The BaSSlines (was High Sensitivity Design)

[---k---]

Dan and Jeff, thanks for pointing out my mis-match measurements. Makes sense. I'm looking forward to seeing your far field measurements.

[dlneubec]

Electrovoice used to say....without high efficiency there is no accuracy. I still remember their marketing quote from the 70's. I guess that dates me a little.

Dan, I didn't think we turned them up much at all. The speaker had a lot in reserve and I really wished I could have heard that, but last Saturday was probably not the venue for it.

I have no doubt that you can voice a crossover that will sound very good and balanced. I heard your Soundrounds, so I already know you can do it. Those were wonderful speaker; Jim and I both thought so, as we discussed it later.

I have to jump on John J.'s bandwagon here - the Lambda woofers he is making are very special drivers, capable of doing things that most drivers can not begin to touch. For example their ability to reach deep in the bass with a good Xmax, and at the same time maintain extremely low non-linear distortion values into the midrange are pretty much unprecedented. The 12" I used had the Le of 4" driver and had a flat response on-axis to 2.5kHz. The Lambda will give you bass slam and a low distortion mid that can cross effectively to a dedicated midrange that you will not get with hardly any other woofer. It was very much a part of what you heard last weekend.

I would not be so picky about the tweeter, the Morel MDT33 is one of the best ever made in terms of flat on and off-axis response and low distortion, but there are other tweeters in its league that I would recommend as well.

With the midrange, though, there is a reason why the last two large three-ways I designed (both for Jim by the way) used the PHL rather than something else, and it had to do with the driver's capabilities in the 400 Hz region that set it apart from the rest. There really is a lot of thinking that goes into this sometimes, and Jim and I were willing to go through three woofers before we were happy, Just like I went through three different woofers AND tweeters before I settled on the drivers in my small monitors you heard on Saturday (I can't even begin to tell you how many crossover variants I listened to ;- )

I got your email and I would be glad to coach you along on this project and make some recommendations, but I don't think I should do that in a public forum since the speaker you heard Saturday will be sold by Salk Sound and I need to keep some of the details of that design proprietary. Some specifics I can't share, but some guidance I am willing to give you privately. Of course, I am sure you understand.

Finally, I am tickled pink that you enjoyed the speaker as much as you did, and I am reminded that imitation is the sincerest form of flattery. Consequently, I am very flattered by all of this attention.

Jeff B.

Hi Jeff,

Glad to see you chime in here. I think you've convinced me that I'll have to dig deeper in my pocket and get the Lambda's and the PHL's, so that leaves the tweeter to try and save some money.

As I said elsewhere, it sounded like Doc S. was in the room playing on that one piece, which is something I'd love to have in my living room. Thanks for the comments about the SoundRounds. I put a lot of thought and work into that design and it always helps to hear that some expereinced ears appreciated the results.

I'll be in contact with you privately to discover whatever you can pass along in terms of guidance on this project, while still protecting Jims investment. Your offer to supply any guidance at all is much appreciated.

[augerpro]

Jeff I think a lot of us here would like any tips or nuances of OB design you are comfortable sharing. Maybe measurement tips, dealing with the dipole peak, determining baffle according to XO point (which isn't real clear to me for dipoles), etc.

[Dennis H]

Hey Jeff,

Good to see you here!

Just a FWIW, Morel is listing the MDT33 as discontinued and Madisound lists it as out of stock. The "replacement" (E something?) doesn't look the same to me. Maybe you and Jim have a source for new out of stock drivers that can let him do a decent run of the design?

[Jeff B.]

Et338

Hey Jeff,

Good to see you here!

Just a FWIW, Morel is listing the MDT33 as discontinued and Madisound lists it as out of stock. The "replacement" (E something?) doesn't look the same to me. Maybe you and Jim have a source for new out of stock drivers that can let him do a decent run of the design?

The ET338 is the same tweeter with a slight change to the face plate and some cosmetic changes. The dome, coil, and motor are all the same as the MDT33. The impedance and response are very close.

Jeff B.

[Mark K]

The measurement you posted for the PHL says it was made at 1" nearfield. measurements above the low bass range are not accurate at this nearfield distance. My PHL's measured nothing like the response shown here. I work with measurements referenced to one meter, which are much more applicable for crossover design. The RS180 measurement on the other hand is at one meter, but then again, it needs to be taken on the baffle to show the baffle effects included so those can be dealt with in the design. I am working with the RS180 right now in a pair of powered studio monitors for a friend's recording studio, and the PHL is much flatter than the RS180 in the range I used it in.

Jeff B.

Ha! They're very accurate! That's why I did them at 1"...

But Jeff is right-they are not for crossover design and not to be used in comparison to the typical 0.5-1m farfield curve.

So, why did I do them at 1"? Well, nearfield or close mic'd measurements have a much better ability to resolve higher q/narrower frequencie irregularities than the typical short windowed 1m measurement. 1m measurements don't have very much resolution unless you have 10-20ms before the first reflection.

So a 1"/nearfield measure is a pretty good indicator of the drivers inherent linearity for narrow Q peaks and dips. However, it will not typically look flat, and you should only compare 1" to other 1" measurements.

Again, the measurements on my site are more geared to look at comparative linearity and nonlinearity between drivers. The measurements are not for crossover work.

I've been curious about this thread, and some pro drivers as well. I've thought (the operative word is "thought"--I've got way too many projects, too slow at each, to even verbalize another project...) about designing a high sensitivity design. Heck, I've even thought about buying Dr. Geddes diy horn kit if it ever comes out. It would be nice to measure and test that!

[Rudy Jakubin]

I guess it was some one else. I tried to search the pe forum but their search function is horrible. Trying to search "b&c" returns all results for "b,c" so just about every post.
Maybe it was Mark K, I cant remember.

