Proposal for a compact three way

[mjb]

I don't know, but every time I see this thread, I think its going to be about something else.

[5th element]

I'd never seen it done either, which is why I did it. It's got me thinking that any time you're pairing up drivers that you should really mount them push-pull, or at least measure to see if they show an appreciable improvement if you do. I always thought the XLS drivers sounded noticeably cleaner push-pull than only a single driver, now I know why.

[Jonasz]

So does this only apply if the drivers are placed next to each other? Will this also work if I turn the drivers in two out of four subs spread around the room?

[5th element]

I wondered that too, but I'd imagine that the point is to have a unified wave-front, where the compression and rarefaction of the moving air molecules is being controlled by the push of one driver and the pull of another with them being close enough to more accurately recreate the correct shape of the waveform as they move.

I sort of imagine this as, if you recall what a perfect sine wave looks like on a scope, with an imbalanced driver, the peak portion looks slightly different to the trough. As there's this difference it lends itself to a distortion mechanic. If you add in the reverse of the first driver, but wire it so its back in phase, then both sides of the sine wave end up equal and that part of the mechanic goes away.

As we're talking waves with very large wavelengths, I would imagine, that if in a room free of room modes, that you could probably get the same benefit, even if the two drivers were some distance from one another. But when the room modes start to affect the way the sound radiates throughout the space in the room, I think the chance that the rearward going motion of one driver could effectively combine with the forwards going motion of another to effectively cancel out, what already are small differences/defects in the waveform, are rather slim. Then again, you don't lose anything at all by trying.

[Jonasz]

Got four of them XLS's running now, however none of them is close to another. Got a few more laying around wich are supposed to be incorporated in the system during the summer (or whenever the boxes will build themselves ) looks like they will be inverted and strategically placed. A pity it's so ugly with reversed drivers though. ops:

Earlier I had SL's Phoenix dipole W-subs in wich the drivers were mounted push-pull, newer tried push-push though. Probably should have to get a feeling for the difference... :unsure:

[5th element]

I actually really like being able to see the rear of the driver

I too have 4 XLS subs about the room. Two are separate and two are together. They were individual stacked boxes 'together'. Now they are properly glued together with the drivers mounted push/pull, with one driver on one side of the cabinet and one on the other for force cancellation. I had no idea how much difference force cancellation would make to how much the cabinet actually feels to vibrate. It's rather fun to see the drivers thrashing around at xmax at 20Hz, but for the cabinet to feel completely dead. The cabinets of the single drivers for comparison visibly shaking. It makes me want to buy another pair and pop them on the rear >.<

[Jonasz]

Hehehe sexy drivers like XLS or Acoustic Elegance with nice rear ends is easy on the eyes, cheap stamped steele chassies not so much... :twisted:

I remember my w-baffles were completely vibration free even with the drivers at xmax, pretty cool feeling!

[5th element]

Force cancellation in an OB is a very good idea IMO. I had the XLS in a H baffle before and those do shake around quite considerably. Linkwitz's move away from the H baffle in his latest Orion was a very good call.

Speaking of AE they have just half released an 8" driver, to get back to where this thread started. It looks to be rather decent, but won't go all that low. 91.5dB for the 8 ohm nominal version. If one picks an EBS alignment you'll get down to around 60Hz in a tiny 10 litre cabinet.

[JonMarsh]

Speaking of other driver possibilities, it's occurred to me that the extended range of the RS52 gives some flexibility for the selection and crossover for the tweeter...

I wonder if I should be considering these, especially in the amorphous core transformer version?

[5th element]

They piqued my interest too a few years ago. If the bigger RAAL shines with a 2.5-3kHz xover, these probably require something along the lines of 5k to work at their best. That's not a lot of design flexibility. If they could work well at 3k it'd be a different matter. I guess you'd need to use the tweeter to see just how capable it is, but with people vastly preferring the big one crossed significantly higher than the manufacturer recommends, it doesn't bode well for the little one liking its advertised lower point either.

[Jonasz]

I've heard the MT version of these and they're crossed at 3,1kHz. Sounds fantastic!



Dennis Murphy crosses them at 2,8kHz in his Philharmonic Audio speakers and as far as I know he has a very good ear!





Jon I think you should give them a try at least!

