The reason I suggested using a full range driver in place of a traditional tweeter is so that the crossover point can be lowered significantly below 1kHz. I was thinking of 700-800Hz, definitely not 300Hz. Why? Well, the full range drivers that I have tested that can go up to 20kHz have a rising distortion profile below 500Hz that would make me not want to use them there. Above 700Hz the distortion is pretty low, at least by my standards. Not as low as the best dome tweeters, however. Also there is the power handling to consider. Despite the rating of 10W that you might see, it is similar to a tweeter in that there is a small voice coil with fine wire. It WILL heat up if too much of the audio band is put through it and you have to keep this in mind. It is likely more robust compared to a 1" dome tweeter but by crossing it lower you are giving it more power as its passband is wider. So it is a balancing act - lower the crossover point just enough to get rid of most of the off-axis holes but not so low that you either have to listen quietly or risk burning up the little driver. If the crossover point is 750Hz, then I would use the MFM arrangement. If you want to buid it as an MMT system, where the (lower) woofer is rolled off earlier, then you have additional flexibility and might even be able to cross over to the F/T at a higher frequency. It's basically a 2-way plus an extra woofer for more low end volume displacement. In the MTM the off-axis holes are coming from destructive interference between the MIDWOOFERS (the M in MTM), not the crossover itself or the F/T driver. So if you don't operate both midwoofers above 300-500Hz then you won't get the off-axis holes appearing at 800Hz as you do with the straight up MTM.
Amen, brother
Charlie,
Thank you. That is probably the most accurate and informative post I have read around here in a long time. Not that I am any know-it-all guru by any means, it's just that I have heard small "full range" drivers being pushed to limits they are not capable of.
Charlie,
Thank you. That is probably the most accurate and informative post I have read around here in a long time. Not that I am any know-it-all guru by any means, it's just that I have heard small "full range" drivers being pushed to limits they are not capable of.
What affect does the ribbons narrow vertical dispersion have on the systems vertical dispersion? Maybe people are thinking in terms of domes.
That was brilliant Charlie.
Yes one of the ideas I had was a 2.5 way MMF. I just could not find a good XO frequency to cross at. 700-800hz is right in the middle of the midrange where our hearing is probably the most sensitive.
No sir, with all due respect we are able to hear loudspeakers. I am 52 today and I must have started playing with loudspeaker drivers in my mid teens (13-14). So over almost 4 decades of listening to various drivers (oh everything from old Philips drivers to the Heil AMT to Decca ribbons and Panasonic ribbons (model 400 and there was a 1000 too), Celestion, Volt, Jordan, Vifa, Dynaudio, Gold Sound, JBL, Focal, Morel, Audax (they made a excellent paper cone P15 I think which we used to replace the 5" mid in a DQ10), etc..... we learn to listen. After 4 decades I have developed a certain degree of confidence in my ears (even as my hearing as slowly lost it's top end).
It might not be good enough for you and other, but it is good enough for me.
Here's the problem with the idea of using a full range driver sandwiched between two woofers for the sole purpose of maintaining a clean power response . . . FRs don't sound as good as proper tweeters in the top octaves.
In the MTM section of Nimbus we use a couple of excellent Accuton midwoofers (or Wavecor) flanking a RAAL ribbon. More than anything else, the character of those drivers is what you hear at the listening position. We take the time to voice the crossover so that the blending of the drivers is seamless presenting a smooth response in the crossover region. Power response issues aside, the resulting sound from those drivers is vastly superior than any FR driver on the planet. And as we learned from real world application, Nimbus produces a very smooth in-room response even through the crossover region between mids and tweeter.
Theory is fine but eventually you have to actually build something and listen to it. As it is, Nimbus wowed the crowd, vertical lobing and all. The quality of the drivers, the seamless blending of those drivers, results in an extremely pleasant and engaging listening experience.
Another of these theoretical chasms to overcome is the dreaded crossover in the critical midrange . . . hogwash.
I've made speakers that cross drivers outside the "critical range" and right in the middle of it. When done well, the result is great. I've done designs where the tweeter crosses with the woofer at 1000Hz and never once did I think, "voices sound off."
We just finished up RMAF displaying a 4-way crossed at 250, 850, 3000 and there was no shortage of people blown away by the result, many of them commenting on how realistic and clear the vocal presentation was.
Speaker design is as much art as it is science. Stick only with the generalized theory and you could miss building a sonic masterpiece.
