Reminds me of someone who suggested that evacuating one side of a condensor microphone would lower the noise 3dB.
Only in that the power level for 90 db spl up close is too low for distortion tests. find for freq response in a room.
Whereas 1W may produce 85-90 db spl at 1 meter, you will need much higher power far away. How much more is shown here.
Simply multiply the power needed for the listening distance and use that power for THD measurements.
At 10-20 feet from spkr, >100W is needed So maybe test from 10-100W for distortion. That will separate the girls from the boys.
THx-RNMarsh
Speaker Power and Distance - PUI Audio | A Projects Unlimited Company located in Dayton, Ohio
Last edited:
One more time? ---
JBL does it right for the PRO's and for home. They give the standard 1W/8 at 1 meter because they have to. But then they give the spec that means a lot more; Continuous/Peak SPL@ 1m: 117dB /123dB; 108dB peak SPL @ 8m
That is a 15 db level loss at 8 meters. It would take 32 Watts to get the same SPL as 1W at 1 meter if listening 8 M away.
So, if you also were listening M2 at 8 M away, test for distortion with 32W input.
THx-RNMarsh
JBL does it right for the PRO's and for home. They give the standard 1W/8 at 1 meter because they have to. But then they give the spec that means a lot more; Continuous/Peak SPL@ 1m: 117dB /123dB; 108dB peak SPL @ 8m
That is a 15 db level loss at 8 meters. It would take 32 Watts to get the same SPL as 1W at 1 meter if listening 8 M away.
So, if you also were listening M2 at 8 M away, test for distortion with 32W input.
THx-RNMarsh
Last edited:
Free field SPL losses with distance aren't a good model for normal in-room use. Also, two speakers driven (assumed not correlated) add as +3dB.
Near-field measurements don't approach infinity either. Within a wavelength or so, SPL is about constant with distance. Easy to try for oneself. The PUI paper is caca.
All good fortune,
Chris
Near-field measurements don't approach infinity either. Within a wavelength or so, SPL is about constant with distance. Easy to try for oneself. The PUI paper is caca.
All good fortune,
Chris
Last edited:
Yes. Beranek, Acoustics of small rooms. Easy to Google to get full book. Anything smaller than a concert hall is a small room and free field SPL rules do not work.
Details. Depends on size of room when the field will flatten out. Not considering air atten or anything else, it is closer to reality than a misery 1W/1meter distortion test. Pick you own power level based on you measurments at the listening distance in your room -- further away than 1 m. And, choise of power based upon your speakers. 10-100W is a lot better test for distortion than a miserly 1W at 1 meter.
Thx-RNMarsh
Beranek signed his book to me himself; in person. Did you notice the "average" living room size that was used? Kinda small by USA standards. My listening room is simply huge and the inverse sq law works for a fairly good distance. 30 x 13 x 18 feet. Cant speak for everyone. Generally the formulas for reverb time in small rooms doesnt work from the formulas originally given for concert halls.
Yes, now we are going to get every possible exception to the rule except for the fact that 1W/1m is no where near reality in listening spaces of almost any size.
You can also measure the distance where the near and far fields meet in your room... I would always sit in near field.
Because of so many variables, pick a number between 10W and 100W to test drivers for their distortion.
THx-RNMarsh
The average UK home - including older and new-build properties is 85 sq m and has 5.2 rooms - with an average area of 16.3 sq m per room.Sep 14, 2011
When doing the sound reinforcement for a symphony I find I need 30 dB of headroom to avoid hearing clipping as mentioned before. For everything else 20 dB is fine. Some recordings have 20 dB of headroom. So that would leave a 1,000 watt amplifier running at an average level of 10 watts if full power was really being used. With inefficient loudspeakers this would be 92 dBa at 1 meter.
Thus the issues are, how efficient are one's loudspeakers, the level at which they listen, the perception of distortion vs time presented and the equivalent acoustic distance. In the near field sound level drops following the inverse square law, in the far field the reverberation keeps the level constant. At the critical distance the reverberant field equals the direct field. In a listening room 1,500 cubic feet or less it is very difficult to get a uniform reverberant field. The distortion perception due to clipping is clearly affected by frequency as there is a need for more energy to reproduce low frequencies in virtually all small room music reproduction systems.
For the types of loudspeakers I tend to use the distortion data for the individual drivers is available in a power level vs distortion format. It is not as common to see the data for a full system. Looking at the power spectrum of music and the efficiency of drivers makes it seem it is the low frequency distortion of most concern. As low frequencies tend to mask higher frequencies, most of that distortion of concern is masked in normal program material. The second distortion perception reduction occurs from multiple driver systems. While second and third harmonic distortion are less perceptable that higher order ones, driver bandwidth limits and crossover selection severely limit such high order transducer distortion creation.
