IMHO wide range drive units have too much doppler distortion, and it is audible. No wide range unit is exempt, regardless of its inherent quality, even Manger and electrostatics: the doppler distortion mucks it up.
The tradeoffs in a multi-drive-unit solution are point source problems and crossover problems. I prefer to work with these because at least they can be minimised, unlike doppler distortion.
Crossover problems used to be chronic, but with modern digital crossover units they are much reduced. Recommended.
Point source problems, especially with waveguides, need care. The Unity waveguide was a very well-executed coaxial solution, although it needed 4 expensive midrange drive units per channel, and they were subject to excessive compression for a standard cone drive unit. I suggest, with a little care, a well-executed overlapping layout will work well, with the tweeter waveguide and supertweeter placed in front of, and overlapping, the midrange waveguide. I draw inspiration from Linn 3-K Driver Array, see www.linn.co.uk.
Regarding other factors I am in agreement with Mr Loesch: high sensitivity units for low distortion at all frequencies, with conical waveguides for controlled dispersion above 200Hz.
Grant
nowater said:
I can't comment on the doppler effect - I'm not sure what I'd be listening for. However, th Manager 103 (Active) was the finest speaker I've EVER listened to. I was in total awe the whole time. I actually lived with the 109 for a few months, and it became a part of me. I would suggest Manger over anything else suggested - and if you want an alternative, a line array of Bandor drivers or actual ESLs are the ony way to go.
Just my opinion of course
Gaz
Re: Re: Re: ultimate DIY speakers - Ideas?
Konnichiwa,
Sorry, not when I had the chance to hear TACT's own Demo's. I'm not saying it sounded bad, but it certainly sounded nothing like a dipole. The key for the dipoles behaviour (at LF) is IMHO is actually in the RT60 and the direct/reflected sound ratio, which TACT (or any other equaliser) can do absolutely nothing about.
Sayonara
Konnichiwa,
amo said:
Sorry, not when I had the chance to hear TACT's own Demo's. I'm not saying it sounded bad, but it certainly sounded nothing like a dipole. The key for the dipoles behaviour (at LF) is IMHO is actually in the RT60 and the direct/reflected sound ratio, which TACT (or any other equaliser) can do absolutely nothing about.
Sayonara
Re: Re: ultimate DIY speakers - Ideas?
Why? What's wrong with a driver with 0.1-0.3% K2 and K3 distortion within the operating range at listening levels when its sensitivity is well below 90dB? The only drawback I see is that you need bigger amp, which should be the only reason why a driver with higher sensitivity is prefered. Right?Kuei Yang Wang said:
Konnichiwa,
No. The TACT RCS (based on what I read on the last releases) still can only correct reflection of less than 1/2 of the wavelength "distance". That is the same any equaliser can corrects and hence the claims of timedomain corrections are a little, shall we say, optimisitic. The correction seems frequency/amplitude domain based with the atendant effects in the timedomain.
What this means is that the RT60 is UNAFFECTED. To reduce the RT60 the system would have to actually calaculate and offset a very large number of reflections, which poses at the very least significant problems. What I know about the TACT RCS at this stage, based on their own technical papers (which seem pulled from the TACT website!!??) suggests that they do not do this. Lst time I looked the RCS had about twice the processing power of Behringers Digital EQ, not enough IMHO to do anything so complex as would be needed.
If the RCS has seen upgrades since I last looked that invalidate the above, fine. Then someone please tell me and show evidence.
BTW, note that I'm not "bashing" TACT, on the contrary, I have been repeatedly on record as saying that conventional speakers require mandatory LF equalisation in order to operate correctly in acostically small rooms. I merely register some disbelieve if such boxes are being ascribed effects that they are unlikely to achieve in reality.
I don't think so. From measurements we did it appears a dipole with a flat LF response (EQ'ed or not) in a moderatly large room has an RT60 that does NOT rise at low frequencies by anywhere the amount a Box Speaker/Room System causes it to rise. Measuring myself other equalised Speaker systems (non TACT) I found the RT60 rise to remain substantially the same with or without correction.
