But, each time you double the amount of drivers, you add 3dB to the efficiency (so less cone excursion and thermal compression). 8 drivers means 9dB more efficient than the spec for one driver. The 3.5 inch drivers I would like to use are the Dayton Audio ND91 at $22/each at Parts Express (Xmax is spec'd at 25mm p-p). Another option for almost no money (to test the theory about room interaction) is to use the Sony 5.25 inch drivers #299-012 "buyout" at $0.82 each, also at Parts express.
Roger Russel, former chief speaker engineer at McIntosh, has done much research with vertical line arrays, and is apparently completely sold on them. Apparently because of the way they interact with typical listening room acoustics. Google his name and check out his large website full of great info.
Last edited:
What you say is true, but I don't think the intent of the thread was to compare a single driver midrange (8" or otherwise) with arrays of multiple drivers. A line array is a completely different animal, and has its own pluses and minuses.
-Charlie
I have to say that Ive only ever owned one 3 way system, and it was a real mess to be frank.
Id have to say I like the idea of a small midbass/midrange:
Audax HM100Z0, HM130Z**, Visaton Ti100, AL130/130M (although a little inefficient)even full rangers like Mark audio in a similar diameter, or the Tangband W5-1611(?)
with a ribbon type tweeter, to limit vertical reflections a little... maybe fountek Neo of some sort?
Bass driver is the one I find hardest to choose
Id have to say I like the idea of a small midbass/midrange:
Audax HM100Z0, HM130Z**, Visaton Ti100, AL130/130M (although a little inefficient)even full rangers like Mark audio in a similar diameter, or the Tangband W5-1611(?)
with a ribbon type tweeter, to limit vertical reflections a little... maybe fountek Neo of some sort?
Bass driver is the one I find hardest to choose
Last edited:
The bass driver should be the easiest one...I have to say that Ive only ever owned one 3 way system, and it was a real mess to be frank.
Id have to say I like the idea of a small midbass/midrange:
Audax HM100Z0, HM130Z**, Visaton Ti100, AL130/130M (although a little inefficient)even full rangers like Mark audio in a similar diameter, or the Tangband W5-1611(?)
with a ribbon type tweeter, to limit vertical reflections a little... maybe fountek Neo of some sort?
Bass driver is the one I find hardest to choose
Small line array gives vertical directivity but no horisontal. That driver is VERY inefficient, 8 driver gives 90 dB + gain of impedance lowering (which bad thing for amp). With one single 6,5"-8" pro mid you can achieve 92-95 dB sensitivity and have no lobing issues between many mid drivers. Plus it gives more beaming (too much?) horizontally. It couple to air better than very small mid.
-----
Interesting matter is how big driver should be at mids? All know that at low frequencies bigger driver is ALWAYS better. There is the border where bigger is no better in mids? I believe that mid driver should also be quite big if you seek "ease" and dynamics of sound. Some say that reproduction of dynamics in natural sounds/music is more important than straight frequency response.
Last edited:
Moving mass and "detail" (whatever that may be, coz I am not aware of the existance of such a parameter) have nothing to do with each other.
Intuitively, it does seem to sound reasonably. In reality it is an unproven statement, and belings inthe same non-sensical category as slow or fast bass drivers.
Sorry to be harsh, but these statements are clear cut examples of loudspeaker myths.
Regards,
Eelco
Intuitively, it does seem to sound reasonably. In reality it is an unproven statement, and belings inthe same non-sensical category as slow or fast bass drivers.
Sorry to be harsh, but these statements are clear cut examples of loudspeaker myths.
Regards,
Eelco
But does the coin has its other side also? You have to feed 3x power to voice coil of smaller SD driver. It also moves 3x more p-p. And they have totally different radiation patterns so could they be compared that easily about energy storage in cone?
That also depends on the motor BL. Assuming both are equal you may have a point, but it is going to be a bit subjective since every midrange is different.
There simply is no simple and hard fast rule for midrange size.
This is why they make so many different types and sizes because it depends on the end application and the desired design goals.