Well it turned out to be Charlie Laub;

http://www.audio.claub.net/projectspage.html

And his WAF-00 concept speaker. Mark K was there so maybe he can still remember how they sounded.

http://www.ncdiyaudio.claub.net/Meeting06.html

I still remember another Pro Sound guy using B&C mids and liking them, maybe Warren G....

Reading Charlie Laubs writeup he did come across a problem at 350 hz that Jeff B brought up.

[Dennis H]

I think Mark Seaton and Tom Danley use a lot of B&C drivers.

[JoshK]

US Speakers shows the Beyman T2010 and the T2030, the prior 92/db softdome $59/ea, the later 95/db harddome $79/ea. I'd think your Audax PR1070's could reach high enough to cross to these tweeters well. Then you'd need to just solve the woofer decision.

Whatever you choose, I'm interested to see your progress. I find your work very interesting and I am most interested in higher efficiency designs.

[looneybomber]

I haven't seen a lot of talk about ribbons. Did I miss a comment somewhere stating something along the lines of "no ribbons please"? Sure their vertical dispersion isn't so great, but there are some pretty sensitive ribbons out there.

[augerpro]

Dunno Looney. I know Dan was at first trying to control cost, and 95dB/watt ribbons tend be the bigger ones. Which are spendy.

[dlneubec]

I haven't seen a lot of talk about ribbons. Did I miss a comment somewhere stating something along the lines of "no ribbons please"? Sure their vertical dispersion isn't so great, but there are some pretty sensitive ribbons out there.

I haven't looked into ribbons much at this point because my goal was to reduce the cost to less than what it would take with the Morel MDT33 and I figured ones that could cross low enough and be sensitive enough would exceed that amount, but I certainly could be wrong in that assumption. I'm certainly not opposed to ribbons if they meet the other goals.

That said, I think that Jim Salk prefers and uses ribbons in most of his designs, so there was probably a pretty strong reason that he and Jeff decided to go with dome. I'll have to ask Jeff if he can shed some light on the reasons for that.

[Jeff B.]

I haven't looked into ribbons much at this point because my goal was to reduce the cost to less than what it would take with the Morel MDT33 and I figured ones that could cross low enough and be sensitive enough would exceed that amount, but I certainly could be wrong in that assumption. I'm certainly not opposed to ribbons if they meet the other goals.

That said, I think that Jim Salk prefers and uses ribbons in most of his designs, so there was probably a pretty strong reason that he and Jeff decided to go with dome. I'll have to ask Jeff if he can shed some light on the reasons for that.

Remember, I built the speaker for myself, then Jim liked it enough that he took it and rebuilt it for Salk Sound. The tweeter choice was based on the fact that I had a pair of MDT33's sitting on the shelf collecting dust ;- ) Yeah, sometimes that's all the reason you need. If we had built it from the ground up for Jim I am sure it would have had a Aurum Cantus G2 ribbon in it. I now have a pair of those sitting on the shelf too, and I have already worked up a crossover for the same speaker but with the ribbon tweeter instead of the Morel. Personally, I have always preferred the better domes over ribbons because they sound more dynamic in their lower range, especially around 2khz, where you are blending it to the midrange driver, but that's just my own preference. Here's a link to a thread where I explain a little more behind the whole thing.

http://www.audiocircle.com/circles/index.php?topic=54161.0

You'll have to plow through a ways to find my posts though.

Jeff

[Scottg]

at some point you will have to integrate Hi. Eff. with a radial config..

Hmm, drivers..

Yes the Audax, you'll not find better for less. It will however require a good enough size baffle AND a fairly high/steep crossover, OR more of them (and that defeats your desire for lower cost).

For the bass driver you will need something that is EXTREMELY efficient to overcome BSC and have any hope at all of matching with a 96 db mid. Then there is getting it at a lower cost.

Note that you cannot go to a lower eff. driver with "boost" and hope of getting a similar sound.

I'd suggest looking at the 18 Sound 15W500. Contact LouspeakerPlus and see what their price is, remember this is their bargain 15 inch midbass and shouldn't cost to much - its 81 euros from Art Studio (..about 125 US), anything under 145 US should be good.

With such a driver you'll want a *LARGE* box and low freq. vent. (near 30 Hz).

With a heavy cabinet (mass loaded) and a stiff coupling to the baffle (say 1/4 inch steel plate ring) - you'll have BETTER sound than what you heard as long as you don't go for extreme SPL's.

The Audax has the capability of being run full-range, with a bit of work you can achieve linearity and a much higher freq. low-pass character than you might ordinarily for a 6.5 inch driver.

I'd suggest looking at the HiVi RT1C for your high freq. "filler". It will be marginally lower in spls vs. the Audax, BUT because its a monopole you should have a slightly reduced spl for it, otherwise it will sound "hot". Work toward a slightly downward sloping response from about 2 kHz with the two drivers. Note that because of its vertical directivity it should be near ear level OR should be on a baffle with "tilt" that will "aim" the driver to the listener.

That should give you what you are looking for this time out.

If you need an even cheaper midbass driver solution look to the Selenium 15PW3-SLF. Needs a larger cabinet, has a "hot"er top-end, isn't as eff. and has less linear movement. BUT it should be about half the price. :T

[Dennis H]

I'd suggest looking at the HiVi RT1C for your high freq. "filler".

That's a very high-distortion driver.

[Scottg]

That's a very high-distortion driver.

How is it used?

Also, does it really make a difference? (..and no, not on a test tone but with actual music.)

Note that above 6 kHz it shouldn't be bad, even 2nd order should be less than 1% at 1 watt with a modest filter. Pretty tough to hear that even with a test tone.

[Jeff B.]

I disagree

at some point you will have to integrate Hi. Eff. with a radial config..

Hmm, drivers..