While not conclusive (too few measurements and too few room correction systems to be certain) this raises a number of interesting points. One might argue that a well designed and digitally corrected dipole is best.
Sayonara
Pan said:
No. The TACT RCS (based on what I read on the last releases) still can only correct reflection of less than 1/2 of the wavelength "distance". That is the same any equaliser can corrects and hence the claims of timedomain corrections are a little, shall we say, optimisitic. The correction seems frequency/amplitude domain based with the atendant effects in the timedomain.
What this means is that the RT60 is UNAFFECTED. To reduce the RT60 the system would have to actually calaculate and offset a very large number of reflections, which poses at the very least significant problems. What I know about the TACT RCS at this stage, based on their own technical papers (which seem pulled from the TACT website!!??) suggests that they do not do this. Lst time I looked the RCS had about twice the processing power of Behringers Digital EQ, not enough IMHO to do anything so complex as would be needed.
If the RCS has seen upgrades since I last looked that invalidate the above, fine. Then someone please tell me and show evidence.
BTW, note that I'm not "bashing" TACT, on the contrary, I have been repeatedly on record as saying that conventional speakers require mandatory LF equalisation in order to operate correctly in acostically small rooms. I merely register some disbelieve if such boxes are being ascribed effects that they are unlikely to achieve in reality.
Pan said:
I don't think so. From measurements we did it appears a dipole with a flat LF response (EQ'ed or not) in a moderatly large room has an RT60 that does NOT rise at low frequencies by anywhere the amount a Box Speaker/Room System causes it to rise. Measuring myself other equalised Speaker systems (non TACT) I found the RT60 rise to remain substantially the same with or without correction.
While not conclusive (too few measurements and too few room correction systems to be certain) this raises a number of interesting points. One might argue that a well designed and digitally corrected dipole is best.
Sayonara
Re: Re: Re: ultimate DIY speakers - Ideas?
Konnichiwa,
Nothing. Except it will not have as low distortion, in virtually all cases. Let's define "listening levels". I define them as "realistic levels". Elsewhere the SPL in the Front Row(s) during an acoustic classical performance was discussed, this suggests an average SPL of >>90db and a peak SPL of >> 105db at the lsitening position.
If you do not listen in the extreme nearfield but (for arguments sake) at 4m listening distance in a moderatly well damepd room then 105db peaks at the listening position equate theoretically to 117db at 1m. If we allow a 2db rise in SPL due to the resonant field and 6db due to the use of Stereo (in reality rarely achieved) we still require a per speaker SPL of 109db/1m. I somehow doubt that many speakers with << 90db/1W/1m sensitivity can reproduce 109db/1m with anything approaching 0.1-0.3% K2 and K3.
Wrong. See above.
Any Speaker Driver that requires significantly over 1 Watt for the required average SPL at the listening position (I would suggest 96db average with 10db crestfactor as realistic but lacking headroom) will be significantly distorted and compressed UNLESS extreme measures are taken to reduce Distortion and Compression, which usually are NOT applied to "high fidelity" drivers, thus making any claims of "Fidelity" suspect.
Of course, maybe your listening levels are way below realistic levels, that's fine with me, but has (obviously) nothing to do with "High Fidelity" again, as the actual dynmics of the performance are a strong factor that allows to delineate recording and original.
Sayonara
Konnichiwa,
Hans L said:
Hans L said:
Nothing. Except it will not have as low distortion, in virtually all cases. Let's define "listening levels". I define them as "realistic levels". Elsewhere the SPL in the Front Row(s) during an acoustic classical performance was discussed, this suggests an average SPL of >>90db and a peak SPL of >> 105db at the lsitening position.
If you do not listen in the extreme nearfield but (for arguments sake) at 4m listening distance in a moderatly well damepd room then 105db peaks at the listening position equate theoretically to 117db at 1m. If we allow a 2db rise in SPL due to the resonant field and 6db due to the use of Stereo (in reality rarely achieved) we still require a per speaker SPL of 109db/1m. I somehow doubt that many speakers with << 90db/1W/1m sensitivity can reproduce 109db/1m with anything approaching 0.1-0.3% K2 and K3.
Hans L said:
Wrong. See above.