So to properly measure perceived distortion a single power level is probably inappropriate. 100 watts at very low frequencies may not be unrealistic. It would however be problematic at the high end and dangerous both to the hearing of the person doing the testing and damaging to the transducer. Now doing a 100 watt test of a low frequency driver would want to include cone effects. I suspect that would limit the closest testing distance to perhaps at least a distance greater than three times the piston diameter and as much as ten times it to get within 1 dB accuracy. That should put the sound pressue level below 110 dB in most cases. A level that a decent measurement microphone can handle but of course with increased distortion in the microphone capsule. Pads may be used to keep the level comfortable for the electronics. Way back in this thread microphone distortion was addressed.
Thus the issues are, how efficient are one's loudspeakers, the level at which they listen, the perception of distortion vs time presented and the equivalent acoustic distance. In the near field sound level drops following the inverse square law, in the far field the reverberation keeps the level constant. At the critical distance the reverberant field equals the direct field. In a listening room 1,500 cubic feet or less it is very difficult to get a uniform reverberant field. The distortion perception due to clipping is clearly affected by frequency as there is a need for more energy to reproduce low frequencies in virtually all small room music reproduction systems.
For the types of loudspeakers I tend to use the distortion data for the individual drivers is available in a power level vs distortion format. It is not as common to see the data for a full system. Looking at the power spectrum of music and the efficiency of drivers makes it seem it is the low frequency distortion of most concern. As low frequencies tend to mask higher frequencies, most of that distortion of concern is masked in normal program material. The second distortion perception reduction occurs from multiple driver systems. While second and third harmonic distortion are less perceptable that higher order ones, driver bandwidth limits and crossover selection severely limit such high order transducer distortion creation.
So to properly measure perceived distortion a single power level is probably inappropriate. 100 watts at very low frequencies may not be unrealistic. It would however be problematic at the high end and dangerous both to the hearing of the person doing the testing and damaging to the transducer. Now doing a 100 watt test of a low frequency driver would want to include cone effects. I suspect that would limit the closest testing distance to perhaps at least a distance greater than three times the piston diameter and as much as ten times it to get within 1 dB accuracy. That should put the sound pressue level below 110 dB in most cases. A level that a decent measurement microphone can handle but of course with increased distortion in the microphone capsule. Pads may be used to keep the level comfortable for the electronics. Way back in this thread microphone distortion was addressed.
Last edited:
Yes! Which is what we should be doing also. A single power level only when you know the distance, effec , SPL desired, etc.
THx RNMarsh
Last edited:
Ideally you'd want to know the distortion at some defined power in, the SPL at that level (ie efficiency) and the relationship between power in and distortion. They you'd know the expected distortion at your typical listening level. And that for each driver in use, then preferably tri amp it.
The average UK home - including older and new-build properties is 85 sq m and has 5.2 rooms - with an average area of 16.3 sq m per room.Sep 14, 2011 I chose the house I live in now primarily because of its large music playback room size. I cannot seriously imagine listening in a room that the average is smaller than my home office room I am typing this from.
I would prefer high quality headphones in that case. I am not being a snot about this but... you do need some generous room size with decent acoustics for a high-end system. I am in the process of moving to Bangkok, Thailand in a couple weeks and one thing I will do is find a larger place for my music to play well in.
THx
-RNMarsh
I would prefer high quality headphones in that case. I am not being a snot about this but... you do need some generous room size with decent acoustics for a high-end system. I am in the process of moving to Bangkok, Thailand in a couple weeks and one thing I will do is find a larger place for my music to play well in.
THx
-RNMarsh
Last edited:
SPL vs. distance from sound source in a small room according to
ACOUSTICS OF SMALL ROOMS MENDEL KLEINER JIRI TICHY © 2014 by Taylor & Francis Group, LLC
And what is a "small room"? Defined here:
Small Room Acoustics
So, 85m^2 area is a small room as well.
ACOUSTICS OF SMALL ROOMS MENDEL KLEINER JIRI TICHY © 2014 by Taylor & Francis Group, LLC
And what is a "small room"? Defined here:
Small Room Acoustics
So, 85m^2 area is a small room as well.
Attachments
Last edited:
Most definitely, as I said also just before you #28754. In fact, too small.. IMO. Especially below 250 Hz. In which case about 16 Watts average would be needed from driver. So, if a single power level was to be chosen for these small listening rooms, 15-20W would be a good number for driver distortion tests.
THx
-RNMarsh
*) it is the overall perception of low frequency in relation to a specific room response which is of the highest value in assessing the quality of reproduction. *https://pdfs.semanticscholar.org/6ce9/30929da97c98af9f5a65cadbd6e223785ea6.pdf
Last edited:
Thanks for showing this.
