Design Concept Study: M8td3

[JonMarsh]

This is probably of limited interest to the general public- what I've basically done is a design study for a three way version of the M8ta, renamed M8td3; d for Duelund, which it's roughly derived from, and 3 for three way. Doh!



For those who don't know what an M8ta is, it's a project i did a few years ago not expecting anyone else to build. Boy, was I wrong. :W

Here's Kris Vardas's-



a set by Mark Andrew,



And of course, the inimitable Colorado Tom...



(a guy who actually knows how to do woodworking, unlike a fumble-fingers like me).

So, one idea I've been kicking around, and this is basically just a proof of concept doing some quick and dirty measurements (well, not that dirty, expect for the lumpy in room bass), to show feasibility and crossover concept for a three way based on this concept with the RS52AN dome midrange, and in a minor twist, the Scanspeak D2608-9130, my current favorite under $100 tweeter. Note that the woofer used is the RS225-4, which was not available when the original M8ta design was tackled; being 4 ohm, the voltage sensitivity is about 3 dB higher, a worthwhile improvement in my opinion for solid state amps. With four dB baffle step compensation, the net system sensitivity for 2.83 VRMS can be about 86 dB with the right inductors (Jantzen C-Core, for example). Some component values still need to be "normalized" to standard available vlues (such as L6 - 2 mH) but I don't see any difficulty there.

This IS a 4 ohm speaker across most of the frequency range. No wimpy HT receivers or vacuum state amplifiers without appropriate impedance taps should apply. I consider this just a proof of concept, but it should need very little dialing in. (provided I haven't screwed up in some inscrutable and unknown manner)

Without further ado, the proposed crossover schematic:



And the calculated SPL response...

[numberoneoppa]

Hey, wait, this wasn't on the list! None-the-less, I look forward to seeing what becomes of this. I hope it's not another [seemingly] 10-year ordeal like the Ardents. =P

[SoundOfNothing]

Jon any thoughts on using the RS28F, Curt seems to really like it. It would keep this design all Dayton (one stop shop) and its cheaper. The HDS is nice, but it can't go as low and is a tad ragged in the top end.

[JonMarsh]

"Ragged on the top end" is a function of whether you use diffraction control or not in taking the measurements, and whether or not you smooth the measurements a lot (not done here).

But it's easy to adapt to the RS28a or f. OTOH, for a three way, going lower doesn't really make any difference or serve any use in a three way speaker. For this Duelund target function with a nominal center frequency of 1500 Hz, the tweeter is nominally crossed at 3500 Hz - which is why the D26 worked fine in the original Modula NeoD CC design. Especially as the D2608 has very low distortion also.



I have design data for the RS28a, and can implement a crossover concept for it which would be compatible with the RS28F, but I don't expect to be popping for an RS28F myself- using a 1-1/8" diameter tweeter in a three way instead of a 1" or 3/4" is not likely to help the off axis performance. So far, I haven't seen any published curves on the RS28F off axis, and the earlier Usher silk tweeter from which the design is based and which I DO have, is not especially good in that regard- not as good as the D2608.



The D2608 is a driver that looks more than fair on paper, but that I think needs to be heard to be heard in a system to be really appreciated- I think Jed and some others would confirm that. In many regards, it's quite competitive with the much more expensive Illuminator series tweeters, especially for a true three way application, or if waveguide loaded. That could even be done in this case- in fact, that was going to be the next concept post when I have time to work on it later this month. Business travel beckons for now.

As I hope I inferred in my first post, this is something I expect mainly to be of interest to those who had built or are in the process of building the M8ta- I'm not building another set myself, so there isn't going to be a big thread on this. Least, I'm not planning on it! :W

[SoundOfNothing]

True forgot about the dome being > 1" on the RS28 which tends to have worse off axis response. I know, the RS52 does go mighty high, hence no need for a low XO for the tweeter; it was just more or less a bonus that it can go lower. Really, the suggestion more or less about cost and convenience by using all Dayton. Sad that the cost of the HDS went up (actually any of the seas/scan tweeters for that matter) over the past 18 months; hopefully the greenback starts gaining value instead of loosing :B

[Jed]

I think Jed and some others would confirm that.

The HDS is one of the best dome tweeters available irregardless of price. The top end dispersion is so good it's almost ribbon like, and you don't have to worry about crossing low. The HDS is fast and detailed and has an immediacy and attack in the treble that is uncanny. It's almost too cheap for some folks to take it seriously, which is unfortunate, but irregardless the Peerless/Scan Speak really nailed it with this one. For Duelund/LR2 slopes, I prefer the Scan Speak Air Circ 6600 mainly for its similar linear and non linear performance but with more extended low end response. The HDS is so good it really has me question the cost/performance ratio if crossing with the good old LR4 targets. Add a waveguide to the mix and that issue is just about eliminated, as we've seen in Jon's updated Modula.

So :T for the HDS.

Lately I've kind-of gone to the darkside with mid/dome ribbon tweeters as my preferred combo.... but I do want to try a larger ribbon and cross really low to an Accuton C90 and see what that gets me (minus the poor vertical dispersion on the top end).

[JonMarsh]

Just curious, why is there so much space between the woofer and tweeter on the Mt8a? I'm guessing you didn't foresee adding a midrange to this at the time you designed it. But, it sounds like a very nice upgrade for those who've built these speakers.

