I came across a post talking about midrange arrays that picked my interest and would like to learn more.
I'd like to consider them in the context of a multiway system, with TPL-150H as tweeter above 2kHz and large woofers taking care below 350Hz.
I like the ability to use multiple smallish midranges (4 to 6") to achieve a large total radiating surface, high sensitivity that can be achieved, and the evening out of reflections. I'm sure there are drawbacks as it would otherwise be more widely used, and I'd like to learn pros/cons of this approach.
Thank you!
I'd like to consider them in the context of a multiway system, with TPL-150H as tweeter above 2kHz and large woofers taking care below 350Hz.
I like the ability to use multiple smallish midranges (4 to 6") to achieve a large total radiating surface, high sensitivity that can be achieved, and the evening out of reflections. I'm sure there are drawbacks as it would otherwise be more widely used, and I'd like to learn pros/cons of this approach.
Thank you!
It's actually used quite often. The first doubling gives the most benefit, with lesser returns until the line is long enough to really change the dropoff-with-distance. My Pentas have four mids with one tweeter nestled in next to the middle. They still fall off at 6db/2x distance, so there is no mismatch with the tweeter's loudness at a given distance.
So go ahead and play with it. The longer your line the more you will restrict the vertical dispersion, this may help in your room; if you are tall and your couch low, you might want a shorter line. If your line is really long you may want a line of tweeters.
So go ahead and play with it. The longer your line the more you will restrict the vertical dispersion, this may help in your room; if you are tall and your couch low, you might want a shorter line. If your line is really long you may want a line of tweeters.
Thank you.
Can you explain why the longer the line the more the vertical dispersion is restricted?
I'm not tall, but my AMT tweeter has an 80x30 degree horn so vertical dispersion mismatch is one of my concerns.
Okay. Look at it this way. You have heard that smaller mids have wider dispersion than larger ones. If we make a column of 4 inchers it's still 4 inches side to side, but four of them are a 16 incher vertically. This is the same effect that is a problem when you lay an mtm on its side as a center channel and you have a very wide seating area close to the screen&speaker. In your case with that tweeter I would say that a line of four little mids or at least two bigger ones(mtm) is not only okay, but probably necessary to match the dispersion of your tweeter. Seems like the analog part of your brain had the right idea- the other part said "wait...what?"
If you had told us of your tweeter, this is what I would have suggested without knowing you were interested in a line of mids. To put it another way, if you had said you wanted a single tweeter for a line, something like what you have is what I would suggest. So have a little more confidence in your instincts, they're pretty good.
If you had told us of your tweeter, this is what I would have suggested without knowing you were interested in a line of mids. To put it another way, if you had said you wanted a single tweeter for a line, something like what you have is what I would suggest. So have a little more confidence in your instincts, they're pretty good.
As far as matching goes, the issue is really level related. If you put 4 x 5" drivers vertically, that's about 2' worth of vertical coverage, assuming the tightest possible dispersion. That's pretty comfortable for sitting or standing in front of the speakers in most homes.
In my humble and inexperienced opinion, a larger problem is the drop off over distance.
At 9' , 12', 16', etc. the tweeter and mid levels may no longer match. So long as you can adjust for this, you should be good. I often see line arrays with L-Pads for just this reason.
With different dispersion patterns the rate at which the level drops off is quite different. This is not unique to line arrays though. Horn loaded systems with direct radiating woofers have the same kinds of issues.
Do consider an AMT without a horn though, as the dispersion characteristics may be much closer.
Best,
E
In my humble and inexperienced opinion, a larger problem is the drop off over distance.
At 9' , 12', 16', etc. the tweeter and mid levels may no longer match. So long as you can adjust for this, you should be good. I often see line arrays with L-Pads for just this reason.
With different dispersion patterns the rate at which the level drops off is quite different. This is not unique to line arrays though. Horn loaded systems with direct radiating woofers have the same kinds of issues.
Do consider an AMT without a horn though, as the dispersion characteristics may be much closer.