[Jonasz]

They piqued my interest too a few years ago. If the bigger RAAL shines with a 2.5-3kHz xover, these probably require something along the lines of 5k to work at their best. That's not a lot of design flexibility. If they could work well at 3k it'd be a different matter. I guess you'd need to use the tweeter to see just how capable it is, but with people vastly preferring the big one crossed significantly higher than the manufacturer recommends, it doesn't bode well for the little one liking its advertised lower point either.

I think I will try to lower the crossover considerably just to see what happens!

[JonMarsh]

My understanding is that the transformer and ribbon have been designed for a 4th order acoustical response at 2.8 kHz. This makes it a bit of a stretch for a classic 2 way, unless the woofer has a more extended top end response... an MTM with that high a crossover frequency with 6-1/2" drivers would tend to have a narrow vertical window in the crossover region, to put it mildly. OTOH, I bet Dennis's big guys with the Neo8 on midrange sounds rather nice- shouldn't have any trouble hooking those up with the RAAL's. Nothing wrong with the 8" revelator on bass, either, though that TL cabinet isn't very petite! Interestingly, his mid to woofer crossover isn't much lower than what I figure is reasonable for the RS52, based on my last project.

Of course, everyone knows the OEM small RAAL is supposed to have some secret sauce that makes it even work better, but it's not theoretically available to the DIY market. I believe I have the contacts to get a pair, but then I wouldn't be able to write about them considering the NDA. So, that's a less attractive path to investigate.

OTOH, I have the feeling that a crossover interface with the RS52 should be pretty easy peasy at a reasonable frequency, and perhaps one should let the distortion analysis be the guide as to what really is an easy frequency...

[Paul K.]

The TL cabinet had to be that large for a line about 5-1/2 feet long containing enough volume to achieve an f3 below 30 Hz.
Paul

My understanding is that the transformer and ribbon have been designed for a 4th order acoustical response at 2.8 kHz. This makes it a bit of a stretch for a classic 2 way, unless the woofer has a more extended top end response... an MTM with that high a crossover frequency with 6-1/2" drivers would tend to have a narrow vertical window in the crossover region, to put it mildly. OTOH, I bet Dennis's big guys with the Neo8 on midrange sounds rather nice- shouldn't have any trouble hooking those up with the RAAL's. Nothing wrong with the 8" revelator on bass, either, though that TL cabinet isn't very petite! Interestingly, his mid to woofer crossover isn't much lower than what I figure is reasonable for the RS52, based on my last project.

Of course, everyone knows the OEM small RAAL is supposed to have some secret sauce that makes it even work better, but it's not theoretically available to the DIY market. I believe I have the contacts to get a pair, but then I wouldn't be able to write about them considering the NDA. So, that's a less attractive path to investigate.

OTOH, I have the feeling that a crossover interface with the RS52 should be pretty easy peasy at a reasonable frequency, and perhaps one should let the distortion analysis be the guide as to what really is an easy frequency...

[JonMarsh]

I understand that- it reminds me of a TL line I did for a 10" woofer in the mid 70's, expect that system was taller and shallower. Long lines are big, no question about it!

Now, as this thread proposes to be about a compact 8" 3 way, I suppose a TL would be ruled out! :W

I suspect that Dennis optimized the height of the drivers and the enclosure design in consideration of typical seating height and the vertical dispersion of the Neo8.

[Jed]

Of course, everyone knows the OEM small RAAL is supposed to have some secret sauce that makes it even work better, but it's not theoretically available to the DIY market. I believe I have the contacts to get a pair, but then I wouldn't be able to write about them considering the NDA. So, that's a less attractive path to investigate. .

Jon, I don't believe that the OEM RAAL is much different than the regular versions of the tweeters available to DIY (similar DNA). However, it does not require the foam pads (probably because of the ribbon geometry), and it can be customized for various applications. Also, in my experience, the AM core sounds a lot better than the standard version. At least it does with the OEM version.

[JonMarsh]

Jon, I don't believe that the OEM RAAL is much different than the regular versions of the tweeters available to DIY (similar DNA). However, it does not require the foam pads (probably because of the ribbon geometry), and it can be customized for various applications. Also, in my experience, the AM core sounds a lot better than the standard version. At least it does with the OEM version.