A dedicated woofer maybe. Plenty of 8" full range drivers have an Mms including air mass load of around 9-11 grams.
One example:
http://www.tb-speaker.com/detail/1230_04/w8-1808.htm
Over 4 times the cone area (and far stronger Bl product) for only 50% more moving mass.
I don't quite see how you translate total cone moving mass to "stored energy". There is no direct relationship between the two independent of other parameters.
Apart from the fundamental bass resonance of a driver, most of the "stored energy" occurs within the cone break-up region, and depends on the material composition and damping of the cone, not just a raw "total mass" figure.
A lot of 4" wide range drivers I've measured have much worse cone breakup resonances in the upper midrange than some 8" full range drivers. Just because a cone is bigger or heavier doesn't automatically give it more "stored energy".
Last edited:
Midrange driver size is a very interesting debate, although I don't think you can make any generalizations without defining what midrange is first, and what crossover frequencies would have one driver covering the midrange instead of splitting it between drivers.
To me, midrange is 300-3000Hz, eg the telephone band, and if a driver can't cover this entire range it's not really a midrange driver.
You then have to consider what drivers its being crossed over to, and their characteristics.
For example lets say you were wanting to cross over to a 12" woofer at 300Hz. Does that make sense to cross to a 4" midrange driver, even if that driver can go down to 300Hz ?
No, because there is a huge difference in dynamic range potential of the two drivers at the crossover frequency. A 12" woofer can make a really loud noise at 300Hz with ease, low distortion, and very little dynamic range compression.
The difference in maximum SPL capability of the two drivers could be as much as 20-30dB at this frequency, with the 4" driver really struggling to provide sufficient SPL levels through the midrange, and with very high distortion. Frequency dependent dynamic range compression would then ensue at anything other than low playback level, where the midrange would have its dynamic range compressed.
Why so much importance at the low end of the midrange ? Because this is precisely where most of the energy in music is centred.
You have the double whammy of maximum excursion at the low end of the drivers range for a given SPL, and also maximum required SPL on typical music also occurring at the low end.
You can't just focus on small-signal polar responses at the high frequency end, without considering the large signal performance at the low frequency end.
Sufficient dynamic range at the low end of the midrange (in a large system) requires a certain minimum cone area for the midrange, period. My personal opinion on the best size for performance at the bottom end of the midrange is that 4" is way too small, (almost a joke as part of a larger system) 6.5" is bare minimum, and 8" is ideal.
Then you have the top end of the range at 3Khz to think about. The main criticism always levelled at large midrange drivers is "too directional". But what is too directional exactly, in terms of beam-width ? Where do we draw the line ?
I took some actual polar measurements of one of my 8" full range drivers for another thread recently, which I've attached below, for 1Khz, 2Khz, and 3.1Khz. How do these individual measurements compare with everyone's preconceived notions of how directional an 8" driver is ? Perhaps not quite as narrow as people expected ?
Depending on how much you're willing to let the pattern narrow you could easily cross it over at 2Khz, or even 3Khz at a pinch. What you cross it over to affects how high you can cross it as well, as mentioned already in the thread.
If you were crossing it over to a really wide dispersion dome, you probably wouldn't want to cross it any higher than 2Khz, (although 2Khz disqualifies it as covering "midrange" for me) or 3Khz at a stretch, but if you cross it over to a wave-guide controlled driver such as a small horn or ribbon, the directivity discontinuity becomes far less, and you can go a bit higher.
Does a small dip in power response below the crossover frequency really matter ? Not according to many pieces of research, which say that it goes largely unnoticed provided the dip is smooth and not excessive.
A 6.5" driver will fare a little bit better at the top end than an 8", but not by a lot, and you lose more at the bottom end than you gain at the top end, since the polar response gets narrower proportionally with the increase in radius, while the bottom end dynamic performance increases with the square of the radius. (eg area) Going from 6.5" to 8" gains you a lot more at the bottom end and only loses a small amount at the top.