The difference between 20 cm and 1 meter is 14dB. An EarthWorks MC30 has no distortion up to 140dB. In other words, I can do near field distortion measurements that equate to 126dB @ 1 meter. Good enough for government use.
Mike was not a problem in my distortion measurements, speaker distortion was much higher, tested. No need for patronizing, RNM. Rather show some of your results.
As I showed an AES paper that the distortion of cap microphones at freq below 100Hz can be very high. And, we are not talking about at max levels. it is the mic capsule to amplifier interface which does not have high enough input Z in typical cases. So, that has to be determined on your mic system.
The measurement of lowered distortion using current feedback at low freq (1985) was with a 3/4 inch omni dynamic mic. BTW. Low enough distortion to see improvement with the current mode technique. I dont plan on doing the distortion tests again. Though I own several cap mics now which are excellent for doing it.
THx-RNMarsh
The measurement of lowered distortion using current feedback at low freq (1985) was with a 3/4 inch omni dynamic mic. BTW. Low enough distortion to see improvement with the current mode technique. I dont plan on doing the distortion tests again. Though I own several cap mics now which are excellent for doing it.
THx-RNMarsh
Last edited:
Totally bummed out.
The visiton bg20-8 has a cone that curves. I thought I could Hotwire a simple cone to hold it locked. Now I have to buy a higher gauge nichrome that can hold a curve, and put a few turns on the power toroid to push current to heat the wire.
Jn
The visiton bg20-8 has a cone that curves. I thought I could Hotwire a simple cone to hold it locked. Now I have to buy a higher gauge nichrome that can hold a curve, and put a few turns on the power toroid to push current to heat the wire.
Jn
Better full range cones are curvilinear. They are stiffer and have better hf dispersion.
Maybe you cal line it with plastic and use plaster to make a tight fit? Seems messy. . .
Maybe you cal line it with plastic and use plaster to make a tight fit? Seems messy. . .
Actually, not a bad idea, I like it.
I was planning on measuring both suspension compliance and BLi vs displacement using a scale to compare displacement force against driver weight, I fear plaster might exceed scale full range..will investigate.
I was worried the styrofoam might not be stiff enough, plaster certainly would. However, having experience fixing plaster walls, there will certainly be alcohol involved.... Not imbibing, although that sounds good..but to slow the hardening.
Jn
I was planning on measuring both suspension compliance and BLi vs displacement using a scale to compare displacement force against driver weight, I fear plaster might exceed scale full range..will investigate.
I was worried the styrofoam might not be stiff enough, plaster certainly would. However, having experience fixing plaster walls, there will certainly be alcohol involved.... Not imbibing, although that sounds good..but to slow the hardening.
Jn
Last edited:
Why not use Great Stuff spray foam from the local hardware store? Expands to conform. Coat the cone in mold release or use a plastic mask.
I have started entering the parameters for the BG20-8 in my model. The inductance measurement allows me not to guess on some important parameters. To not guess on the shorting ring parameters, I then need to know the number of turns in the coil (or the inductance when removed from the motor).
Luckily the specsheet gives the plate thickness and coil length, which makes it even easier.
The LF inductance helps me to model the reluctance force which tries to pull the cone inward regardless of polarity as the coil is magnetizing the motor.
The number of turns helps me to model the eddy currents which try to push the cone away regardless of polarity as the eddy currents are demagnetizing the coil. In the current model the eddy currents definitely seem to be winning over the reluctance force...
The inductance of the coil in free air helps me to model the permeability of the motor geometry.
Am I right that eddy currents mainly push on the coil radially from the pole piece outwards, and only a fraction of that force actually pushes the cone out? Because in my model the force is quite significant.
I have started entering the parameters for the BG20-8 in my model. The inductance measurement allows me not to guess on some important parameters. To not guess on the shorting ring parameters, I then need to know the number of turns in the coil (or the inductance when removed from the motor).
Luckily the specsheet gives the plate thickness and coil length, which makes it even easier.
The LF inductance helps me to model the reluctance force which tries to pull the cone inward regardless of polarity as the coil is magnetizing the motor.
The number of turns helps me to model the eddy currents which try to push the cone away regardless of polarity as the eddy currents are demagnetizing the coil. In the current model the eddy currents definitely seem to be winning over the reluctance force...
The inductance of the coil in free air helps me to model the permeability of the motor geometry.
Am I right that eddy currents mainly push on the coil radially from the pole piece outwards, and only a fraction of that force actually pushes the cone out? Because in my model the force is quite significant.
- Status
- Not open for further replies.