I wrote Madisound asking what the deal is with the AM core because the spec sheet was not very clear, and both talked about 8 ohm input impedance and using the amporphous core gapped with drive from a SET amp, where the transformer primary is carrying the plate current. The answer I got was vague and not convincing, that it was only for the SET application. I'm familiar with powdered amorphous alloys, and would expect a core using them to have higher flux density capability and less distortion at low frequencies, while coming close to retaining the HF properties of ferrite. Thanks for the feedback and heads up!

[Paul K.]

You mean sort of like this?


This design, Centilana, used the RS180-4, RS52 and XT19.
Paul

Now, as this thread proposes to be about a compact 8" 3 way, I suppose a TL would be ruled out! :W

[JonMarsh]

That's a very nice example; now what would happen if we just scaled that up to an 8" woofer, and investigated a range of tweeter options? Nice job building the crossover into the base...




And nice packaging of a 60" single fold MLQWTL or whatever it's being called these days! Any response measurements of the system? Looks very nice! :T

I bought a bunch of Martin's MathCAD files some years ago; I should go back and look at them again...

[Paul K.]

Thanks, Jon. You sure have a good memory and were able to dig up the crossover photo. I often locate the crossover in the base like that, or I build it on top of the base assembly with the base attached to the bottom of the cabinet with 1/4-20 bolts and T-nuts. The design is an ML-TL, not an ML-TQWT, because there's no tapering. I can't find any response measurements but I am attaching the modeled system bass response.



Paul

That's a very nice example; now what would happen if we just scaled that up to an 8" woofer, and investigated a range of tweeter options? Nice job building the crossover into the base...




And nice packaging of a 60" single fold MLQWTL or whatever it's being called these days! Any response measurements of the system? Looks very nice! :T

I bought a bunch of Martin's MathCAD files some years ago; I should go back and look at them again...

[Paul K.]

I manage to dig up a measurement of the system response that was made at DIY Iowa 2008. I had a printout that I scanned in. The quality of this scan is not great but I hope it's adequate.
Paul

[kevinm]

Hey ET,

Is this designed aimed to be used with a sub in a 2.1 system?

I have a friend that's looking for a 3 way that can be used without a sub and is having trouble finding a subwoofer/woofer to handle the low frequency duties.

[Evil Twin]

Interesting contributions and inputs...

One would not usually expect extended low frequency response in a "compact" system. Naturally this is a function of Hoffman's Iron Law, which guides a reasonable understanding of the tradeoffs- there is no free lunch, and no substitute for Sd and Xmax.

With the SW223BD01, a relatively compact system is possible with extended LF response, but at low sensitivity. Amplifier power is relatively inexpensive, unless you like SET...

With the 8" Revelator, Xmax is still fairly good, but the driver parameters require a larger enclosure for low frequency extension, due to the driver parameters, mainly cone mass and suspension compliance.

Then, there is the matter of SPL- two 8" drivers would never match the quantity of low frequency output that a long travel 15" sub is capable of- but with the right drivers in some room arrangement, the results may be satisfactory in 2.0, unless you are seeking the ultimate in LF output for movie effects- music is generally a different and less demanding matter.

So, it is a wise man that understands and qualifies his true needs before planning a system build. The intent for this system is not to construct a technological terror, but to achieve satisfying musical performance in a moderate size room- somewhat extended low frequency response would also allow some flexibility in choosing crossover frequencies for the addition of an LFE subwoofer.

It may well be that a few variations of this should be considered- while retaining a core midrange approach, perhaps... evaluating several woofer and tweeter combinations, but keeping the RS52 on midrange duties might be interesting... very interesting.

[Jed]

I wrote Madisound asking what the deal is with the AM core because the spec sheet was not very clear, and both talked about 8 ohm input impedance and using the amporphous core gapped with drive from a SET amp, where the transformer primary is carrying the plate current. The answer I got was vague and not convincing, that it was only for the SET application. I'm familiar with powdered amorphous alloys, and would expect a core using them to have higher flux density capability and less distortion at low frequencies, while coming close to retaining the HF properties of ferrite. Thanks for the feedback and heads up!

I'd put in an email to RAAL to answer the technical questions, but the AM core is one option, as is the Direct Drive AM core transformer for SET amps (sounds like Madisound was attempting to describe this version). You can also get silver/gold transformer windings. In conclusion, there are many options and applications, not just for SET amps. I'm not sure which of the above Madisound is carrying, because I get mine directly from RAAL built to what I need.