More important than the polar response at the top end of a midrange drivers range is its cone breakup characteristics IMHO, and I pay far more attention to this than a small amount of narrowing of the polar pattern. The effects of a small dip in wide off axis response are debatable, while the effects of poorly controlled cone breakup affecting the on-axis response are clearly audible.
In theory a smaller driver should have less trouble with cone breakup and push that cone breakup higher in frequency, but for whatever reason its extremely common for cone breakup to fall in the same critical upper midrange region whether 5", 6.5" or 8". (Even some 4" drivers have bad breakup in this region)
It then becomes a matter of evaluating the performance of individual drivers rather than "smaller is better". It's perfectly possible to find an 8" driver with a smoother response in the breakup region than a lesser quality 6.5" or 5" driver.
If better dynamics is more important to you, find a good 8" driver. If wider polar pattern is more important to you, find a good 6.5" driver. I wouldn't stray outside that range though.
One more thing to consider about directivity at higher frequencies is that a moderate amount can be an ally not a foe. Baffle diffraction is a major design problem for many speakers particularly at higher frequencies, and what sometimes gets overlooked by those who favour small drivers with very wide polar patterns is that this gives maximum possible baffle diffraction ripple for a given baffle design, while a more directional driver greatly reduces the "illumination" of the baffle.
An 8" driver has enough directivity over about 1Khz to significantly reduce baffle diffraction, and over 2Khz it's almost a non issue.
Looking at the 90 degree angle response of a driver is a good indication of how much attenuation there will be for the signal travelling along the baffle which will diffract off the edge. Unfortunately I didn't take my polar plots right out to 90 degrees for practical reasons, but you can see roughly what the response would be at 90 degrees, -6dB at 1Khz, -18dB at 2Khz, and -22dB at 3.1Khz.
Even 6dB is a worthwhile reduction in diffraction signature, and 18dB is enough to practically eliminate it. Eliminating baffle diffraction signature in the presence and crossover region is well worth doing in my opinion.
Of course if you use a wide dispersion tweeter like a dome you will have diffraction signature from that, but a controlled directivity wave-guide tweeter can largely eliminate baffle diffraction from the tweeter as well, giving a fairly diffraction free crossover region even with an ordinary rectangular baffle.
I guess my overall position is that in many cases too much emphasis is placed on trying to prevent a driver becoming moderately directional at its top end, (which is a small-signal characteristic, and of debatable importance) at the expense of using a driver that is too small to have an adequate large-signal characteristic, (headroom, distortion etc) and for that reason I lean towards 8", when used with a controlled directivity tweeter, and wouldn't consider less than 6.5".
To me, midrange is 300-3000Hz, eg the telephone band, and if a driver can't cover this entire range it's not really a midrange driver.
You then have to consider what drivers its being crossed over to, and their characteristics.
For example lets say you were wanting to cross over to a 12" woofer at 300Hz. Does that make sense to cross to a 4" midrange driver, even if that driver can go down to 300Hz ?
No, because there is a huge difference in dynamic range potential of the two drivers at the crossover frequency. A 12" woofer can make a really loud noise at 300Hz with ease, low distortion, and very little dynamic range compression.
The difference in maximum SPL capability of the two drivers could be as much as 20-30dB at this frequency, with the 4" driver really struggling to provide sufficient SPL levels through the midrange, and with very high distortion. Frequency dependent dynamic range compression would then ensue at anything other than low playback level, where the midrange would have its dynamic range compressed.
Why so much importance at the low end of the midrange ? Because this is precisely where most of the energy in music is centred.
You have the double whammy of maximum excursion at the low end of the drivers range for a given SPL, and also maximum required SPL on typical music also occurring at the low end.
You can't just focus on small-signal polar responses at the high frequency end, without considering the large signal performance at the low frequency end.
Sufficient dynamic range at the low end of the midrange (in a large system) requires a certain minimum cone area for the midrange, period. My personal opinion on the best size for performance at the bottom end of the midrange is that 4" is way too small, (almost a joke as part of a larger system) 6.5" is bare minimum, and 8" is ideal.