Any Speaker Driver that requires significantly over 1 Watt for the required average SPL at the listening position (I would suggest 96db average with 10db crestfactor as realistic but lacking headroom) will be significantly distorted and compressed UNLESS extreme measures are taken to reduce Distortion and Compression, which usually are NOT applied to "high fidelity" drivers, thus making any claims of "Fidelity" suspect.
Of course, maybe your listening levels are way below realistic levels, that's fine with me, but has (obviously) nothing to do with "High Fidelity" again, as the actual dynmics of the performance are a strong factor that allows to delineate recording and original.
Sayonara
Ultimate speaker? <evil grin>
Following Linkwitz's advice: dipole bass from 40 up to say 100Hz with dual 15" TCSound (for example the Stryke HE15). And a pair of W26 for the range between 100 and say 350Hz.
Depending on the room(gain), add monopole subs for <40Hz: another pair of 15" TCSound in two seperate housings per side. Seperate housings gives the posiibility of optimizing the roommodes at listeningposition.
Although I have no experience with waveguides whatsoever, it seems a near ideal way to obtain a controlled, high and gradual directivity on and off axis. What to put into the monopole mid and tweet: how about a C44/8 and a C23/6 crossed at 3,5KHz. No wait, that ain't cost-no-object. Get the diamand Accuton instead and cross a little higher (4.5KHz).
Make it active with multiple digital (flexible, high cpu capacity) filters and a bunch of your favorite amps.
No where did I put that lottery ticket.
Following Linkwitz's advice: dipole bass from 40 up to say 100Hz with dual 15" TCSound (for example the Stryke HE15). And a pair of W26 for the range between 100 and say 350Hz.
Depending on the room(gain), add monopole subs for <40Hz: another pair of 15" TCSound in two seperate housings per side. Seperate housings gives the posiibility of optimizing the roommodes at listeningposition.
Although I have no experience with waveguides whatsoever, it seems a near ideal way to obtain a controlled, high and gradual directivity on and off axis. What to put into the monopole mid and tweet: how about a C44/8 and a C23/6 crossed at 3,5KHz. No wait, that ain't cost-no-object. Get the diamand Accuton instead and cross a little higher (4.5KHz).
Make it active with multiple digital (flexible, high cpu capacity) filters and a bunch of your favorite amps.
No where did I put that lottery ticket.
KYW,
have not done such measuremetns myself but realise dipoles have lower RT60 than a normal boxed speaker in the low range.
I assumed (but where wrong obviously) that the TacT did eat "all" errors in the time domain in low frequencies, meaning the RT60 would be shortened seriously.
Problem still excists if listeing in a node... the TacT does not amplify to try to fill in a null right? (or maybe it does.... by correcting in the time domain?). With this resoning as base it seems like a dipole and/or (active) bassabsorbers has its place even in a rig containing a RCS unit.
/Peter
have not done such measuremetns myself but realise dipoles have lower RT60 than a normal boxed speaker in the low range.
I assumed (but where wrong obviously) that the TacT did eat "all" errors in the time domain in low frequencies, meaning the RT60 would be shortened seriously.
Problem still excists if listeing in a node... the TacT does not amplify to try to fill in a null right? (or maybe it does.... by correcting in the time domain?). With this resoning as base it seems like a dipole and/or (active) bassabsorbers has its place even in a rig containing a RCS unit.
/Peter
Re: Re: Re: Re: ultimate DIY speakers - Ideas?
AFAIK, high sensitivity says nothing about maximum SPL or the level of distortion. If there is a relationship, I'd like to learn about it.
Oh and my listening level is bigger than yours
I read you answer carefully but failed to find any argument why apart from you stating that 'extreme measures' against distortion aren't used in "high fidelty" units...Kuei Yang Wang said:
AFAIK, high sensitivity says nothing about maximum SPL or the level of distortion. If there is a relationship, I'd like to learn about it.
Oh and my listening level is bigger than yours
Re: Re: Re: Re: Re: ultimate DIY speakers - Ideas?
Konnichiwa,
Hmmm. My argument is simple.