Yes the Audax, you'll not find better for less. It will however require a good enough size baffle AND a fairly high/steep crossover, OR more of them (and that defeats your desire for lower cost).

For the bass driver you will need something that is EXTREMELY efficient to overcome BSC and have any hope at all of matching with a 96 db mid. Then there is getting it at a lower cost.

Note that you cannot go to a lower eff. driver with "boost" and hope of getting a similar sound.

I'd suggest looking at the 18 Sound 15W500. Contact LouspeakerPlus and see what their price is, remember this is their bargain 15 inch midbass and shouldn't cost to much - its 81 euros from Art Studio (..about 125 US), anything under 145 US should be good.

With such a driver you'll want a *LARGE* box and low freq. vent. (near 30 Hz).

With a heavy cabinet (mass loaded) and a stiff coupling to the baffle (say 1/4 inch steel plate ring) - you'll have BETTER sound than what you heard as long as you don't go for extreme SPL's.

The Audax has the capability of being run full-range, with a bit of work you can achieve linearity and a much higher freq. low-pass character than you might ordinarily for a 6.5 inch driver.

I'd suggest looking at the HiVi RT1C for your high freq. "filler". It will be marginally lower in spls vs. the Audax, BUT because its a monopole you should have a slightly reduced spl for it, otherwise it will sound "hot". Work toward a slightly downward sloping response from about 2 kHz with the two drivers. Note that because of its vertical directivity it should be near ear level OR should be on a baffle with "tilt" that will "aim" the driver to the listener.

That should give you what you are looking for this time out.

If you need an even cheaper midbass driver solution look to the Selenium 15PW3-SLF. Needs a larger cabinet, has a "hot"er top-end, isn't as eff. and has less linear movement. BUT it should be about half the price. :T

Unfortunately, I have to disagree with nearly everything you stated. Sorry. I don't have time to rebut the points one by one, but I will simply say -

You don't need a real high efficiency woofer to overcome baffle step for this design. There is a general misconception in speaker design about how baffle diffraction operates in this region. The woofers Dan is looking at are more than adequate for achieving his goals.

You do not need a large baffle. My orginal baffle was even smaller than the one shown in the Salk speaker.

He is trying to avoid a large box. You don't necessarily need one. Ours was 1.5 cuft, Dan can tell you if the bass was adequate or not.

Dan got the Audax drivers from me. They are very impressively built midranges, but they failed a critical test in the simulations for me for use in this design, and chose to not used them. Therefore I would not recommend them, or any over several other similar drivers out there. The PHL's met the requirement.

Finally, I have used the HiVi RT1C and I do not consider it a very high quality driver. It has high distortion and in order to have any high frequency extension requires quite a bit of response shaping that gives up a lot of sensitivity. I would not recommend this tweeter at all for an application like this one. The power response would sound very dull due to the narrow high frequency dispersion of the line source. You need a wide dispersion driver instead.

I put a lot of thought and testing into what I was doing here.

Jeff

[ttan98]

Dan got the Audax drivers from me. They are very impressively built midranges, but they failed a critical test in the simulations for me for use in this design, and chose to not used them. Therefore I would not recommend them, or any over several other similar drivers out there. The PHL's met the requirement.


Jeff

Jeff, if don't mind can you tell us the critical test in which the Audax fails? I believe the Audax is a very dynamic mid-range.

[dlneubec]

I can only say that Jeff's Salk speaker was not lacking in anything as far as I'm concerned. The bass was excellent and very clean and the speaker was incredibly dynamic and that was in a room that was probably 40'x100' long. In my living room of 15'x20' I don't think high enough spl or dynamics will be a issue, if I can come at all close to replicating his design. I don't recall any hints of strain in the tweeter output.

A 15" driver is just going to be physically too large to meet my goal of smaller form factor on the overall design (and to obtain SAF :B ). As it is, I have made the bass bin deeper and probably a little taller than in Jeff's and I expect to have in the range of 70-75 net liters available in the box for the woofer. I have yet to determine whether I need that much.

After hearing Jeff's thoughts on the Lambda, I'm covinced they are exceptional drivers and at this point I plan to use them. However, I do have a quesiton in my mind about which one. Jeff used the TD12H, which has a lower Fs and can use a smaller box, but the TD12X is a couple db more sensitive and in a 10l larger box has a very similar extention, though the FS is up about 7hz. Which one is the best to use in this application and which is the best match for the midranges under consideration.

I'm pretty certain I will go with the PHL 1120, since it is a known quantity and very well thought of and almost the same as the driver I heard in Jeff's speaker audition. However, the B&C 6MND44 is very similar in specs, is about $40 ea. cheaper and has the benefit of a small neo magnet that allows for an more open back and potentially closer matching rear FR than the PHL. The PHL has about 18 cm2 more SD and is a little more sensitive, however.

Below is a Unibox graph comparing these 4 drivers. I used a sealed ouput for the mids and I don't know how close that simulates an open baffle output for these mid drivers (after the dipole peak has been attenuated). The TD12H is shown in a 65l box, while the TD12X is shown in a 75L box. All have 100w applied (showing 112-114db at 1m).

I'd like to explore everyones thoughts on these drivers and how they match up. What do you all think?

[Saurav]

They are very impressively built midranges, but they failed a critical test in the simulations for me for use in this design, and chose to not used them.

Jeff, if don't mind can you tell us the critical test in which the Audax fails?

Also, if you measured these drivers and can share the results, that would be very helpful too. I haven't seen any numbers or graphs other than the ones from the manufacturer.

[Scottg]

Unfortunately, I have to disagree with nearly everything you stated. Sorry. I don't have time to rebut the points one by one, but I will simply say -

You don't need a real high efficiency woofer to overcome baffle step for this design. There is a general misconception in speaker design about how baffle diffraction operates in this region. The woofers Dan is looking at are more than adequate for achieving his goals.