[JonMarsh]

I'd put in an email to RAAL to answer the technical questions, but the AM core is one option, as is the Direct Drive AM core transformer for SET amps (sounds like Madisound was attempting to describe this version). You can also get silver/gold transformer windings. In conclusion, there are many options and applications, not just for SET amps. I'm not sure which of the above Madisound is carrying, because I get mine directly from RAAL built to what I need.

Ah hah.... interesting. And a very good suggestion. I'll follow up with RAAL direct. Thanks for the suggestion, Jed.

[Evil Twin]

Well, after some recent discussion, another possibility has become apparent, though to realize it would mandate changing the name of the thread- Jon Marsh has been "persuaded" that his M8ta cabinets are not the enclosures he's looking for, and will relinquish them into my custody for possible rough prototype evaluations.

[JonMarsh]

Whatever helps the cause... I here these might be fodder for a wedding present for an up and coming Sith apprentice and his princess to be...

[dwk]

I'm glad this idea hasn't completely fallen off the radar.

This caused me to go back and read some of the M8ta threads. Engage time machine! It was a pretty cool read. Interesting to see that turning it into a 3-way with the RS52 was an idea that came up just after the driver was released. (good grief - it's actually been SEVEN years. zoiks). I do have to say that your view of budgetary constraints has evolved a bit over the years :-)

It was also a bit of a sobering read - almost 2 years or so for you to finish the design, and for those that built it 8 months to a year seems typical. (and I guess this means you never did get around to actually veneering the originals)

Are you still thinking of going sealed with the Wavecors or similar, or is it more of a revival of the 3-way M8ta concept?

I'm still hoping to be able to target an Avalon style floor stander as a long term project, but since this is for the living room I have to get the wife on board. So far, she's liked the Martin Logan EM-ESLs the best of any ideas I've floated (go figure). Not much DIY opportunity going that route.
The short-term approach is going to be to cobble together a decent setup from existing pieces. A mini-dsp nano-digi along with my ProAc SuperTablettes and my existing bass modules (RSS-210HFs sealed in the 1 cu ft MTM cabs). Looking to cross at 200-300, assuming the RSS holds up. This should hopefully give a couple years of breathing room to consider the long-ish term approach. An evolution of the M8ta into a 3-way might fit that pretty well.

[5th element]

With Jon, nothing ever really falls of the radar, time on the other hand seems to break the laws of physics and disappear faster then it's supposed to.

[dwk]

With Jon, nothing ever really falls of the radar, time on the other hand seems to break the laws of physics and disappear faster then it's supposed to.

True enough, particularly since this can almost be seen as revisiting the 3-way w/ RS52 idea from years ago. I should have said something like "with work pressures and all the time on the Ardents, I'm glad it hasn't (yet) been pushed out to some indeterminate future opportunity".

[Evil Twin]

With Jon, nothing ever really falls of the radar, time on the other hand seems to break the laws of physics and disappear faster then it's supposed to.

When immersed in the Dark side, conventional viewpoints of time and progression are seen for the mundane perspective they are... a true Sith must take the long view, and be willing to plan, wait, and develop until the unfolding in the Force shows the time is right...

As my mentor would say, everything is unfolding as I have foreseen...

On the other hand, sometimes disturbances in the Force lead to unanticipated events and opportunities- the very rapid progression of events with the so-called "Wavecor Ardents" are a clear example of how the Force and long term trends may lead to unexpected serendipity, which should not be ignored lest potentialities slip away...

I do have to say that your view of budgetary constraints has evolved a bit over the years :-)

Obviously, perspective is different from different viewpoints... and a certain build might be considered "low budget" simply because the pieces happen to be lying around at hand, purchased previously for test or investigation or another project concept that has since been put on hold or discarded... I do believe in the same methodology a certain computer and mobility company in Cupertino CA espouses, of saying "no" to a great many more things than are ultimately given a "yes", even though they have promise...


In this light, a budget build with parts on hand might include the RS52 midrange, the Scanspeak revelator 22w/8857t-00-8-woofer-aluminum-cone, and the Scanspeak D2904/71000, all of which are on hand and not spoken for yet. Another woofer alternative is the W22, which also are on hand, but which may not offer the same level of overall performance as the SS. They will be testing in standard enclosures along with the RS225 and new SB Acoustics before final choices are made- they're all here, so the incremental cost is identical.

