At Pano's suggestion on another thread...
...I ran the test.
My usual amp - a 40wpc into 8 ohms JVC JA-S44 integrated is out of commission. So I used a desktop NAD-D3020 (30wpc into 8 ohms) instead and turned up the volume as loud as I'd likely listen (which, admittedly, isn't saying much).
I connected one speaker only - a Tannoy Monitor Gold 15 of ~90dB efficiency.
Playing the 120 Hz test tone from my mobile phone connected to the amp via bluetooth, the voltmeter climbed and held at 6.75 volts.
Can this be right? Or did I setup/measure incorrectly?
...I ran the test.
My usual amp - a 40wpc into 8 ohms JVC JA-S44 integrated is out of commission. So I used a desktop NAD-D3020 (30wpc into 8 ohms) instead and turned up the volume as loud as I'd likely listen (which, admittedly, isn't saying much).
I connected one speaker only - a Tannoy Monitor Gold 15 of ~90dB efficiency.
Playing the 120 Hz test tone from my mobile phone connected to the amp via bluetooth, the voltmeter climbed and held at 6.75 volts.
Can this be right? Or did I setup/measure incorrectly?
Sorry for the delay in replying. It sounds like you did the test correctly, tho the voltage does seem high. 6.75 volts would indicate that you need an amp rated at least 45 WPC, which you didn't have. Did you also use your mobile phone as the music source?
If you are testing a solid state amp, then you don't need to have the speaker connected for the test. If a tube amp, then it should have a load. It would be interesting to know if you get different readings with the Tannoy connected or not.
If you are testing a solid state amp, then you don't need to have the speaker connected for the test. If a tube amp, then it should have a load. It would be interesting to know if you get different readings with the Tannoy connected or not.
Isn't the signal well compressed by virtue of being played over BT, regardless of the source?
Yes, could well be. Which would certainly throw off the numbers. It's worth testing for. If the test tone is played and the power amp volume raised slowly up to the setting that was used in the first test, then the voltage ought to come up smoothly, then suddenly get much higher. That would indicate clipping.
Of course it's easier to see the clipping on a oscilloscope, but seeing the voltage suddenly jump is also a good indication.
It could be that the source music is very quiet.
Ground point 9, did you play music at your listening level, and then switch to a test tone to measure the voltage? Perhaps choose different music genres to find your listening level.
I would also check if your player has a "normalize" function, so that different songs have the same average volume. In Spotify, the setting is "Normalize volume - Set the same volume level for all songs and podcasts". I wonder if this normalize function would offset the test results significantly, since the average in music is much lower than the average for a sine wave?
Ground point 9, did you play music at your listening level, and then switch to a test tone to measure the voltage? Perhaps choose different music genres to find your listening level.
I would also check if your player has a "normalize" function, so that different songs have the same average volume. In Spotify, the setting is "Normalize volume - Set the same volume level for all songs and podcasts". I wonder if this normalize function would offset the test results significantly, since the average in music is much lower than the average for a sine wave?
Interesting, could be.But wouldn't a normalize or volume match attenuate the sine wave? The test tone is pretty hot, at 9dB below the max allowed for a sine.
I don't know. It probably would because the average for a sine wave is so much higher than the average for music. It is a variable though.
The initial procedure should be followed for consistency, but I can't remember it exactly. I just think that for consistency's sake, the normalize function is probably better left off. This is my humble, and likely meaningless opinion.
My measurement was 12Vac. I'm running a big ~1980 Accuphase amp (modified) into 1978 Tannoy Ardens (modified) in a room 9 x 11 x 3 metres (~10,000 cubic feet). When I play loud, I like to play trouser flapping loud, ~110dBC on my SPL meter at 3.5m (12'). Not a hint of harshness or distortion, but my ears don't stand up to that volume for very long before the highs roll off.