You do not need a large baffle. My orginal baffle was even smaller than the one shown in the Salk speaker.

He is trying to avoid a large box. You don't necessarily need one. Ours was 1.5 cuft, Dan can tell you if the bass was adequate or not.

Dan got the Audax drivers from me. They are very impressively built midranges, but they failed a critical test in the simulations for me for use in this design, and chose to not used them. Therefore I would not recommend them, or any over several other similar drivers out there. The PHL's met the requirement.

Finally, I have used the HiVi RT1C and I do not consider it a very high quality driver. It has high distortion and in order to have any high frequency extension requires quite a bit of response shaping that gives up a lot of sensitivity. I would not recommend this tweeter at all for an application like this one. The power response would sound very dull due to the narrow high frequency dispersion of the line source. You need a wide dispersion driver instead.

I put a lot of thought and testing into what I was doing here.

Jeff

I would assume that you had put a lot of thought into your design - it appears to be a very nice speaker.

I also didn't realize Dan was shooting for a clone, I thought it was more about getting something similar or better with much less expensive drivers.. my mistake. ops:

Nor did I realize he was shooting for a small(er) loudspeaker. But I would suggest a read of Hoffman's Iron Law for this particularly imposed restraint.

Though its a moot point now -

The Lambda driver is fairly expensive (not that it isn't worth it), rather that it seems contrary to his original idea of keeping the price low. However IF he wanted a driver to "keep up" with the *audax* without substantial filtering then yes, he would VERY much need a substantially more eff. midbass driver.

Likewise he would also need a relatively large baffle to avoid additional excursion.. and still need a fairly steep high-pass.

Also note that a more eff. driver does in fact make a difference. After all, he could always get an additional +6 db with two normal eff. 8 ohm drivers in parallel - that should be comparable to the TD12M *AND* likely get a lower freq. response near the average. Ah, but it isn't the same is it? Then there is the eff. of the 15w500 - I can assure you that that will not be the same as the TD12M. (..and yes, it will sound better so long as the driver is in a proper enclosure and isn't pushed to extreme spl's.)

As to the Hi Vi driver, sure it has higher non-linear distortion - again, what of it? (Morel isn't exactly high on the list for low non-linear distortion either, but you'll not find me making anything of it.) For the Hi Freq. extension I presume you are referring to the declining response above 10 kHz, True.. But like the non-obvious nature of having a slightly depressed treble response for "conjoining" a monopole tweeter with a dipole mid, the driver does have an excellent horizontal dispersion that when averaged in-room will to most sound somewhat more dominate than your normal tweeter's dispersion. Additionally unless low freq. extension near the average is *incredible*, its typically preferred to have a "touch" less treble as well (..freq. balance an all that). The narrow vertical directivity can actually be a good thing, but as I mentioned before, its critical that it is "aimed" at the listener vertically. Still, I'm not "sold" on this driver - it was just a suggestion and not nearly as poor a one as you seem to indicate.

Again though, its all a moot point - Dan's seems to have made his decision for this design and I more than respect it! :T

[dlneubec]

The decisions are far from all made on this design. I'm pretty certain of using the 12" Lamda drivers for the woofer, but unsure if it will be the TD12H or the TD12X. I have as much as 75L available to work with in the box. I'm still looking for feedback on these two. One of the other reasons I gave up on the Ciare is that I have contected them twice and recieved nothing in return, which leaves me concerned about them as an option. I also have not heard from anyone who has actually heard or used one. I did consider combining two smaller drivers and modelled about everything I could find, but did not find anything that worked from an spl, box size, extension or cost standpoint that was better than a single Lambda's. As mentioned, reducing the apparent mass or size of the speaker is a major goal as well.

For the mids, I know the PHL 1120 will work, but I'm still considering the B&C 6mdn44 as a less expensive and a little less sensitive alternative. I'd like to hear thoughts on these mids and how they match the potential woofers mentioned.

Tweeters are completely up in the air at this point, but if I go with both the Lambda and PHL drivers, the tweeter is the only place left to reduce the cost much, so I will probably lean toward trying to do that.

I hope that clarifies where I'm at with this design.

[Jeff B.]

I would assume that you had put a lot of thought into your design - it appears to be a very nice speaker.

I also didn't realize Dan was shooting for a clone, I thought it was more about getting something similar or better with much less expensive drivers.. my mistake. ops:

Nor did I realize he was shooting for a small(er) loudspeaker. But I would suggest a read of Hoffman's Iron Law for this particularly imposed restraint.

Though its a moot point now -

The Lambda driver is fairly expensive (not that it isn't worth it), rather that it seems contrary to his original idea of keeping the price low. However IF he wanted a driver to "keep up" with the *audax* without substantial filtering then yes, he would VERY much need a substantially more eff. midbass driver.

Likewise he would also need a relatively large baffle to avoid additional excursion.. and still need a fairly steep high-pass.

Also note that a more eff. driver does in fact make a difference. After all, he could always get an additional +6 db with two normal eff. 8 ohm drivers in parallel - that should be comparable to the TD12M *AND* likely get a lower freq. response near the average. Ah, but it isn't the same is it? Then there is the eff. of the 15w500 - I can assure you that that will not be the same as the TD12M. (..and yes, it will sound better so long as the driver is in a proper enclosure and isn't pushed to extreme spl's.)