If the speaker has 0.1-0.3% Distortion at listening levels (implied to be 109db/1m for each speaker of the stereo pair) than that's a fine speaker. However, based on experience, most speakers struggle severely to produce such SPL's at such low distortion.
The lower the sensitivity, the more they strugle, as we have two fundamental mechanisms causing distortion, one of them dierectly related to applied electrical power (and cubic in nature). Thus, once you pass the point of around 1 Watt applied to a drive unit your eddy current induced distortions go up quite fast and steeply, unless serious measures are taken to combat this. These measured usually result in lowering sensitivity too, thus reducing the sensitivity and requiring more power for a given SPL again raising distortion and thus giving away some of what has been achieved.
Equally, the 1 Watt electrical power level is in many cases that at which we can observe thermal compression in drive units. Thus, the more SPL is produced for a given power level (all else being equal of course) the higher the SPL for a given distortion limit, in other words, higher sensitivity (all else being equal) means lower distortion.
I am unaware of any widely used "HiFi" Electrodynamic Transducer that implements sufficient measures to minimise the Distortion mencanisms and thermal compression to get anywhere near 109db/1m with distortion levels as low as you suggest.
Of course, a number of technologies exist (eg. ESL) that at least in theory can manage this, however they are outside the scope of my analysis, as they are not easy to DIY and commercially more or less unavailable.
Correct.
Incorrect, due to the majority of the inherent distortion mechanisms in electrodynamic speakers. Please consider the recent "Distortion in Speakers" thread for starters.
Then you need >> 100db/1Watt Speakers or you need to like distortion a lot! ;-)
Sayonara
Konnichiwa,
Hans L said:
Hmmm. My argument is simple.
If the speaker has 0.1-0.3% Distortion at listening levels (implied to be 109db/1m for each speaker of the stereo pair) than that's a fine speaker. However, based on experience, most speakers struggle severely to produce such SPL's at such low distortion.
The lower the sensitivity, the more they strugle, as we have two fundamental mechanisms causing distortion, one of them dierectly related to applied electrical power (and cubic in nature). Thus, once you pass the point of around 1 Watt applied to a drive unit your eddy current induced distortions go up quite fast and steeply, unless serious measures are taken to combat this. These measured usually result in lowering sensitivity too, thus reducing the sensitivity and requiring more power for a given SPL again raising distortion and thus giving away some of what has been achieved.
Equally, the 1 Watt electrical power level is in many cases that at which we can observe thermal compression in drive units. Thus, the more SPL is produced for a given power level (all else being equal of course) the higher the SPL for a given distortion limit, in other words, higher sensitivity (all else being equal) means lower distortion.
I am unaware of any widely used "HiFi" Electrodynamic Transducer that implements sufficient measures to minimise the Distortion mencanisms and thermal compression to get anywhere near 109db/1m with distortion levels as low as you suggest.
Of course, a number of technologies exist (eg. ESL) that at least in theory can manage this, however they are outside the scope of my analysis, as they are not easy to DIY and commercially more or less unavailable.
Hans L said:
Correct.
Hans L said:
Incorrect, due to the majority of the inherent distortion mechanisms in electrodynamic speakers. Please consider the recent "Distortion in Speakers" thread for starters.
Hans L said:Oh and my listening level is bigger than yours
Then you need >> 100db/1Watt Speakers or you need to like distortion a lot! ;-)
Sayonara
My understanding of drive units is that any one unit possesses a strict relationship between amplitude and harmonic distortion. More amplitude => more distortion.
So it is more informative to compare drive unit distortion at a given (high) SPL level than at a given input voltage.
For example, say 105dB SPL at 1m. An 85dB/w-m sensitivity unit may distort a low 0.3% at 2.83V input, but that is only 85dB SPL. Crank it to 105dB and you may have 10% distortion. But a 105dB/w-m unit with "more" distortion of 1% at 2.83V input, will at 105dB SPL produce only 1% distortion.
Hence the critical advantage of the high sensitivity drive unit, provided your design allows for its inherently narrower bandwidth. Probably 4-way at a minimum if your design goal is not to compromise.