Maximum loud is my normal if I am listening to music, and I tend play something related to my professional career in theatre and attempt to play at the volumes that it was reproduced when I was first exposed, for example Wim Merten's piano, drums and vocal composition 'The Power of Theatrical Madness' which was played at over 125dB in a 700 seat auditorium. (I actually expected the audience to walk out, but they didn't.) In the auditorium it felt like being inside the piano with the hammers hitting your chest.Thanks John. That does seem LOUD.
Can you measure again at what you would call a normal volume? It would be interesting to see what's the difference in voltage between your normal and max loud.
Not that I would know but it's a bit like being a drug addict I guess, I need a fix every now and again. I do have the luxury of no neighbours within a mile or so.
If I am listening to music with my partner then the level will be well below 2V and still apparently loud at our 3.5m listening position, ~mid 80s dB.
Hello Pano, thanks for this very helpful test.
I did it on my mid-bass few months ago and got numbers that I was expecting but it was a good check.
I redid it recently once I changed few things in the system, new dsp tune and new amps with less power.
Since it's a 5 way active system I wanted to get a little more precise so needed a measurement per amp.
Also I don't really have a "loud" level reference, it moves all around so I kind of used an opposite way, to get an idea of the max power needed at -0db right now.
It's more to get a reference and maybe downgrade again the amps, based on these results I could refine future choices and maybe ditch the party mode, when I came close to -6/-3db.
I first searched for the right sine-waves frequency to be used per bandpass.
Guessing that the maximum power would be measured at the maximum FR of each bandpass in dsp, including crossover and eq (but maybe I'm wrong here).
So I found some that are close enough to the middle of each bandpass with: 50Hz/160Hz/400Hz/1Khz & 8Khz.
And played them at -30/-20/-10db on the main dsp and measured each (I could have used only one level but wanted to confirm my calculations).
Using the rapidtables it gave me these needed power numbers to reach -0db:
It should not radically change except maybe on mid-bass that are wired in //.
Needless to say I use less than 1 watt most of the time near -20db, and less than 10w when loud.
1500 funny watts for the subs, I don't need that and I'm learning how to use the compressor anyway as the current 500w are more than enough, passed -10db, it already rattles everything around.
The midbass number is a surprise, as it's much lower than few months ago with your version of the test with 120Hz (it asked for 360w to reach -0db). But I didn't have the subs at that time, so I had more boost there.
Do you see anything wrong in the process? Thx!
I did it on my mid-bass few months ago and got numbers that I was expecting but it was a good check.
I redid it recently once I changed few things in the system, new dsp tune and new amps with less power.
Since it's a 5 way active system I wanted to get a little more precise so needed a measurement per amp.
Also I don't really have a "loud" level reference, it moves all around so I kind of used an opposite way, to get an idea of the max power needed at -0db right now.
It's more to get a reference and maybe downgrade again the amps, based on these results I could refine future choices and maybe ditch the party mode, when I came close to -6/-3db.
I first searched for the right sine-waves frequency to be used per bandpass.
Guessing that the maximum power would be measured at the maximum FR of each bandpass in dsp, including crossover and eq (but maybe I'm wrong here).
So I found some that are close enough to the middle of each bandpass with: 50Hz/160Hz/400Hz/1Khz & 8Khz.
And played them at -30/-20/-10db on the main dsp and measured each (I could have used only one level but wanted to confirm my calculations).
Using the rapidtables it gave me these needed power numbers to reach -0db:
- 1500w on sub, lol
- 90w on 2 mid-bass
- 35w on low mid
- 6w on high mid
- 11w on tweeter
It should not radically change except maybe on mid-bass that are wired in //.
Needless to say I use less than 1 watt most of the time near -20db, and less than 10w when loud.
1500 funny watts for the subs, I don't need that and I'm learning how to use the compressor anyway as the current 500w are more than enough, passed -10db, it already rattles everything around.
The midbass number is a surprise, as it's much lower than few months ago with your version of the test with 120Hz (it asked for 360w to reach -0db). But I didn't have the subs at that time, so I had more boost there.
Do you see anything wrong in the process? Thx!