As to the Hi Vi driver, sure it has higher non-linear distortion - again, what of it? (Morel isn't exactly high on the list for low non-linear distortion either, but you'll not find me making anything of it.) For the Hi Freq. extension I presume you are referring to the declining response above 10 kHz, True.. But like the non-obvious nature of having a slightly depressed treble response for "conjoining" a monopole tweeter with a dipole mid, the driver does have an excellent horizontal dispersion that when averaged in-room will to most sound somewhat more dominate than your normal tweeter's dispersion. Additionally unless low freq. extension near the average is *incredible*, its typically preferred to have a "touch" less treble as well (..freq. balance an all that). The narrow vertical directivity can actually be a good thing, but as I mentioned before, its critical that it is "aimed" at the listener vertically. Still, I'm not "sold" on this driver - it was just a suggestion and not nearly as poor a one as you seem to indicate.

Again though, its all a moot point - Dan's seems to have made his decision for this design and I more than respect it! :T

Many of your ideas will make a fine speaker. It will just be a significantly different speaker than the one Dan listened to.

I enjoy answering questions, but I gave up debating these things a long time ago on the Mad board, so now when it feels like it is becoming a debate I usually move on.

The speaker Dan heard was originally designed and built for my own personal use in my home and then after Jim Salk listened to them they ended up evolving into an even better commercial product. It was displayed this past weekend at the Audio Kharma high-end show in Detroit and received many votes for "Best Sound at Show". This was a nice treat, and Jim receive several orders for the speaker as well.

Dan met Jim and I at the Richmond Indiana gathering and later expressed that he felt the speaker may have been the best speaker he has ever listened to. When he expressed that he would like to approach that same sound I agreed to help him. To accomplish that is a lot more complicated than just selecting some varied drivers and making something that looks similar. There are some very specifical things that make this speaker what it is.

The speaker was not actually an attempt to target highish sensitivity. It was an attempt to make a very musical and accurate loudspeaker from a specific design perspective and it simply turned out to have decent sensitivity.

There are a lot of great woofers out there, some even play well into the midrange. We recently used a JBL 2206 for a similar speaker. The Lambda appears to significantly out-perform the more expensive JBL though. Aside from having a 22Hz Fs and 11mm of Xmax, it has full length Faraday sleeve that reduces Le to .30mH, which is about the level you find in a 4" driver. Consequently, one of the most significant sources of distortion (inductance) is reduced to a very low level. This driver has very low levels of distortion in the midrange for being able to produce good bass at the same time. I am not aware of another driver that approaches this. Like I said, many drivers will work, but will they sound the same?

You are correct, using two smaller drivers in parallel may work very well, and is certainly an option. But chances are, to get the same performance in terms of bass extension and midrange distortion he may pay more money in the end. It is still a good option though.

Tweeter-wise I already said there are other tweeters than can work. I don't think the HiVi is up to the task though. Having used it, I don't think it is a very good tweeter. And yes, non-linear distortion is audible and makes a difference. And no, the Morel is much better than you give it credit for. Mark K. has some measurement on his site that show how this tweeter does. I will give Dan some other recommendations via email.

Your comments about the larger baffle and steep highpass are correct as well. They just result in a very different loudspeaker, and assume some things about this speaker that are incorrect and make a difference in how it sounds, but I won't go into that here, because I don't want to share those details publicly.

Now, as for your first statement above, "something similar or better with much less expensive drivers" - I would certainly be interested in that. I just don't know how to accomplish it within the design goals of this speaker. You are certainly welcome to try though.

Jeff B.

[Mark Seaton]

Hi Dan,

I've been reading along here and there while working on the final version of my Catalyst, which actually uses a pair of 12s that are a custom variation on the TD12 driver. The Lambda drivers always have, and still do, many things right and have a range of parameters and qualities that aren't readily available elsewhere.

You earlier mentioned stereo subwoofers, and then also mentioned size constraints. Is there any reason you aren't designing with the hand-off to the subwoofers in mind from the get-go? You could easily use a TD12X in a sealed box as small as 45L and not trade a whole lot in the working range.

There is a slightly different crossover interaction with the impedance curve of the sealed vs. vented alignment, as any time you start working on a passive crossover anywhere close to the LF corner things get more complicated, especially in terms of total network impedance.

The combo that looks the easiest to execute IMO is a sealed TD15X driver in 50-80L, but obviously the dims of the 15" can make the form more visually imposing, but that's why you stash the subwoofers somewhere less conspicuous.

[Mark Seaton]

Dan met Jim and I at the Richmond Indiana gathering and later expressed that he felt the speaker may have been the best speaker he has ever listened to. When he expressed that he would like to approach that same sound I agreed to help him. To accomplish that is a lot more complicated than just selecting some varied drivers and making something that looks similar. There are some very specifical things that make this speaker what it is.

What! You mean there's more to this speaker design mumbo-jumbo than cool drivers in an interesting looking box?!? :roll:

Congrats on the very well received new design. It's even more of a pleasure when you have someone like Jim to make a design look as cool/elegant as all the effort put into the design.

There are a lot of great woofers out there, some even play well into the midrange. We recently used a JBL 2206 for a similar speaker. The Lambda appears to significantly out-perform the more expensive JBL though. Aside from having a 22Hz Fs and 11mm of Xmax, it has full length Faraday sleeve that reduces Le to .30mH, which is about the level you find in a 4" driver. Consequently, one of the most significant sources of distortion (inductance) is reduced to a very low level. This driver has very low levels of distortion in the midrange for being able to produce good bass at the same time. I am not aware of another driver that approaches this. Like I said, many drivers will work, but will they sound the same?

I mostly responded to this regarding the above comment about distortion. I agree that the drivers have very low audible distortion. To save us some silly arguements from some nitpickers, I'm not sure I would label inductance modulation (vs. position & current) as a most significant source of distortion (% distortion wise), but rather a highly audible source of distortion. In the end it is the stuff we can actually hear that really matters.

[dlneubec]

Hi Dan,

I've been reading along here and there while working on the final version of my Catalyst, which actually uses a pair of 12s that are a custom variation on the TD12 driver. The Lambda drivers always have, and still do, many things right and have a range of parameters and qualities that aren't readily available elsewhere.