Re: Re: Re: Re: Re: Re: ultimate DIY speakers - Ideas?
Konnichiwa,
Here some links to related discussions and background info, which may illustrate and clarify the distrotion menchanisms and help to a guestimate as to their quantity.
http://www.diyaudio.com/forums/showthread.php?threadid=37541
http://www.google.co.uk/search?q=cache:9eJeI9ZBSOUJ:f18.parsimony.net/forum31999/messages/64114.htm+Ragnar+lien+loudspeaker+Distortion&hl=en
http://db.audioasylum.com/cgi/m.mpl?forum=tweaks&n=50278
http://www.soundonsound.com/sos/Jun02/articles/monitors.asp
http://www.soundonsound.com/sos/Jul02/articles/monitors2.asp
http://www.listeninc.com/files/pdf/B&K%20Audio%20Distortion%20Measurements.pdf
I hope that helps for starters....
Sayonara
Konnichiwa,
Here some links to related discussions and background info, which may illustrate and clarify the distrotion menchanisms and help to a guestimate as to their quantity.
http://www.diyaudio.com/forums/showthread.php?threadid=37541
http://www.google.co.uk/search?q=cache:9eJeI9ZBSOUJ:f18.parsimony.net/forum31999/messages/64114.htm+Ragnar+lien+loudspeaker+Distortion&hl=en
http://db.audioasylum.com/cgi/m.mpl?forum=tweaks&n=50278
http://www.soundonsound.com/sos/Jun02/articles/monitors.asp
http://www.soundonsound.com/sos/Jul02/articles/monitors2.asp
http://www.listeninc.com/files/pdf/B&K%20Audio%20Distortion%20Measurements.pdf
I hope that helps for starters....
Sayonara
Re: Re: Re: Re: Re: Re: ultimate DIY speakers - Ideas?
And there are a ton of hifi / highend drivers that have (multiple) faraday rings and various measures against compression. Very low distortion drivers with 'low' sensitivity compared to the 'average' pro driver can be found at: Seas Excel, Audio Technology, TCSound, Accuton and ATC.
I don't think I need to comment on their quality.
Summarizing: there are mechanisms that come into play when one needs more than 1 or 2 watts: eddy currents and heat compression. High excursion and non pistonic behaviour are also major sources of distortion, but these have nothing to do with the sensitivity, only freq, required spl, Sd and cone material/construction. Combine this with the fact that sensitivity doesn't come for free and these questions remain: how important are the first two sources compared to the others? And at what price (meaning dollars) do the benifits you attribute to pro drivers come?
'My listening levels are bigger than yours' was a joke... Of course I don't have a speaker with that kind of distortion at 109db and neither have you. No patronizing ppl who ask questionsKuei Yang Wang said:
Now we're talking
Exactly! So a lower sensitivity can indeed go together with a high quality unit. Let's not be vague on these (extreme) measures: a Faraday ring is such a measure and is readily found in many hifi / highend speakerunits.
If you are aware of 'pro' units that do, I'd like to know which ones. I do know units that specifically attacks both sources of distortion you descibe above. The Seas M15 comes to mind right away because of the design and construction of its engine. If you're not into pp cones there is a magnesium version as well (with 3-4db less sensitivity). Surely this must be one of the best midranges available anywhere.
And there are a ton of hifi / highend drivers that have (multiple) faraday rings and various measures against compression. Very low distortion drivers with 'low' sensitivity compared to the 'average' pro driver can be found at: Seas Excel, Audio Technology, TCSound, Accuton and ATC.
I don't think I need to comment on their quality.
I read a few bits and pieces of the links you provided. It's a shame there is so much noise in these threads
Summarizing: there are mechanisms that come into play when one needs more than 1 or 2 watts: eddy currents and heat compression. High excursion and non pistonic behaviour are also major sources of distortion, but these have nothing to do with the sensitivity, only freq, required spl, Sd and cone material/construction. Combine this with the fact that sensitivity doesn't come for free and these questions remain: how important are the first two sources compared to the others? And at what price (meaning dollars) do the benifits you attribute to pro drivers come?
Re: Re: Re: Re: Re: Re: Re: ultimate DIY speakers - Ideas?