Best,
E
Last edited:
It's a particularly good idea for the mid-bass transition for baffle-step (up to about 450 Hz - depending on baffle size). (..say 4-6 drivers from floor, up to tweeter with mid array above that.) The midbass array becomes a "lumped" grouping of reflections that minimize suck-out from floor-bounce while also allowing for an easier crossover that doesn't need baffle-step compensation. It also "loads" or is bounded to the floor to improve loudness (which is particularly useful for a very narrow baffle).
For the mid.s, when you have 4 or more drivers you'll get added gain between about 400 Hz and 1.3 kHz (..ex. with 4 drivers, maybe an added +1 db in addition to other sources of gain). The more drivers, the more added gain.
As far as difficulty with line-source behavior (..lack of pressure loss vs. point source), you can look at the line-source transition for listening distance on page 8 fig. 5 here:
http://www.audioroundtable.com/misc/nflawp.pdf
-basically, a half-meter (about 20"s) array of mid.s at a 2 meter distance will display point-source pressure loss behavior (up to almost 5 kHz) - so that should be sufficient for most low-pass crossovers around 2 kHz. Note: the further the listening distance the higher the freq. where it starts becoming a line source.
The mid. array allows for enough height off of the floor to lower the freq.s where any floor-bounce suck-out would occur.
Floor/Ceiling Reflection Calculator
The one thing you should plan-on with a design like this is a stepped baffle, with the tweeter being inset the most, then the mid.s, and no inset for the woofer array. (..and typically with a bit of felt placed above and below the tweeter to avoid reflections.)
Overall, a design like this can provide good eff. for a narrow baffle while simplifying the crossover design and providing a fair bit of rejection from floor and ceiling bounce suck-out.
Frankly I'm surprised I don't see more commercial designs based on this format.
Last edited:
Have you ever looked at Nola Speakers - specifically the Baby Grandes?
Nola by Accent Speaker Technology Ltd.
I forgot to mention - open baffle too.
Nola by Accent Speaker Technology Ltd.
I forgot to mention - open baffle too.
Last edited:
An array of drivers simply create a wavefront that's shaped like the array.
IE, if you have a flat array that's one meter tall, you're going to get a flat wavefront at frequencies that are one meter and smaller. (340Hz is one meter long.)
If you have a curve array that's one meter tall, you're going to get a curve wavefront at frequencies that are one meter and smaller.
Thanks for the link to Griffin's paper!! It has been a wonderful read.
I need to run some numbers to get a grip of what it would mean in my case, but can see some interesting pointers as well as some concerns to be further understood.
I didn't mention in my initial post but I'm looking to drive the midrange with 45 or 2a3 SET, so sensitivity is a significant issue as is impedance. Before reading the paper I was more concerned about this, but now I see arrays could have a very interesting match with SETs. A design such that the listening position is within the near field, keeping at 3dB per doubling of distance, would make the midrange extra efficient and suitable for SETs. The concern is the mismatch between the AMT tweeter behaving far field vs the midrange behaving near field; to be further analyzed.
Stepped baffle to time-align the drivers? Making sure I'm not missing something, as my system is active and time-aligned through the digital xo and DSP, so wouldn't need to step the baffle for time-alignment, while I do need to manage diffraction and reflections of course.
Indeed I had. And I keep going back to see them again
Do you happen to know which 4.5" midrange drivers they use?
Phase issue is not unavoidable with Array or MTM. I compared single midrange and multiple ones with the 1500Hz 4 pole crossover. Although the difference can't be clearly measured with my low cost environment, the multiple midranges certainly blur the transient. Vocal is noticeably less intimate, more artificial than single unit.
Dynaudio doesn't seem to mind the problems of using multiple mids (or woofers):
Kind of reminds me of the Dunlavy WMTMW setups...
Dunlavy Audio Labs SC-IV/A loudspeaker | Stereophile.com
Lots more examples like that around. Gryphon comes to mind:
Or these arrays from the same firm:
Maybe even more appropriate to this thread:
Fisher & Fisher:
Cost may be a very good reason we don't see it more often.
Just browse this page for even more examples: pro-arrakis-01 - Speakers High End | Pinterest
Kind of reminds me of the Dunlavy WMTMW setups...