You earlier mentioned stereo subwoofers, and then also mentioned size constraints. Is there any reason you aren't designing with the hand-off to the subwoofers in mind from the get-go? You could easily use a TD12X in a sealed box as small as 45L and not trade a whole lot in the working range.

There is a slightly different crossover interaction with the impedance curve of the sealed vs. vented alignment, as any time you start working on a passive crossover anywhere close to the LF corner things get more complicated, especially in terms of total network impedance.

The combo that looks the easiest to execute IMO is a sealed TD15X driver in 50-80L, but obviously the dims of the 15" can make the form more visually imposing, but that's why you stash the subwoofers somewhere less conspicuous.

Hi Mark,

Thanks for the comments and thoughts. My current plan is to restrict the subwoofers to HT use only. The subs are not high quality drivers. They are the higher end Dayton 12" DVC subwoofers in sealed boxes that use an active crossover (designed by John K.) for use with my NaO mini's. They have what some would call an LT circuit to push them lower and are set to crossover to my current mains at 40 hz (my HOSS omni design in my avatar)). That design includeds the Dayton RSS265HF subs in a tapered TL that can cover from about 30hz-180hz.

For music, which represents the majority of use, I would prefer the deeper extension of the ported option for this new design. I agree that a sealed 15" might be ideal, however, when I looked again at the possibility of the 15" I found it would result in just too massive a physical presence in my situation, even if was located at the very bottom of the cabinet, so I have to rule that option out.

Do you have any thoughts about using the TD12H versus the TD12X, given the up to 75L size range avialable?

Similarly, do you have any thoughts on the PHL versus B&C midrange drivers and how they might work with the woofer options?

[Scottg]

And yes, non-linear distortion is audible and makes a difference. And no, the Morel is much better than you give it credit for.

Jeff B.

Not to "harp on this", nor to get you into an argument.. But.

Of course non-linear distortion is audible, But that is not the problem you have specifically posed.

The problem you have stated (in context) is that specifically the HiVi RT1C is high enough in distortion levels that you can hear and it is objectionable under normal conditions. That I *would* dispute. Why?

Because I've tested this myself, and others on the effects of non-linear distortion with normal recordings (a variety of music) - and on average it is EXTREMELY difficult to detect 2nd order distortion below 10%, its also EXTREMELY difficult to detect higher order harmonics below 1-.1% (depending on the harmonic). (..that is NOT the case with test tones however, but that is not what people listen to.)

Any number of studies bear this out as well.

Note that I don't mind it when people are specifically looking for a low non-linear distortion driver. (..of course even then you have to look at how its used in a design at what the results are at differing power levels.) But to put others OFF a particular tweeter for something that (even more than casual) listeners will almost certainly not find objectionable is misleading (..not by intent of course, but in result).

As to the Morel - wow! 8O Did I give it credit (for or against)?

What I posted was:

"(Morel isn't exactly high on the list for low non-linear distortion either, but you'll not find me making anything of it.)"

Morel as a manufacturer generally has more non-linear distortion than competing products - especially similarly priced competing products. This is not specifically regarding the exact tweeter you have chosen. In any event I expressly believe that its non-linear distortion level is pretty much IRRELEVANT.

Better?

Jeff - I can understand that there will be several areas where performance will differ. I can also understand what areas Dan has found very appealing with this design - especially in contrast to his pursuit with radials (which few have done). The real "key" with your design for him (and perhaps your self and others) is very likely to be the gain in efficiency - particularly in the midbass to midrange. That (and many other factors), is why I recommended the drivers that I have. Yes, specifically I do think he can achieve a better sounding, (to him and most others), loudspeaker with the drivers I mentioned following good design and execution. Note though that this does not in any way detract from what you have accomplished. Nor is what I've suggested something that can be done effectively in a smaller form factor - your design considering size/eff./extension is simply elegant. :T

[augerpro]

Jeff, Mark,
Just to echo a question of Dan's, I too would like some clarity on the difference between the various Lambda woofers. The TDxxM is obviously the high sensitivity king and purportedly can cross higher for use in a 2 way, but has lower xmax. The TDxxH is more of woofer by looking at the specs and mass. The one that intrigues me is the TDxxX though. Is this slightly lower sensitivity but higher xmax version of the TDxxM? Or a slightly higher sensitivity version of the TDxxH with slightly less bass extension? Can it run as high as the TDxxM for 2 way use?

The reason I ask is I'd like to get one that compares to the B&C 15NW76 for a project. This project will use the woofer only up to 200Hz or so in U frame, but I would like to use them in 2way some day, which the B&C can do, and want the one that can work well in both projects.

[Dennis H]

The wayback machine's archive of the Lambda site gives some explanations of the differences in the drivers. Some of the specs don't agree with those John published in his forum so I dunno what's right. For example, John says 6mm Xmax for the M versions and 14 mm for the others. The old Lambda site says 3 mm for the M and 10 for the others.

http://web.archive.org/web/20030604071000/lambdacoustics.com/drivers/TDdrivers.html

[Jeff B.]

. To save us some silly arguements from some nitpickers, I'm not sure I would label inductance modulation (vs. position & current) as a most significant source of distortion (% distortion wise), but rather a highly audible source of distortion. In the end it is the stuff we can actually hear that really matters.

"a highly audible source of distortion" Sure, I can agree with that just fine. As long as it keeps the nitpickers away, of course. Thanks for the save there.

[Jeff B.]

I am slow, but I am getting there.

Not to "harp on this", nor to get you into an argument.. But.

Of course non-linear distortion is audible, But that is not the problem you have specifically posed.

The problem you have stated (in context) is that specifically the HiVi RT1C is high enough in distortion levels that you can hear and it is objectionable under normal conditions. That I *would* dispute. Why?