Konnichiwa,
Depends on your definition of "high quality". In on of the the references I gave three different 6.5" Drivers in commercial 2-Way speakers where evaluated. The "best" unit was Dynaudio with a little under 3db compression at 50W compared to 0.5 Watt, the worst was a B& custom Job from the DM303 with almost 5db compression. Midband distortion levels for all of these speakers/drivers will exceed 0.5% @ 96db/1W/1m.
The Focal unit in the KRK Monitor and the Dynaudio unit in the Dynaudio Speaker would be classed by most people as being "high quality". Based on measured performance they are "clearly incapable of high fidelity reproduction".
So, no, no matter how you slice it, a low sensitivity drive unit CANNOT be high quality.
A farraday ring is usually applied at a place other than the actual pole piece. In order to be truely effective you must have a significant copper sleeve on BOTH pole pieces IN THE AIRGAP. In addition you must choose a magnet geometry that evens out the magent field with coil travel.
The commonly found magnet systems with simple flat outer pole pieces, the inner pole piece flush with upper surface of the outer pole piece and a farraday ring on the inside of the ceramic magnet provide a very poor motor system that inherently remains high distortion, even though the faraday ring does a little to reduce distortion, it mostly does so for 2nd harmonics.
The problem in the midrange is invariably 3rd harmonics from the cubic function of the eddy current distortion and can only be addressed by either making the magnet pole pieces materially non-conducting (ATC) or fitting significant copper sleeves, both of which requires a significant increase in magnet size to at least partially compensate the loss of efficiency.
So no, if I am saying extreme measures I am thinking of something a little more involved (and expensive) than a farraday ring, which is mostly good for advertising it's presence. I would think either the JBL symmetric field geometry magent system or that developed and used by Lambda on their better woofers.
It still leaves other issues unaddressed which would require the use ideally of field coil magnets or "super strength" neodymium magnets.
Serious Pro-Units usually incorporate much more sophisticated measures of distortion control than Faraday Rings. You will not usually find these employed in drivers readily available to generic customers, they do tend to be applied in the OEM drivers for the better "High End" Studio Monitor and PA System manufatcurers that have their own drivers designed.
It is one of the better Midranges available, but the performance at realistic levels still leaves much to be desired, compared to a basic high performance pro driver which uses only HIGH SENSITIVITY to reduce distortion.
Of those I am familar with Seas and ATC and in OEM versions with Accuton (Thiel). Non that I have encountered are audibly low in distortion at realistic levels and compression becomes very audible.
I think you do. And you need to provide measurements of their Distortion at 96db/2m and preferably also of the compression at 50 Watt applied. I know how ATC (despite attempting to address the problems) compares and have a good idea howSeas manages (both are better than average but still rather poor).
And modulation of the Magnets field!
Actually, there is a certain link. In order to have higher sensitivity at a given frequency you usually require a larger cone. Now for the same SPL a larger cone will move much less and hence will be less subject to these problems. Simply compare a 15" Coaxial Speaker aimed at a 50Hz LF corner to a 6.5" aimed at the same LF corner. How much less excursion will the 15" Driver have for a given SPL at 50Hz and how much more sensitivity at (say) 100Hz will it display?
No, it doesn't. But compared to the prices charged for the agressively "High tech" appearing HiFi drivers a plain old Pro Audio Coax of large format compares surprisingly well when it comes to SPL vs Money, Distortion at a given SPL vs Money and and compression for a given SPL vs. Money. Indeed, some of the really good ones (Volt, certain Tannoy Models, certain JBL Models) will seriously whoop the best "HiFi" Drivers bum objectively and subjectively. Now just like HiFi drivers all Pro Drivers are not created equal, but on similar levels of engineering excellence the "HIFi" units are comprehensively outclassed, simply due to basic physics and acoustics.
Well, eddy current distortion follows the square of the power (cube of the current). So, if we have 0.03% 3rd harmonics eddy current distortion at 1 Watt we will get 3% at 10 Watt and (theoretically) 300% (meaning more distortion than fundamental) at 100 Watt. Based on measurements in the sadly long defunked "Audio" (US) most better "High End" speakers managed to exceed single number percentages 3rd Harmonics in the lower midrange (400Hz) once approaching the 50 Watt Mark.