What about one version that uses the Morel MDM55? I know it's more expensive, but maybe It'd be a nice upgrade(?) that would interest some. You probably don't have measure data on that, so maybe that's not possible?

Hi John,

I may be the lone voice crying in the wilderness on this one, but other than the smaller footprint, I don't think the MDM 55 is superior to the RS52AN in any fashion. I know it's pretty popular in DIY circles, probably because of the mounting ease and Morel's reputation in past years. But whether you look at it in terms of frequency extension or non-linear distortion, the Morel doesn't match up. Morel's motors are nothing to write home about for linearity- HiVi spanks them in some categories, like 7" woofers. The RS52 is one of the best dome mids you can buy at any price.

90 dB


The only one I've seen or owned and tested that gives it a run for the money is the hideously expensive (in comparison) Accuton C50-8-44 (at $276).

The driver spacing is a function of having a near zero acoustic offset on the horizontal axis. There was no RS52 then, either, or it might have been a three way!! :B

[JonMarsh]

Lately I've kind-of gone to the darkside with mid/dome ribbon tweeters as my preferred combo.... but I do want to try a larger ribbon and cross really low to an Accuton C90 and see what that gets me (minus the poor vertical dispersion on the top end).

Well, ThomasW wants me to do something like that for him with his RAAL ribbons, maybe this year. It will be interesting to see how that works for you.

Your comments on the HDS echo my feelings about it, too. If someone told me I could only use one tweeter the rest of my life (and I still had to pay for them, not just get them for free!) it would be that one, I think. Though I do like what I'm hearing with the 6640 a lot. But for a real world tweeter at a reasonable price, the HDS is a winner for SS. I think people have to hear it though to really appreciate it, not just bench race it. :W

[kvardas]

Looking forward to this one!

[ThomasW]

The discussion of the MDM55 vs the RS52 that evolved in this thread has been split to it's own thread here....

https://www.htguide.com/forum/showthread.php?t=36183

Let's keep this thread on topic

[Winter]

Mark K has tested both the Scanspeak D2608-9130 (as the Peerless 8109210) and the Dayton RS28F. Surprisingly, per Mark K's frequency response plots on his website the two tweeters have very similar dispersion characteristics.

Mark K's Website:
http://www.audioheuristics.org/default.htm

[Bear]

Jon - For the baffle, would you mount the RS52 dead center between the '225 and the tweeter?

[cjd]

I'd expect it to be far closer to the tweeter. This is one of those designs where a custom faceplate for the mid and tweeter combined would really be nice.

[kvardas]

Without further ado, the proposed crossover schematic:




Jon, do you see much variation in the part values in the crossover schematic presented above to that of the final schematic? Also, do you think a 225-8 woofer and Seas millennium could be used with the RS52? I know you are busy with a bunch of other designs plus work, but thought I would ask anyway.

Thanks

Kris

[JonMarsh]

With the RS225-8, sensitivity will be reduced by 3 dB. I will have that network modeled soon, too, with the Millenium Excel, as many have used that in the M8ta, as I did. The "grind" is on to get the Xtreme's ready for Nothern CA DIY later this month, but I will be getting back to this soon. Bump it later in April... I can always use a reminder.

[kvardas]

M8td3

Checking in with you (Jon) on the m8td3. I know you have been busy.

[gainphile]

Very nice looking loudspeakers!

Is that distance between Tweeter and Woofer a non-issue?

[ThomasW]

The spacing between the woofer and tweeter is dealt with in the crossover.

Jon's life been swallowed by his job, this means no work has been done on any speaker unless Jon's specifically posted about it.

[1Michael]

It's a shame he gets worked like a dog like he does...but then I have been out of work since Nov 2008 so having lots of work isn't always bad...

[JonMarsh]

Haven't gotten back to this yet, other than some thought experiments about how I'm going to hog out the existing cabinets to mount the midranges...

If I had it all to do over again, I'd probably consider a dual 8" configuration with flux capacitors. Or not. Who knows? I will evaluate crossovers for both 8 and 4 ohm versions of the driver, as I have some of the 4 ohm, and it will help the system sensitivity by 3 dB.

[kvardas]

Haven't gotten back to this yet, other than some thought experiments about how I'm going to hog out the existing cabinets to mount the midranges...

.

Does the hole for the RS52 need to go through the entire 4-inch baffle, or can it perhaps be a depth a 1/2-inch beyond depth of the driver and then drill two smaller holes for the wires. If this is okay, then perhaps use a long spiral router bit for the hole. If not, a large diameter forstner bit...Rockler has some that go up to about 3-inches. l

[Bear]

Does the hole for the RS52 need to go through the entire 4-inch baffle, or can it perhaps be a depth a 1/2-inch beyond depth of the driver and then drill two smaller holes for the wires. If this is okay, then perhaps use a long spiral router bit for the hole. If not, a large diameter forstner bit...Rockler has some that go up to about 3-inches. l

The problem with the no-through-hole idea is that you have to turn that entire volume into sawdust in the process. The problem with the through hole is 4" baffles. One could get something like a Blue Rhino TCT hole saw. It will chew-up MDF fairly well, but 4" is pushing it for a hole saw and drill driver.

[knifeinthesink]

Has anyone built this?

I have a feeling it might be awhile before Jon can turn his attention back to this one.