Because I've tested this myself, and others on the effects of non-linear distortion with normal recordings (a variety of music) - and on average it is EXTREMELY difficult to detect 2nd order distortion below 10%, its also EXTREMELY difficult to detect higher order harmonics below 1-.1% (depending on the harmonic). (..that is NOT the case with test tones however, but that is not what people listen to.)

Any number of studies bear this out as well.

Note that I don't mind it when people are specifically looking for a low non-linear distortion driver. (..of course even then you have to look at how its used in a design at what the results are at differing power levels.) But to put others OFF a particular tweeter for something that (even more than casual) listeners will almost certainly not find objectionable is misleading (..not by intent of course, but in result).

As to the Morel - wow! 8O Did I give it credit (for or against)?

What I posted was:

"(Morel isn't exactly high on the list for low non-linear distortion either, but you'll not find me making anything of it.)"

Morel as a manufacturer generally has more non-linear distortion than competing products - especially similarly priced competing products. This is not specifically regarding the exact tweeter you have chosen. In any event I expressly believe that its non-linear distortion level is pretty much IRRELEVANT.

Better?

Jeff - I can understand that there will be several areas where performance will differ. I can also understand what areas Dan has found very appealing with this design - especially in contrast to his pursuit with radials (which few have done). The real "key" with your design for him (and perhaps your self and others) is very likely to be the gain in efficiency - particularly in the midbass to midrange. That (and many other factors), is why I recommended the drivers that I have. Yes, specifically I do think he can achieve a better sounding, (to him and most others), loudspeaker with the drivers I mentioned following good design and execution. Note though that this does not in any way detract from what you have accomplished. Nor is what I've suggested something that can be done effectively in a smaller form factor - your design considering size/eff./extension is simply elegant. :T

First, I think you misunderstood why I was negative about the HIVi tweeter. My reason was a simplistic one - I have used it and I just don't believe it is a very good sounding driver. I think there are countless other options that far exceed it in a number of ways. I am sorry that you think it is wrong to "put others OFF a particular tweeter". Unfortunately, I guess I have the audacity to believe my opinion counts as much as anyone else's does.....which, by the way, is very little. If you like the tweeter, then OK. I don't.

As for the other comments, thanks. I see things much more clearly now. You seem to have tremendous insights into what my speaker sounds like and how to make it better.

[Jeff B.]

Why not the Audax?

Jeff, if don't mind can you tell us the critical test in which the Audax fails? I believe the Audax is a very dynamic mid-range.

Yes, I agree. The Audax is a very robust driver, and a very dynamic midrange. They rate it at 100dB, but in reality once the response is flattened it is only about 95dB, but hey, that's still very high. However, here's why it wouldn't work in my speaker, and it was the same reason for the B&C (at least the one I was going to use).

I made a really cool tool. I built together some aspects of my "Passive Crossover Designer" and my "Woofer and Box Designer" programs. This allowed me to model any driver; woofer, midrange, or tweeter by entering the T/S parameters, including Xmax, and the system parameters and model the T/S based transfer function. Then using the actual measured impedance I model the passive crossover circuit to get the transfer function of the crossover terminated by the impedance. Then I load this transfer function into the T/S model and enter the target power level. This allows me to see exactly how much power and at what frequency the driver will reach Xmax - whether it is a woofer, a midrange, or even a tweeter, based on the crossover it is used with.

The crossover I ended up with for this speaker was fairly demanding on the low end due to the dipole roll-off of the open baffle configuration. The listed Xmax of the Audax is only 0.5mm. My program showed that I reach this amount of travel with only 25 Watts of input. I realized from the simulations that I needed a driver that had at least 2mm of Xmax available if I wanted to remain linear to higher power level and avoid the distortion that comes with driving the coil out of the gap. The PHL was the only driver I could find that gave me the combination of linear, high resolution midrange reproduction, relatively low distortion, fairly decent sensitivity, and high enough excursion to meet demands. The PHL will not exceed its Xmax in this speaker until very high power levels. High enough that it probably won't matter at that point ;-)

Yes, there may be some midbasses sensitive enough to work too, but I found the midrange reproduction of the PHL to be exquisite, so I kept it. It is a very good sounding midrange driver in its own right.

Jeff B.

[dlneubec]

Hi Jeff,

That sounds like a very useful tool! Just to clarify, you need an acutal measurement file of the drivers impedance in order ot make this work, correct? For example, would you be able to input the published TS parameters of a driver, like the B&C 6mdn44 (it does claim 2.5mm xmax, BTW) and do any meaningful comparison, or do you need the driver in hand?

I have seen Zaph do some power to xmax studies of midranges and tweeters using SoundEasy, which I own and use, but I'm not knowlegeable enough with it at this point to do this myself and I'm not sure that what he is doing is comparable to what you are doing.

[exojam]

Dan,

Until Jeff has a chance to get back to you. Below is the link to those tools. They are also listed in FAQ section. Since you have way more knowledge in this area, it will probably not make you cry like a baby as it did me when I tried to tinker around with them.

http://audio.claub.net/software/jbagby.html

[augerpro]

Dan in the newest version of SE in the enclosure designer is new feature that can apply generic filters to the box model. I use it to see the excursion of say, a 6" woofer that is crossed to a sub at 60hz or whatever. This is not precisely the same as Jeff's method I don't think. but it works well. It also helps to visualize how said 6" driver will sum with a woofer using the typical HT reciever bass managements settings.

[dlneubec]

I guess I need to break down and upgrade from V13. :B

[dlneubec]

Thanks Exojam, I'll take a look at that tomorrow.

[ttan98]

Jeff,

Thanks for taking the time to write a detailed explanation about not using Audax mid-range.