Then you have the top end of the range at 3Khz to think about. The main criticism always levelled at large midrange drivers is "too directional". But what is too directional exactly, in terms of beam-width ? Where do we draw the line ?
I took some actual polar measurements of one of my 8" full range drivers for another thread recently, which I've attached below, for 1Khz, 2Khz, and 3.1Khz. How do these individual measurements compare with everyone's preconceived notions of how directional an 8" driver is ? Perhaps not quite as narrow as people expected ?
Depending on how much you're willing to let the pattern narrow you could easily cross it over at 2Khz, or even 3Khz at a pinch. What you cross it over to affects how high you can cross it as well, as mentioned already in the thread.
If you were crossing it over to a really wide dispersion dome, you probably wouldn't want to cross it any higher than 2Khz, (although 2Khz disqualifies it as covering "midrange" for me) or 3Khz at a stretch, but if you cross it over to a wave-guide controlled driver such as a small horn or ribbon, the directivity discontinuity becomes far less, and you can go a bit higher.
Does a small dip in power response below the crossover frequency really matter ? Not according to many pieces of research, which say that it goes largely unnoticed provided the dip is smooth and not excessive.
A 6.5" driver will fare a little bit better at the top end than an 8", but not by a lot, and you lose more at the bottom end than you gain at the top end, since the polar response gets narrower proportionally with the increase in radius, while the bottom end dynamic performance increases with the square of the radius. (eg area) Going from 6.5" to 8" gains you a lot more at the bottom end and only loses a small amount at the top.
More important than the polar response at the top end of a midrange drivers range is its cone breakup characteristics IMHO, and I pay far more attention to this than a small amount of narrowing of the polar pattern. The effects of a small dip in wide off axis response are debatable, while the effects of poorly controlled cone breakup affecting the on-axis response are clearly audible.
In theory a smaller driver should have less trouble with cone breakup and push that cone breakup higher in frequency, but for whatever reason its extremely common for cone breakup to fall in the same critical upper midrange region whether 5", 6.5" or 8". (Even some 4" drivers have bad breakup in this region)
It then becomes a matter of evaluating the performance of individual drivers rather than "smaller is better". It's perfectly possible to find an 8" driver with a smoother response in the breakup region than a lesser quality 6.5" or 5" driver.
If better dynamics is more important to you, find a good 8" driver. If wider polar pattern is more important to you, find a good 6.5" driver. I wouldn't stray outside that range though.
One more thing to consider about directivity at higher frequencies is that a moderate amount can be an ally not a foe. Baffle diffraction is a major design problem for many speakers particularly at higher frequencies, and what sometimes gets overlooked by those who favour small drivers with very wide polar patterns is that this gives maximum possible baffle diffraction ripple for a given baffle design, while a more directional driver greatly reduces the "illumination" of the baffle.
An 8" driver has enough directivity over about 1Khz to significantly reduce baffle diffraction, and over 2Khz it's almost a non issue.
Looking at the 90 degree angle response of a driver is a good indication of how much attenuation there will be for the signal travelling along the baffle which will diffract off the edge. Unfortunately I didn't take my polar plots right out to 90 degrees for practical reasons, but you can see roughly what the response would be at 90 degrees, -6dB at 1Khz, -18dB at 2Khz, and -22dB at 3.1Khz.
Even 6dB is a worthwhile reduction in diffraction signature, and 18dB is enough to practically eliminate it. Eliminating baffle diffraction signature in the presence and crossover region is well worth doing in my opinion.
Of course if you use a wide dispersion tweeter like a dome you will have diffraction signature from that, but a controlled directivity wave-guide tweeter can largely eliminate baffle diffraction from the tweeter as well, giving a fairly diffraction free crossover region even with an ordinary rectangular baffle.
I guess my overall position is that in many cases too much emphasis is placed on trying to prevent a driver becoming moderately directional at its top end, (which is a small-signal characteristic, and of debatable importance) at the expense of using a driver that is too small to have an adequate large-signal characteristic, (headroom, distortion etc) and for that reason I lean towards 8", when used with a controlled directivity tweeter, and wouldn't consider less than 6.5".