Dunlavy Audio Labs SC-IV/A loudspeaker | Stereophile.com
Lots more examples like that around. Gryphon comes to mind:
An externally hosted image should be here but it was not working when we last tested it.
Or these arrays from the same firm:
Maybe even more appropriate to this thread:
Fisher & Fisher:
Cost may be a very good reason we don't see it more often.
Just browse this page for even more examples: pro-arrakis-01 - Speakers High End | Pinterest
Waysayso, an average audiophile just wants to see more units on the speakers sadly, and I believe that is the main reason why mainstream audiophile speakers have a lot of units. The cost of the unit is trivial compared to the whole speaker price.
I have been using MTM monitors for 15 years, and I know their pros and cons.
I have been using MTM monitors for 15 years, and I know their pros and cons.
I ran some numbers and looking to get feedback from experienced users.
Glad to see wesayo here - loved your tower build thread! I checked you use the Vifa drivers.
I looked around some and decided to run my excercise with Faital Pro 4FE35: nominal 8 ohm, 91 dB sensitivity, 100mm flange diameter. It has very nice treble extension and should easily get to 3kHz and my xo point is targeted at 2kHz.
Keeping center-to center under 1 wavelength at 2kHz, 170mm, I choose 150mm, or 50mm in between drivers. So a 6-driver array would be 85cm tall (h).
The paper talks about a tweeter low frequency cutoff as 1/(3h), or 392Hz. Also talks about a woofer low frequency cutoff as 1/(9h), or 131Hz. Considering I'm thinking a midrange array with a low xo point around 450Hz I don't know if these have any bearing in my calculations? Anyone?
A key consideration is the transition distance between near and far-field, defined as 1.5*f*h^2. Using 450Hz for f, I get 49cm transition distance...Using 2kHz for f I get 2.17m which is close enough I guess to my 2.4m listening distance (front baffle to ears). So basically all the midrange will be operating in the far field.
Power tapering: I modeled drivers 3 and 4 connected in parallel to 1 and 2 connected in series between them, and also 5 and 6 in series between them. The driver is 8 ohm nominal. BTW, should I use nominal or minimun impedance for this?
So the array impedance:
1/Za = 1/8 + 1/8 + 1/16 + 1/16, so Za = 2.7 ohm.
Hence the efficiency gain would be 7.8dB and the system gain would be 4.8dB, therefore a total midrange sensitivity of 103.6dB (91+7.8+4.8).
Eager to receive feedback!!
Glad to see wesayo here - loved your tower build thread! I checked you use the Vifa drivers.
I looked around some and decided to run my excercise with Faital Pro 4FE35: nominal 8 ohm, 91 dB sensitivity, 100mm flange diameter. It has very nice treble extension and should easily get to 3kHz and my xo point is targeted at 2kHz.
Keeping center-to center under 1 wavelength at 2kHz, 170mm, I choose 150mm, or 50mm in between drivers. So a 6-driver array would be 85cm tall (h).
The paper talks about a tweeter low frequency cutoff as 1/(3h), or 392Hz. Also talks about a woofer low frequency cutoff as 1/(9h), or 131Hz. Considering I'm thinking a midrange array with a low xo point around 450Hz I don't know if these have any bearing in my calculations? Anyone?
A key consideration is the transition distance between near and far-field, defined as 1.5*f*h^2. Using 450Hz for f, I get 49cm transition distance...Using 2kHz for f I get 2.17m which is close enough I guess to my 2.4m listening distance (front baffle to ears). So basically all the midrange will be operating in the far field.
Power tapering: I modeled drivers 3 and 4 connected in parallel to 1 and 2 connected in series between them, and also 5 and 6 in series between them. The driver is 8 ohm nominal. BTW, should I use nominal or minimun impedance for this?
So the array impedance:
1/Za = 1/8 + 1/8 + 1/16 + 1/16, so Za = 2.7 ohm.
Hence the efficiency gain would be 7.8dB and the system gain would be 4.8dB, therefore a total midrange sensitivity of 103.6dB (91+7.8+4.8).
Eager to receive feedback!!
Well, that's my taste and belief. I simply prefer Voxativ, Altec or Tannoy to Dynaudio, although I still use 2 different MTM monitors for some reasons.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.