1 - 5 of 5 Posts

#### aceinc

·
##### Registered
Joined
·
453 Posts
Discussion Starter · ·
As will beocme obvious, I am not an electrical engineer.

It is my understanding that if you have a 4 ohm driver, and you input ~2 volts the driver should be consuming ~1 watt of power. If you then put a mic about 1 meter away you can try and match sensitivity specs of well documented drivers (I know that some documentation cheats by specifying 2.8 volts) Most mfrs test at 1khz, some do +-x db from a hz to b hz.

At any rate, my issue is if I hook up REW to an amp into a 4 ohm load and set the generator to play a sine wave at 20 hz, and measure the voltage across the speaker leads I get say 2.15v. If I then increase the freq to 44hz without changing anything else I get 2.65v. I assume that the difference is that the impedance is changing with frequency.

So how does one with some level of "honesty" measure the sensitivity of a speaker?

Please point me to "speaker measurement for dummies" if that's is what I need...

Paul

#### keelay

·
##### Registered
Joined
·
145 Posts
You are right speaker impedance varies over frequency. But so does amplifier response, even if you turn off all equalization. I'm considering doing the same thing you are - starting speaker measurements, so can't help much but am along for the ride.

Kyle

#### SturmMD

·
##### Banned
Joined
·
165 Posts
Amplifiers are ideally designed with a low output impedance so the driver impedance variation between those two measured values has minimal effect.
If your equalizer is set to flat this means that sum of all the equipment in your signal chain is giving you -1.8dB (20*log10(2.15/2.65) at 20Hz (with respect to 44Hz).
If you record both the SPL at 1m and also the voltage measured at the speaker you can calibrate out the changes caused by the varying input signal.

To "honestly" measure SPL you need a large room and to be far from walls and floor/ceiling - or to go outside. Also SPL measurements are usually taken at higher frequencies where the driver+baffle is radiating in 2pi space (forward) instead of 4pi (omnidirectional) giving 3dB higher response.

#### aceinc

·
##### Registered
Joined
·
453 Posts
Discussion Starter · ·
If I use the example in my original post and went outside would this affect the voltage, and hence the wattage?

In other words I am outside, or in a large anechoic chamber and I want to measure the "efficiency" of a speaker how do I do that?

Do I measure DB vs voltage/impedance (resistance?) at all frequencies and average? Use white/pink noise?

Paul

#### SturmMD

·
##### Banned
Joined
·
165 Posts
Since impedance varies with frequency that is why 2.83V is used. It is easier than 1 watt because this changes depending on what impedance it references, Re, nominal, @ 1Khz, etc.

Just measure the spl over many frequencies. Also record the voltage to compensate the spl measurements. Plot the results and eyeball an average. This will be close enough considering the accuracy of non-labratory grade spl meters.

1 - 5 of 5 Posts