would only apply to the linear portions in any case.