In the MTM section of Nimbus we use a couple of excellent Accuton midwoofers (or Wavecor) flanking a RAAL ribbon. More than anything else, the character of those drivers is what you hear at the listening position. We take the time to voice the crossover so that the blending of the drivers is seamless presenting a smooth response in the crossover region. Power response issues aside, the resulting sound from those drivers is vastly superior than any FR driver on the planet. And as we learned from real world application, Nimbus produces a very smooth in-room response even through the crossover region between mids and tweeter.
Theory is fine but eventually you have to actually build something and listen to it. As it is, Nimbus wowed the crowd, vertical lobing and all. The quality of the drivers, the seamless blending of those drivers, results in an extremely pleasant and engaging listening experience.
Another of these theoretical chasms to overcome is the dreaded crossover in the critical midrange . . . hogwash.
I've made speakers that cross drivers outside the "critical range" and right in the middle of it. When done well, the result is great. I've done designs where the tweeter crosses with the woofer at 1000Hz and never once did I think, "voices sound off."
We just finished up RMAF displaying a 4-way crossed at 250, 850, 3000 and there was no shortage of people blown away by the result, many of them commenting on how realistic and clear the vocal presentation was.
Speaker design is as much art as it is science. Stick only with the generalized theory and you could miss building a sonic masterpiece.
Hello Pete. I'm interested in the horizontal dispersion you mention. Given a midrange driver, let's say the Wavecor WF182 I believe you use in the Nimbus, does the horizontal dispersion change if using TM vs MTM with the same drivers? My understanding is it doesn't change, and the choice of using two Ms has to do more with sensitivity, sense of impact, etc. is this correct?
Thanks so much for sharing here!
We use the WF152 in Nimbus (and C158 Accuton). You are correct in your assertion that the horizontal dispersion is unaffected by the dual mids in a vertical array. The use of two mids has two purposes and you identified one of them . . . adequate sensitivity to keep up with the woofer and tweeter. The other purpose is visual impact and looks matter when designing speakers for sale.
We are developing a version of Nimbus using only a single midrange for a couple of customers by utilizing more sensitive and larger midrange units.
We are developing a version of Nimbus using only a single midrange for a couple of customers by utilizing more sensitive and larger midrange units.
Can you share a polar map of the speaker you're discussing above to support that claim?
Also, your claim that "many" MTMs have very smooth horizontal response strikes me as at best only technically true. (That because "many" just means "more than one.") Most measured commercial MTM speakers exhibit midrange mushroom cloud polars. There are exceptions with good horizontal performance, such as David Smith's old Snells.
Here Pallas, just for you.
Below is the horizontal series we took for the Stiff Breeze which is basically the top half of Nimbus. 0-50 in 10 degree steps, 2dB per division.
As I said, response stays smooth as you move off axis. Could it be better? Possibly but the improvements would be marginal at best at the listening position. As is, this particular MTM is a great sounding speaker because of the components used.
Below is the horizontal series we took for the Stiff Breeze which is basically the top half of Nimbus. 0-50 in 10 degree steps, 2dB per division.
As I said, response stays smooth as you move off axis. Could it be better? Possibly but the improvements would be marginal at best at the listening position. As is, this particular MTM is a great sounding speaker because of the components used.
If you compare apples to apples, meaning the same driver or at least the same diameter driver, and you remain on the tweeter axis in a horizontal plane, an MTM will have essentially the same horizontal pattern as an MT.
The main objection about off axis response is in the vertical plane where the distance to the two M's becomes different enough such that, at even middle frequencies, interference between the two causes lobing or "beaming".
It is all a matter of the compromises you make and all speakers have significant compromises. Any XO where the C-C is greater than a quarter-wavelength is a compromise.
Price is as well. The cost of the FASTs we have done are less than the tweeter and still provide superb performance.
dave
A quote from Earl who have done proper scientific experiments on the subject:
We have a FAST where FF85wKeN is XOed at 240Hz 1st order, and have gotten very good performance.
dave
I'm not armchair quarterbacking here, Pete. I current have an MTM system that I built that is crossed over low (800Hz) and that uses a full-ranger as the "tweeter" to solve the off axis holes problem. It's built AND listened to and sounds pretty good on and off axis. NOT THEORY! Really, Pete, it works. 🙂
Saying all fullrangers "don't sound as good a proper tweeters" is an over generalization. It's like saying "all ribbon tweeters are just distortion making machines compared to proper tweeters" (another generalization). Your RAAL unit for instance is certainly a standout among ribbons. In fact it's probably the best of the best, with a pretty extraordinary price, too. Honestly, I rarely spend that much on all the drivers in a speaker. I'm sure you would agree that you could not have used just any old ribbon and have achieved the same level of fidelity. The same true is for fullrange drivers. There are certainly many poorly performing examples of this type of driver. The design has to balance several aspects of direct radiator design and performance, but nonetheless there are some that can, and do, sound very good although you have to live with somewhat lower sensitivity (e.g. 85dB/W).