[John_E_Janowitz]

The wayback machine's archive of the Lambda site gives some explanations of the differences in the drivers. Some of the specs don't agree with those John published in his forum so I dunno what's right. For example, John says 6mm Xmax for the M versions and 14 mm for the others. The old Lambda site says 3 mm for the M and 10 for the others.

http://web.archive.org/web/200306040...TDdrivers.html

A few things you will see different in the future. We measure parameters using 1/3 of the surround, so you will see the Sd and as a result the 1w/1m efficiency be spec'd a little lower than in the past. Regarding Xmax, Nick previously published just the physical overhang as Xmax. This was back when DUMAX testing became the craze and before the 70% rest Bl point became significant. The TDM for example has 3.3mm physical overhang. It however has a very tall 3/4" thick gap plate. If you look at the BL curve of the driver you have over 6mm of travel before you get to that 70% value. The S, H, and X all have just over 10mm physical overhang and again looking at a BL curve they have over 14mm before the Bl drops to 70%. As these numbers based on BL curves are basically the standard now days, we will be publishing those now.

John

[Scottg]

First, I think you misunderstood why I was negative about the HIVi tweeter. My reason was a simplistic one - I have used it and I just don't believe it is a very good sounding driver. I think there are countless other options that far exceed it in a number of ways. I am sorry that you think it is wrong to "put others OFF a particular tweeter". Unfortunately, I guess I have the audacity to believe my opinion counts as much as anyone else's does.....which, by the way, is very little. If you like the tweeter, then OK. I don't.

As for the other comments, thanks. I see things much more clearly now. You seem to have tremendous insights into what my speaker sounds like and how to make it better.

Ah, you don't "like" the tweeter - I can respect that! (I would have course liked some description of what you don't like.)

Quite defensive on the rest. (..if I'm getting the sarcastic undertone correct.) Tis a shame.

[augerpro]

A few things you will see different in the future. We measure parameters using 1/3 of the surround, so you will see the Sd and as a result the 1w/1m efficiency be spec'd a little lower than in the past. Regarding Xmax, Nick previously published just the physical overhang as Xmax. This was back when DUMAX testing became the craze and before the 70% rest Bl point became significant. The TDM for example has 3.3mm physical overhang. It however has a very tall 3/4" thick gap plate. If you look at the BL curve of the driver you have over 6mm of travel before you get to that 70% value. The S, H, and X all have just over 10mm physical overhang and again looking at a BL curve they have over 14mm before the Bl drops to 70%. As these numbers based on BL curves are basically the standard now days, we will be publishing those now.

John

Have you tested to see what xmax is due to suspension non-linearity Cms? I'm guessing Le(x) is already pretty good.

[Jeff B.]

Ah, you don't "like" the tweeter - I can respect that! (I would have course liked some description of what you don't like.)

Quite defensive on the rest. (..if I'm getting the sarcastic undertone correct.) Tis a shame.

No undertone. But your compliments seemed backhanded to me, as each was accompanied with a criticism of a speaker you have never seen or heard. And, you know very little about the design, because I have been careful to not share too much of that. As I said, quite sincerely, the speaker you propose may sound quite nice, it's just a very different speaker design in enough ways that I don't think we are talking about the same things anymore. I am sure you are a very knowledgeable DIYer and can build a very nice sounding speaker. The problem is on my end because I won't explain everything is a public forum. This design now belongs to Salksound and is a commercial product that he is taking orders for. I learned a lot to get to this point and had reasons for doing some things a little differently than some, but unfortunately, those are some of the things I shouldn't share. I have also been around forums a long time now and I recognize when people want to argue and debate, this is something I want to avoid, but sometimes I get pulled in anyway. I'm sorry, but I seem to see that in your posts. If I have stepped on your toes I apologize. Step the tone down a little, and we'll have more fun, otherwise, I have enjoyed the discussion, but I may be involved in much less of it.

Jeff

[Jeff B.]

Dan,

Until Jeff has a chance to get back to you. Below is the link to those tools. They are also listed in FAQ section. Since you have way more knowledge in this area, it will probably not make you cry like a baby as it did me when I tried to tinker around with them.

http://audio.claub.net/software/jbagby.html

Sorry. I did not mean to make you cry. If you have a specific question you can contact me at jeffb1836@yahoo.com, and I may be able to help.

[Jeff B.]

Hi Jeff,

That sounds like a very useful tool! Just to clarify, you need an acutal measurement file of the drivers impedance in order ot make this work, correct? For example, would you be able to input the published TS parameters of a driver, like the B&C 6mdn44 (it does claim 2.5mm xmax, BTW) and do any meaningful comparison, or do you need the driver in hand?

I have seen Zaph do some power to xmax studies of midranges and tweeters using SoundEasy, which I own and use, but I'm not knowlegeable enough with it at this point to do this myself and I'm not sure that what he is doing is comparable to what you are doing.

I only need the impedance to determine the precise transfer function of the passive elements. Once I have that I just calculate excursion vs frequency for the input voltage across this transfer function combined with the system parameters. You can easily do this now using my PCD and saving the midrange filter response as an frd file. Then model the driver in the Woofer and Box Designer and click the button to import the external filter (and make sure it is turned on with the other button). However, I would be glad to do this myself today for this driver using a simulated impedance (which will be close enough, since it is free air anyway) and I will use the same filter I had in the speaker you heard. I will send you the results later. I can't see any reason why it shouldn't work OK though.

What is your target sensitivity, Dan? There may even be other options too, but this should be a good mid for the design.

For Tweeters you might consider the Peerless HDS as an option.

Jeff

[exojam]

Jeff,

Not a problem. :P I had started to look at designing my own speaker a while ago. I went into big chested and thought I could pull it off. After downloading and trying to go through your program, I left my office and told my wife how ignorant I felt.