I thought I was the only one to use this trick to "break-in" speakers.
Indeed. While some speakers produce a superb null across most of the band, there are some where you get audible "hot spots." And with some pathological examples, this unmasks cabinet panel resonances.
John Atkinson
Editor, Stereophile
Placing the speakers face-to-face closely and running them out of phase is a neat idea, but how well does this work for speakers with sloped baffles where the distance between the baffles may become substantial as we move from bottom to top?
Cheers,
Bob
BoB,
Well, depending how far slope, it matters not a lot, IF they match well and the positioning is precise and surfaces in the room are a lot more distanct than the distance between the speakers. AMR's Speakers slope a bit and drivers are time aligned (waveguide on HF)...
I find the null in output surprisingly tolerant of minor position errors but strongly intolerant to any errors in FR. I guess the constant directivity design helps...
Ciao T
Well, depending how far slope, it matters not a lot, IF they match well and the positioning is precise and surfaces in the room are a lot more distanct than the distance between the speakers. AMR's Speakers slope a bit and drivers are time aligned (waveguide on HF)...
I find the null in output surprisingly tolerant of minor position errors but strongly intolerant to any errors in FR. I guess the constant directivity design helps...
Ciao T
The back rooms I have visited used a kryton tube to discharge into needle like electrodes of an undisclosed very wear resistant material. Calibration was using a B&K 1/4" capsule. Both places were set up by the same expert who charges even more than I do and that was 20+ years ago!
The last time I bought measurement microphones I got a quantity price on half a dozen 1/4" complete units. The idea is that microphones are sometimes used as three pairs. Jim left my employe with 2 of them that he actually paid for. Chuck stole one as did Ross, while we were cleaning up from the flood. Dave left with a cheapie he stole from a client! I dropped one while rigging it in a stadium, landed right on the capsule with the mic cable to guide it! So I have one left. I was planning on buying a new set. Then I looked at the price!!!
In addition to those I have a classic B&K 1" an ACO 1/2" and two GR 1".
I of course also have a pistonphone with the calibrated in db barometer. I got an extra but I gave that to Charlie for a favor. I have several calibrators and SLM's (2 B&K, 2 GR, and 2-3 Off brand) but I think Dave also left with one of each (GR). The GR's are loaners, one NFL football client used a set to keep another client honest!
(You know I might just look at getting some webcams!)
Oh yeah I also have a fairly recent TEF system in addition to the AP system 2, Smaart rig and Zoom stereo recorder.
Then I have the test equipment we don't talk about! It is handy to measure things my competitors don't know about!!!
So that is what I consider basic equipment to make acoustic measurements. But even with all that I will be adding gear next year!
Then I occasionally get inquiries to use my facilities, but we really aren't interested in that.
So Scott if you need to borrow something to do a calibration don't be afraid to ask.
Yeah and you make the amateur's mistakes as well!
Time aligned is a really big issue when you build big arrays. The "Acoustic Center" moves with frequency and to a lesser extent level. To make things worse there are several different definitions.
Some folks consider the center to be at the voice coil. I ask them if I connect the voice coil to a mass-less rigid tube and then connect the tube to the piston does this change the point of origin?
Then there is the issue of a standard cone. It is a cone! When you move the back of the cone the front moves the same distance almost instantly. The air pushed by the back of the cone moves slower than that. So the air from the front of the cone arrives a little bit ahead of the dust cap air. For combinatorial calculations the barycenter is a better choice.
Then we can talk about horn loads. At low frequencies the air movement conforms to the horn shape and behaves as if the outer edge of the horn where maximum pressure change occurs is the combinatorial center. At higher frequencies the center moves back as the air motion lifts earlier.
Now for the case of a single full range loudspeaker, these issues are generally most important at crossover frequencies between drivers. In a multiple loudspeaker system you have many drivers covering the same frequency range but different seating areas. The combinations of adjacent drivers can narrow the coverage patterns, so the ability to predict that is a wee bit important. The we get to worry about how drivers in different frequency ranges combine!
My last post to the idea of current drive of a dynamic speaker. 1st plot shows impedance plot vs. frequency calculated from the equivalent diagram shown. The 2nd plot shows transfer function of this equivalent speaker circuit with increasing voice coil resistance as a parameter (higher resistance = closer to current drive). We can see that when increasing voice coil resistance, i.e. approaching to current drive condition, the only thing we get is a resonance peak at bass frequencies and decreasing efficiency. Exactly as expected.
Attachments
Did you consider that some drivers are more suitable for current drive than others? I can imagine taking a driver with a Qts of 0.7 (perfect for an infinite baffle), showing what happens when you put it in a vented box, then triumphantly claiming that, see?, vented boxes are no good!
Mills and Hawksford worked this current drive stuff out in pretty good detail. This was expanded on in a recent book which caused some lively discussion here. Nelson Pass wrote a very accessible "dummy's" guide, and right in our own Articles section, Morgan Jones demonstrated the utility of matching the right speaker with an amp having a high-ish source impedance.
Mills and Hawksford worked this current drive stuff out in pretty good detail. This was expanded on in a recent book which caused some lively discussion here. Nelson Pass wrote a very accessible "dummy's" guide, and right in our own Articles section, Morgan Jones demonstrated the utility of matching the right speaker with an amp having a high-ish source impedance.
He's adding a physical resistance rather than creating a "virtual" one in the source impedance of the driving amp. The efficiency argument is, IMO, not totally honest.