If you invest comparable sums in "plain vanilla" pro Drivers as you would into middle of the road HiFi drivers you can, with carefull selection reap all these benefits. The German DIY Magazine "Klang + Ton" has shown many designs using Pro-Drivers aimed at providing High Fidelity (and beyond) reproduction when it comes to freedom from coloration and frequency response aberations while keeping the sensitivity in the 95db+ region and distortion compared to "High Fidelity" systems rather drastically lower (sadly they usually do not measure compression - no audio mag ever does, pitty - what are they afraid off?)
It is just that the US and UK audio weenies cannot comprehend basic physics and desire to ferverently believe that such a thing as a "good small speaker" can actually exist (wake up and smell the cofee - physics says it can't).
Sadly Audax has mostly dropped their pro audio line, using their apropriate 13" Woofer (it used to be around $ 150), 6.5" Cone Midrange (around $ 70) and titanium horn tweeter you could make a truely exceptional speaker. You could even build the 6.5" & Tweeter as seperate "minimonitor" Box and hide the big and badassed Woofer a little, people would walk into the shop (there was one inWest Berlin who had done such a system in the late 80's) and people would be shocked how a minimonitor could actually sound like a grown up Monitor.
Anyway, I'm rambling.
Try building a decent quality 3-Way using a 6.5" Cone Midrange (PHL?), a nice Tweeter (Fountek/Aurum Cantus Ribbon?) and a substantial 12 - 15" Pro Woofer and compare that to you favourite "HiFi" 3-Way system for the same budget. The Improvement is not subtle....
Sayonara
Konnichiwa,
Hans L said:Now we're talking
Depends on your definition of "high quality". In on of the the references I gave three different 6.5" Drivers in commercial 2-Way speakers where evaluated. The "best" unit was Dynaudio with a little under 3db compression at 50W compared to 0.5 Watt, the worst was a B& custom Job from the DM303 with almost 5db compression. Midband distortion levels for all of these speakers/drivers will exceed 0.5% @ 96db/1W/1m.
The Focal unit in the KRK Monitor and the Dynaudio unit in the Dynaudio Speaker would be classed by most people as being "high quality". Based on measured performance they are "clearly incapable of high fidelity reproduction".
So, no, no matter how you slice it, a low sensitivity drive unit CANNOT be high quality.
Hans L said:
A farraday ring is usually applied at a place other than the actual pole piece. In order to be truely effective you must have a significant copper sleeve on BOTH pole pieces IN THE AIRGAP. In addition you must choose a magnet geometry that evens out the magent field with coil travel.
The commonly found magnet systems with simple flat outer pole pieces, the inner pole piece flush with upper surface of the outer pole piece and a farraday ring on the inside of the ceramic magnet provide a very poor motor system that inherently remains high distortion, even though the faraday ring does a little to reduce distortion, it mostly does so for 2nd harmonics.
The problem in the midrange is invariably 3rd harmonics from the cubic function of the eddy current distortion and can only be addressed by either making the magnet pole pieces materially non-conducting (ATC) or fitting significant copper sleeves, both of which requires a significant increase in magnet size to at least partially compensate the loss of efficiency.
So no, if I am saying extreme measures I am thinking of something a little more involved (and expensive) than a farraday ring, which is mostly good for advertising it's presence. I would think either the JBL symmetric field geometry magent system or that developed and used by Lambda on their better woofers.
It still leaves other issues unaddressed which would require the use ideally of field coil magnets or "super strength" neodymium magnets.
Hans L said:
Serious Pro-Units usually incorporate much more sophisticated measures of distortion control than Faraday Rings. You will not usually find these employed in drivers readily available to generic customers, they do tend to be applied in the OEM drivers for the better "High End" Studio Monitor and PA System manufatcurers that have their own drivers designed.
Hans L said:
It is one of the better Midranges available, but the performance at realistic levels still leaves much to be desired, compared to a basic high performance pro driver which uses only HIGH SENSITIVITY to reduce distortion.