The M8ta cabinets are in very clean shape, and are partially veneered in Walnut. This could be completed, or a complete re-veneer may be in the works. Further consultation and evaluation of the Wavecor Ardent results will guide this decision.

[Paul K.]

Speaking of the SS 22W/8857T in a 3-way with the RS52 reminds me of my 3-way from 2007 using the very same woofer in a folded ML-TL with an MDM55 and NeoCD1.0, the canTiLena.



Paul

When immersed in the Dark side, conventional viewpoints of time and progression are seen for the mundane perspective they are... a true Sith must take the long view, and be willing to plan, wait, and develop until the unfolding in the Force shows the time is right...

As my mentor would say, everything is unfolding as I have foreseen...

On the other hand, sometimes disturbances in the Force lead to unanticipated events and opportunities- the very rapid progression of events with the so-called "Wavecor Ardents" are a clear example of how the Force and long term trends may lead to unexpected serendipity, which should not be ignored lest potentialities slip away...



Obviously, perspective is different from different viewpoints... and a certain build might be considered "low budget" simply because the pieces happen to be lying around at hand, purchased previously for test or investigation or another project concept that has since been put on hold or discarded... I do believe in the same methodology a certain computer and mobility company in Cupertino CA espouses, of saying "no" to a great many more things than are ultimately given a "yes", even though they have promise...


In this light, a budget build with parts on hand might include the RS52 midrange, the Scanspeak revelator 22w/8857t-00-8-woofer-aluminum-cone, and the Scanspeak D2904/71000, all of which are on hand and not spoken for yet. Another woofer alternative is the W22, which also are on hand, but which may not offer the same level of overall performance as the SS. They will be testing in standard enclosures along with the RS225 and new SB Acoustics before final choices are made- they're all here, so the incremental cost is identical.

The M8ta cabinets are in very clean shape, and are partially veneered in Walnut. This could be completed, or a complete re-veneer may be in the works. Further consultation and evaluation of the Wavecor Ardent results will guide this decision.

[JonMarsh]

How do you like the canTiLena compared with your Amarosos ML TL?



I see you used the Neo8 in the Amarosos; a lot of guys have been doing stuff with the Neo10 of late, but it seems to me like there is some considerable variability in performance of that planar- any thoughts or experience from your side?

[Paul K.]

To be honest, Jon, I have yet to do a side-by-side comparison between the two. The Amarosos have remained in my living room since they were finished and the canTiLenas are in the basement (covered so they don't get dusty). I love the midrange sound, especially on violins, from the Amarosos, and even though the SS woofer has more capability than the RS225, I don't play any speaker loud enough to be able to hear that difference I don't think. IIRC I've only heard the Neo10s on one speaker, Dan Neubecker's Echelons (I think that's the correct name) from his win at MWAF 2012. Dennis Murphy sure likes the Neo8 and I didn't have room for the larger Neo10, so I settled on the Neo8 (and I have no input or experience on variability in the Neo10).
Paul

How do you like the canTiLena compared with your Amarosos ML TL?

I see you used the Neo8 in the Amarosos; a lot of guys have been doing stuff with the Neo10 of late, but it seems to me like there is some considerable variability in performance of that planar- any thoughts or experience from your side?

[Spanky Ham]

Consider me very interested in this build. I remember the original Modula Neo that you said was one of the best sounding speakers you have heard.

[Squidspeak]

Speaking of the SS 22W/8857T in a 3-way with the RS52 reminds me of my 3-way from 2007 using the very same woofer in a folded ML-TL with an MDM55 and NeoCD1.0, the canTiLena.
Paul

Paul do you have a cabinet drawing of the interior layout of this design,just interested in your
different approches to TL design Thanks Mike

[Paul K.]

Yes I do but what I've attached may not be all that great since it's a scan of a pencil on paper drawing. This is a single-fold ML-TL, starting at the bottom of the front, rising to and making a U-turn at the top, then going down the back to the bottom. The line is 76.5" long, the woofer's center is located 23.75" from the beginning of the line, and the mass-loading port's center is located 13" from the end of the line. The first half of the line is stuffed at a density of 0.8 lb/ft3 with poly fiber.



Paul

Paul do you have a cabinet drawing of the interior layout of this design,just interested in your
different approches to TL design Thanks Mike