Ed,
It is one even in the fairly small arrays we used two decades ago.
That depends on he drivers and various of their parameters...
When AMR's Speaker came out I had a whitepaper that covered everything I considered really relevant to say about why it is what it is. It came to around 30 Pages and covered less than halve I wanted. Marketing said "too long" and condensed it to one page...
Not quite, BUT it depends on driver construction. In the Eckmiller driver it is at the crossover frequency very close to the voice coil for both drivers. In the TH315 it needs LC Delay lines to make it work. In AMR's speaker I measured the acoustic centre with pulses...
Hopefully not. Cones are nice for IScream...
My ideal cone driver (for bass or bass low mid in a multiway system) has a shallow NAWI cone and non-spherical section dustcap that are mirror symmetrical (kind of like a big ring radiator) with equal distance (but not surface) from centre to voice coil and from voice coil to edge of cone. Cone termination is still a major PITA, but with a 4" Voice Coil for an 8" cone it is possible to push the trouble out of band. The acoustic centre in this case is reasonably constant as the whole driver is easier to make behave inunison.
But the barycentre moves with frequency for most drivers. However, there is a saving grace. If the driver does not work as ideal piston parts of the diaphragm decouple. And long wavelength have more "tolerance" than short ones...
Never really encountered any, outside musical instruments.
I have found the view of "horn" of little use. I have found the "waveguide" concept more in line with observations I have made, but I notice that badly designed waveguides show "bad" behaviour normally considered to be a functional characteristic of "horns".
So we are not talking about a single full range loudspeaker, but a multiple loudspeaker, multiway system.
Ed, I used to work with guys that your kind of installs in the 80's and sat at mixing desks for bands. I know that. But the concepts and challenges do not directly translate and scale to domestic systems.
If you have the time, look up the Eckmiller Driver. It gives you an idea of my background in serious audio. It was like really ancient news when I was wet behind the ears...
In the 80's we used column speaker designs for PA (as in public address, not sound reinforcement) with a very defined directivity across the whole usable bandwidth (< 100Hz) and as we could not really deal well with the delays (multiple output digital delays where still years in the future for generic public installs and what existed was not available in the eastern block) we had to use things very differently...
Anyway, here we are not normally talking about massive stadium or hall systems, but home audio... So some problems disappear, others raise their ugly head...
Ciao T
It is one even in the fairly small arrays we used two decades ago.
That depends on he drivers and various of their parameters...
When AMR's Speaker came out I had a whitepaper that covered everything I considered really relevant to say about why it is what it is. It came to around 30 Pages and covered less than halve I wanted. Marketing said "too long" and condensed it to one page...
Not quite, BUT it depends on driver construction. In the Eckmiller driver it is at the crossover frequency very close to the voice coil for both drivers. In the TH315 it needs LC Delay lines to make it work. In AMR's speaker I measured the acoustic centre with pulses...
Hopefully not. Cones are nice for IScream...
My ideal cone driver (for bass or bass low mid in a multiway system) has a shallow NAWI cone and non-spherical section dustcap that are mirror symmetrical (kind of like a big ring radiator) with equal distance (but not surface) from centre to voice coil and from voice coil to edge of cone. Cone termination is still a major PITA, but with a 4" Voice Coil for an 8" cone it is possible to push the trouble out of band. The acoustic centre in this case is reasonably constant as the whole driver is easier to make behave inunison.
But the barycentre moves with frequency for most drivers. However, there is a saving grace. If the driver does not work as ideal piston parts of the diaphragm decouple. And long wavelength have more "tolerance" than short ones...
Never really encountered any, outside musical instruments.
I have found the view of "horn" of little use. I have found the "waveguide" concept more in line with observations I have made, but I notice that badly designed waveguides show "bad" behaviour normally considered to be a functional characteristic of "horns".
So we are not talking about a single full range loudspeaker, but a multiple loudspeaker, multiway system.
Ed, I used to work with guys that your kind of installs in the 80's and sat at mixing desks for bands. I know that. But the concepts and challenges do not directly translate and scale to domestic systems.
If you have the time, look up the Eckmiller Driver. It gives you an idea of my background in serious audio. It was like really ancient news when I was wet behind the ears...
In the 80's we used column speaker designs for PA (as in public address, not sound reinforcement) with a very defined directivity across the whole usable bandwidth (< 100Hz) and as we could not really deal well with the delays (multiple output digital delays where still years in the future for generic public installs and what existed was not available in the eastern block) we had to use things very differently...
Anyway, here we are not normally talking about massive stadium or hall systems, but home audio... So some problems disappear, others raise their ugly head...
Ciao T
I already said I have access to traceable standards, afterall those cell phone mics need .2dB calibration.
Maybe you forgot, I'm trying to help DIY'ers gratis (my wife figures Jan is paying me about $1 an hour for my time). There is also the challenge factor how good can you do on $50 worth of scavenged parts. I already know you can equal an AP with a $200 sound card, free software, and some cleverness but the plots don't say AP in the corner. Sorry Bruce ignore that statement it's merely rhetorical.
It does not, actually... Our office went to Circus, but I could not go... I took already too much of Excedrin. Does not help.
Edit: sorry for off-topic... It hurts...
Last edited:
John it's not about the money. It's all about the money thats why more was invented and get it up front like you said.
- Status
- Not open for further replies.