Hans L said:
Of those I am familar with Seas and ATC and in OEM versions with Accuton (Thiel). Non that I have encountered are audibly low in distortion at realistic levels and compression becomes very audible.
Hans L said:
I think you do. And you need to provide measurements of their Distortion at 96db/2m and preferably also of the compression at 50 Watt applied. I know how ATC (despite attempting to address the problems) compares and have a good idea howSeas manages (both are better than average but still rather poor).
Hans L said:
And modulation of the Magnets field!
Hans L said:
Actually, there is a certain link. In order to have higher sensitivity at a given frequency you usually require a larger cone. Now for the same SPL a larger cone will move much less and hence will be less subject to these problems. Simply compare a 15" Coaxial Speaker aimed at a 50Hz LF corner to a 6.5" aimed at the same LF corner. How much less excursion will the 15" Driver have for a given SPL at 50Hz and how much more sensitivity at (say) 100Hz will it display?
Hans L said:
No, it doesn't. But compared to the prices charged for the agressively "High tech" appearing HiFi drivers a plain old Pro Audio Coax of large format compares surprisingly well when it comes to SPL vs Money, Distortion at a given SPL vs Money and and compression for a given SPL vs. Money. Indeed, some of the really good ones (Volt, certain Tannoy Models, certain JBL Models) will seriously whoop the best "HiFi" Drivers bum objectively and subjectively. Now just like HiFi drivers all Pro Drivers are not created equal, but on similar levels of engineering excellence the "HIFi" units are comprehensively outclassed, simply due to basic physics and acoustics.
Hans L said:
Well, eddy current distortion follows the square of the power (cube of the current). So, if we have 0.03% 3rd harmonics eddy current distortion at 1 Watt we will get 3% at 10 Watt and (theoretically) 300% (meaning more distortion than fundamental) at 100 Watt. Based on measurements in the sadly long defunked "Audio" (US) most better "High End" speakers managed to exceed single number percentages 3rd Harmonics in the lower midrange (400Hz) once approaching the 50 Watt Mark.
Hans L said:
If you invest comparable sums in "plain vanilla" pro Drivers as you would into middle of the road HiFi drivers you can, with carefull selection reap all these benefits. The German DIY Magazine "Klang + Ton" has shown many designs using Pro-Drivers aimed at providing High Fidelity (and beyond) reproduction when it comes to freedom from coloration and frequency response aberations while keeping the sensitivity in the 95db+ region and distortion compared to "High Fidelity" systems rather drastically lower (sadly they usually do not measure compression - no audio mag ever does, pitty - what are they afraid off?)
It is just that the US and UK audio weenies cannot comprehend basic physics and desire to ferverently believe that such a thing as a "good small speaker" can actually exist (wake up and smell the cofee - physics says it can't).
Sadly Audax has mostly dropped their pro audio line, using their apropriate 13" Woofer (it used to be around $ 150), 6.5" Cone Midrange (around $ 70) and titanium horn tweeter you could make a truely exceptional speaker. You could even build the 6.5" & Tweeter as seperate "minimonitor" Box and hide the big and badassed Woofer a little, people would walk into the shop (there was one inWest Berlin who had done such a system in the late 80's) and people would be shocked how a minimonitor could actually sound like a grown up Monitor.
Anyway, I'm rambling.
Try building a decent quality 3-Way using a 6.5" Cone Midrange (PHL?), a nice Tweeter (Fountek/Aurum Cantus Ribbon?) and a substantial 12 - 15" Pro Woofer and compare that to you favourite "HiFi" 3-Way system for the same budget. The Improvement is not subtle....
Sayonara
Try looking at www.steenduelund.dk
The page is unfortunately in Danish but for those familiar with the language there's a lot of interesting information about how to build the optimum speaker, "cost no object".
"Cost no object" does not necessary have to be very expensive. The best units are not always the most expensive ones.
The page is unfortunately in Danish but for those familiar with the language there's a lot of interesting information about how to build the optimum speaker, "cost no object".
"Cost no object" does not necessary have to be very expensive. The best units are not always the most expensive ones.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.