Has anyone measured their speakers sensitivity?

[insider9]

Would it be as easy as measuring 2.83V from amp speakers terminals and then taking an acoustic measurement at 1m at 1kHz? (that's for 8 ohm speakers)

A little afraid to put multimeter into amp terminals when it's outputing signal.

 

[Gray]

I don't mean you're a weirdo (you'll be pleased to hear) but like a scientist who likes to prove things.

Let us know how you get on, worst case it'll be better than the usual indoor fireworks you can buy.

Seriously, with your meter set to the voltage range you can't do any harm (AC as opposed to the DC you used for the offset measurements)

No doubt you'll take the right precautions of starting from zero output with your probes attached to the terminals to avoid inadvertent clumsy shorts.

Knowing you, you'll be building an anechoic chamber for the test. Have fun.

 

[lindsayt]

My multi-meter is hopeless at measuring 1 khz AC. It's ok at measuring 50 hz AC.

I wouldn't trust my sound pressure meter at 50 hz, due to falling sensitivity at bass frequencies. It's OK at 1 khz.

I'm not going to go out and buy an HP signal generator just to check my speaker sensitivity.

And a 1khz sine wave at 2.83 v would be an unpleasant, deafening, noise through my bigger speakers. Ear muffs required.

 

[nick8858]

And what would the point of doing this be?

 

[insider9]

And what would the point of doing this be?

A number of points actually
- how do these correspond to specs?
- how loud is it actually at 1W of power?
- what amp setting (volume) corresponds to 1W?
- what minimum amp required to run the speakers?

All of the above and trying to estimate at what continuous power I'm listening to. And whether if I come across a cheap low powered Class A amp I should try it, or not

 

[insider9]

My multi-meter is hopeless at measuring 1 khz AC. It's ok at measuring 50 hz AC.

I wouldn't trust my sound pressure meter at 50 hz, due to falling sensitivity at bass frequencies. It's OK at 1 khz.

I'm not going to go out and buy an HP signal generator just to check my speaker sensitivity.

And a 1khz sine wave at 2.83 v would be an unpleasant, deafening, noise through my bigger speakers. Ear muffs required.

Good point.

But I could measure electrically a 50Hz sinewave to output 2.83V. Then change to 1kHz once volume is established and connect speakers. Measure speakers at 1kHz to get the results.

It's not only room interaction at 50Hz I'm worried about but impedance changes. With dips to 4ohm I'd be measuring 2W not 1W.

 

[insider9]

I don't mean you're a weirdo (you'll be pleased to hear) but like a scientist who likes to prove things.

Let us know how you get on, worst case it'll be better than the usual indoor fireworks you can buy.

Seriously, with your meter set to the voltage range you can't do any harm (AC as opposed to the DC you used for the offset measurements)

No doubt you'll take the right precautions of starting from zero output with your probes attached to the terminals to avoid inadvertent clumsy shorts.

Knowing you, you'll be building an anechoic chamber for the test. Have fun.

 

 

:biggrin:

Thanks for reassurance.

 

[davedotco]

Even at 1 metre, the in room reflections would make a mockery of sensitivity measurement. Back in the seventies, JBL Professional, the premium loudspeaker (components) manufacturer in the industry at that time used to measue sensitivity in the open air. No anechoic chamber in those days.

A horizontal platform was built several feet of the ground with an interchangeable baffle to take different drive units facing directly upwards. Measurements were taken at 30 feet, with pink noise, band limited to suit the application of the driver under test.

Attempting to measure sensitivity yourself, in a normal room is fraught with difficulty and I doubt that they would be meaningful, might be amusing and quite educational though. Dispersion will be a big variable, as will bandwidth, both will affect the 'in room' response, probably more variation in that than in actual sensitivity.

 

[insider9]

Thanks Davedotco! I've not considered this. Taking the speakers outside isn't possible. I wouldn't risk damaging them.

My room is acoustically treated though and early reflections in particular have been reduced to not being a factor. That's at listening position. I could make sure the reflections are kept to minimum. Especially at 1m distance it would be rather easy. It would however reduce an already low reverberation. I take it this could have a bigger impact and showing up as speakers being even less sensitive.

 

[davedotco]

Thanks Davedotco! I've not considered this. Taking the speakers outside isn't possible. I wouldn't risk damaging them.

My room is acoustically treated though and early reflections in particular have been reduced to not being a factor. That's at listening position. I could make sure the reflections are kept to minimum. Especially at 1m distance it would be rather easy. It would however reduce an already low reverberation. I take it this could have a bigger impact and showing up as speakers being even less sensitive.

The problem is reflection and reverberation, irrespective of room treatment, a lot of the energy collected by the mic will be reflected, the mic will not discriminate the way the ear does, so later reflections that the ear does a good job of rejecting, are still measured.

Bass bandwidth is an issue too, bass radiation paterns are near spherical, so think of all that bass energy reaching the mic, little of it will be 'direct' but it will still be measured and generally speaking, increase measured sensitivity. In room measurements will, because of the reflections, be much higher than measurements taken in free space, this is what is known as room gain and it can mess your measurements up a lot.

Remember, even in a treated room, once you get beyond a metre or so all measurements will be of the room response, not the speaker and in addition the speakers off axis response will make a big difference too.

All is not lost though, if you can set up different speakers in exactly the same way, in the same room, you should be able to get some very interesting comparisons but you will not get close to measuring 'real' sensitivity, for that you will need an anechoic chamber or a large open space in the Calefornian sun,

 

[insider9]

Thanks Davedotco! I've not considered this. Taking the speakers outside isn't possible. I wouldn't risk damaging them.

My room is acoustically treated though and early reflections in particular have been reduced to not being a factor. That's at listening position. I could make sure the reflections are kept to minimum. Especially at 1m distance it would be rather easy. It would however reduce an already low reverberation. I take it this could have a bigger impact and showing up as speakers being even less sensitive.