Attachments
Last edited:
well, in music midrange is more like 160-1600 Hz, just as bass is 16-160 Hz and highs 1600-16000 Hz
and here is midrange driver that - quite against laws of physics
- has advantages of small and large, in fact extremely large - effortless dynamics and power handling, low non-linear distortions, low energy storageMms around 2g or so and Sd around 2000 or so
extremely high directivity though - so for fans of total direct sound only
OTOH we are told by experts that it is objectively wise to take the room out of the sound equation, so no problem really
Last edited:
well, this is also my opinion exactly but some differ and prefer it this way and I think that such an array is best available solution for those people
actually typically it's Mms of about 1g and Sd of about 1300 cm2 for such a driver
sorry but not really - this is not math - this is physics
Last edited:
Simon,
I agree with your last post to a certain extent. However, to make matters a bit more specific, I ran a simple SPL vs excursion simulation in LspCAD. I simulated an Audax 10 cm/4" bass midrange in a 3.0 liter enclosure, a realistic boxvolume for a midrange. What is most interesting is cone excursion below 400 Hz.
Some data: with 2.83 V (1W in 8 Ohms) input , the output is approx 86.5 dB @ 1 meter for al frequencies above approx. 220 Hz. Peak-Peak excursion is less then 1 mm for any frequency above 180 Hz.
Doubling the input to 5.7 volts yields approx 92.5 dB at 1 meter at any frequency above approx 210 Hz. PP excursion is still less than 1 mm for any frequency above 250 Hz.
Increasing Vin to 8Volts (=8W) brings the Spl at 1 meter to 95 dB. X pp is still less than 1 mm for any frequency above 280 Hz.
The simulation above shows in my opinion that even a 100 mm midrange has sufficient output (Spl) capability in the 400-3000 Hz range. Normal listening levels rarely exceed 90 db at 1 M. I do not think dynamic compression is an issue with such small excursions. Keep in mind we normally listen to a stereo pair, so acoustic in room output is even 3 dB higher.
So in real life I do not think 4 inch midrange suffer from insufficient output capability. I do, however have to acknowledge there is al lot of energy in the midrange.
Coming to think of it, I vaguely recall John K. debunking dome tweeter dynamic compression issues here at diyaudio.
Eelco
I agree with your last post to a certain extent. However, to make matters a bit more specific, I ran a simple SPL vs excursion simulation in LspCAD. I simulated an Audax 10 cm/4" bass midrange in a 3.0 liter enclosure, a realistic boxvolume for a midrange. What is most interesting is cone excursion below 400 Hz.
Some data: with 2.83 V (1W in 8 Ohms) input , the output is approx 86.5 dB @ 1 meter for al frequencies above approx. 220 Hz. Peak-Peak excursion is less then 1 mm for any frequency above 180 Hz.
Doubling the input to 5.7 volts yields approx 92.5 dB at 1 meter at any frequency above approx 210 Hz. PP excursion is still less than 1 mm for any frequency above 250 Hz.
Increasing Vin to 8Volts (=8W) brings the Spl at 1 meter to 95 dB. X pp is still less than 1 mm for any frequency above 280 Hz.
The simulation above shows in my opinion that even a 100 mm midrange has sufficient output (Spl) capability in the 400-3000 Hz range. Normal listening levels rarely exceed 90 db at 1 M. I do not think dynamic compression is an issue with such small excursions. Keep in mind we normally listen to a stereo pair, so acoustic in room output is even 3 dB higher.
So in real life I do not think 4 inch midrange suffer from insufficient output capability. I do, however have to acknowledge there is al lot of energy in the midrange.
Coming to think of it, I vaguely recall John K. debunking dome tweeter dynamic compression issues here at diyaudio.
Eelco
I strongly agree with the paragraph that I am quoting above.
In fact, your whole post is right in line with my thinking. Thanks for taking the time to write it!
I also think that a midrange should be able to get down to 200-300Hz at system output with low distortion, and that a 6.5" driver is the minimum useful size for a midrange when used in this way.
-Charlie
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.