I you are writing off full range drivers as a whole, that would seem to be a bit disingenuous on your part. Your whole argument that "[full range drivers] don't sound as good in the top octaves" doesn't hold water in my opinion, at least for the full range drivers I am thinking of. I'm not talking about an 8" driver with a whizzer cone... The truth is that the large majority of people can't hear above 15kHz or so and between about 7kHz and 15kHz the ear is rather insensitive to frequency response deviations as long as there is some sound power in that band. So if by "top octaves" you mean 2.5-10kHz, then sure, but IMHO a small full range driver can reproduce this band well. Still higher frequencies just don't matter nearly as much. "Airy details" are happening around 5-10kHz not 50kHz, and the "ultrasonic response" of a ribbon is simply solving what amounts to a non-problem.
Sure, but when I measure a driver and even at very moderate SPLs the distortion is already climbing above 1% I tend to avoid using it in that part of its passband...
Earl is right. Distortion from loudspeakers is surprisingly high before you really notice it. But since I am able to reduce distortion by simply increasing the crossover point I did that. Perhaps this was not a possibility for your system or you had some other design constraint.
Moving the XO up is an option but i like to keep as much of the midrange as possible in the FR. A higher XO also means that driver spacing needs to be tighter. The insignificance of distortion means tome that i can mostly ignore it, it seems a compromise a design to minimize something unimportant at the sacrifice of something that is.
dave
dave
And there you have it. Until you actually build systems using those expensive, top-shelf drivers you really don't know what they can do and how far they separate themselves from little full range drivers.
We don't hear polar response as much as we hear first arrival. Much of the character of the system is in that signal. The detail, instrument separation, the ability to handle complex passages, are revealed in the first arrival. There is no full range driver that can match the character of the drivers we use in the Nimbus, especially at higher SPL levels.
I'm not saying that FAST systems can't sound good. But there is no FAST system that will sound as good as a properly implemented 3-way using top shelf components like Accuton and RAAL.
To me, there is a much larger compromise attempting FAST than there is building MTM with vertical lobing. The polars might look better with FAST but the ability to properly render complex passages using a 3" driver covering 5 octaves is not my idea of high fidelity. That poor little sucker would become a blurred mess during complex orchestral passages at near real life SPL.
I decided to stay away from the Wavecor drivers since their sensitivity is lower than what I need. The TPL-150H is 102dB sensitive and I'm aiming for a midrange section that gets as close as possible so I can drive each (active xo) with a 1W SET. So a pair of midranges in MTM is appealing in this regard. Might need to experiment and listen for myself. I have a couple B&W FST drivers in my 804S and they look like great midranges, at 94dB sensitivity...
Alternatively, I was considering an 8" midrange with 99dB sensitivity or so, such as PHL. I see you mentioned using a larger midrange with the Raal ribbon. How high are you planning to xo and how large a midrange are you considering?
I'm very familiar with the TPL150H. One thing you'll find with it is a very narrow vertical window. It's a LARGE AMT and that large vertical span produces a very tight beam above 5000Hz. When you are in line with that tweeter, the delivery of highs is exceedingly good. Get too far above or below and the top octaves are gone. It also takes a bit of EQ to flatten it out properly. We did it passively in our Arcus model using 7 components (3 caps, 2 inductors, 2 resistors).
We are building a Nimbus that will use the Accuton C173 neo midrange paired with the 140-15D, the same drivers we use in the Joule White. XO point will likely be near 2000Hz LR4.
We are building a Nimbus that will use the Accuton C173 neo midrange paired with the 140-15D, the same drivers we use in the Joule White. XO point will likely be near 2000Hz LR4.
Thanks for the data, but they clearly fail to support your assertion. Even though the plot is more difficult to read than a polar map, there is clear evidence of midrange mushroom cloud from 2.5-4 kHz.
I know part of the job of a luxury goods salesman like you is cultivating parts fetishism. But so is knowing your audience. 😉
As for "resolution" of widebanders...my first instinct is to agree with you, just because widebanders tend not to have as smooth axial frequency response and have resonances that are lower down in frequency than typical tweeters. That said, I've been meaning to order those WG's Dave Pellegrene made for the Aura Whisper for a while, because that driver on a good WG could be the (wonderfully cheap) exception to prove the rule if they can play loudly enough. Especially for less critical applications such as height speakers in immersive setups.
- Status
- Not open for further replies.