The problem is reflection and reverberation, irrespective of room treatment, a lot of the energy collected by the mic will be reflected, the mic will not discriminate the way the ear does, so later reflections that the ear does a good job of rejecting, are still measured.

Bass bandwidth is an issue too, bass radiation paterns are near spherical, so think of all that bass energy reaching the mic, little of it will be 'direct' but it will still be measured and generally speaking, increase measured sensitivity. In room measurements will, because of the reflections, be much higher than measurements taken in free space, this is what is known as room gain and it can mess your measurements up a lot.

Remember, even in a treated room, once you get beyond a metre or so all measurements will be of the room response, not the speaker and in addition the speakers off axis response will make a big difference too.

All is not lost though, if you can set up different speakers in exactly the same way, in the same room, you should be able to get some very interesting comparisons but you will not get close to measuring 'real' sensitivity, for that you will need an anechoic chamber or a large open space in the Calefornian sun,

Thank you! Much appreciated. Absolutely agree and know where you're coming from. I'll give it a go anyway to see what results I get.

My thinking is when using 1kHz sinewave there will be no gains as there could be in bass regions. Measured early reflections (all frequencies under 20ms) are about 5% of direct sound, so not much gain there. Will have to check how they behave at 1kHz. Before room treatment reflections were 20-30%. The ceiling was most reflective surface, with just over 30% of the sound coming from there. It wouldn't happen with Classics with their limited vertical dispersion but I though it's worth a mention.

Finally in room reverberation is around 0.2-0.25s throughout midrange. I could have it easily under 0.2s for purpose of these measurements. I don't have it as low normally as it results in a dry sound. Particularly audible on vocals where notes feel cut short on some recordings with little reverb present.

Having all this in mind I guess I could get fairly close. But ultimately it will be interesting to see what do I need to set an amp to get 1W out of it and how does it sound like. I predict like with many class A/B designs the bias might be set that the first "enter number here" is pure class A. It does sound mighty nice at low volume. Not sure if it's related to that.

 

[QuestForThe13thNote]

One question, why want to measure?

 

[insider9]

One question, why want to measure?

See post #5

 

[QuestForThe13thNote]

Why do you want to know if it conforms to specs insider. Also minimum power needed as I’m sure any commercial amplifier would drive any speakers to get sound out of them, whether it does it well or sufficiently another matter. Are you trying to achieve something.

I can understand if you want to check your car if the mpg is as advertised given you want to know if you are saving money or not etc.

 

[newlash09]

I would be very interested in your findings. Will get to see how much power we actually need at a particular db level. Though it can't account for dynamic peaks, still would be a good way to get an idea of the relationship between power, sound level and qouted sensitivity. All the best for a safe experiment

 

[insider9]

Just curious. Specs are often either misrepresented or providing incomplete information so if it's something easy to do why not try it.

Example
Focal Sopra 2 is an 8 Ohm load with 91dB sensitivity at 1m (2.83V).

These dip to 3.1 Ohm according to specs and 2.2 Ohm according to reviews. If you apply 2.83V to 2 Ohm load you will be supplying 4W not 1W. That's quite a difference.

 

[QuestForThe13thNote]

I’d only really be interested if it had a point to buying decisions and selecting one amp over another etc with given speakers, positioning, and I’m not sure it does....

 

[insider9]

I would be very interested in your findings. Will get to see how much power we actually need at a particular db level. Though it can't account for dynamic peaks, still would be a good way to get an idea of the relationship between power, sound level and qouted sensitivity. All the best for a safe experiment

Sure not accounting for peaks but continuous power doesn't account for peaks either.

 

[steve_1979]

And what would the point of doing this be?

We're audiophiles. We do weird **** like this for fun.

 

[QuestForThe13thNote]

And what would the point of doing this be?

We're audiophiles. We do weird **** like this for fun.

i wouldn’t unless I understand it and it has a purpose but mostly just about enjoying the music for me. But each to their own. I could take my car apart now and find out if it has a 5mm or 10mm washer in an inaccessible part of the engine I couldn’t reach. Have been wanting to find out for years about that washer. lol.

 

[insider9]

And what would the point of doing this be?

We're audiophiles. We do weird **** like this for fun.

 

Haha, so true

 

[Diamond Joe]

Each to their own but I can't help thinking that life's just a bit too short...

 

[newlash09]

Do you also plan to leave the multimeter plugged in and listen to music at your normal listening level to see how much real time power your speakers are consuming. Considering that bass is the most power sapping spectrum to reproduce, will you also be listening to bass heavy tracks to get a better estimate of the worse case scenario power consumption in your room with your setup

 

[insider9]

Do you also plan to leave the multimeter plugged in and listen to music at your normal listening level to see how much real time power your speakers are consuming. Considering that bass is the most power sapping spectrum to reproduce, will you also be listening to bass heavy tracks to get a better estimate of the worse case scenario power consumption in your room with your setup

It would probably fry the multimeter. Someone correct me if I'm wrong. I wouldn't dare to do it.

It's not really the right equipment to do it but it should handle low voltage

 

[Gray]

Do you also plan to leave the multimeter plugged in and listen to music at your normal listening level to see how much real time power your speakers are consuming. Considering that bass is the most power sapping spectrum to reproduce, will you also be listening to bass heavy tracks to get a better estimate of the worse case scenario power consumption in your room with your setup

It would probably fry the multimeter. Someone correct me if I'm wrong. I wouldn't dare to do it.

It's not really the right equipment to do it but it should handle low voltage

Your meter is (almost certainly) capable of measuring beyond mains voltages, you won't harm it with anything you'll measure accross speaker terminals.