Planar Speaker Asylum

Welcome! Need support, you got it. Or share your ideas and experiences.

Return to Planar Speaker Asylum

Message Sort: Post Order or Asylum Reverse Threaded

DIY speaker design comments

91.152.9.226
Posted on March 29, 2014 at 03:59:16
AkuAnkka
Audiophile

Posts: 167
Joined: February 3, 2013



What would you think, which one of the two options in a picture is a better choice for a DIY setup; A on the left or B on the right?

DEQX HDP-3 will be used as a 3-way crossover, and subs & midbass would separated by subs own crossover, so this would be 4- way setup. XO frequencies would be about 100hz, 400hz and 4khz, but this of course has to be tested by listening.

GR12's would be in an OB box such as: http://www.gr-research.com/pdf/obsub.pdf. All other drivers would naturally be in a I- baffle just as maggies are.

My understanding of their benefits are:
A)
- since mids & highs are on a narrower frame, there should be less frame distortions for them if frames are not positioned side by side
- since mids & highs are separated from mid basses, there would be less distortion caused by it's movement

B)
- larger midbass driver -> maybe a bit better midbasses
- option to use even 3 GR12's per side -> maybe a bit better sub basses
- better integration between midbasses and mids, since drivers are closer to each other and are about the same height

So bacically I think that A *could* be better for mids & highs, and B surely could be better for basses. Any opinions on this?

 

Hide full thread outline!
    ...
RE: DIY speaker design comments, posted on March 29, 2014 at 05:03:01
JLindborg
Audiophile

Posts: 1037
Location: Uppsala
Joined: April 26, 2010
 If You ARE serous about sound, then You should NEVER EVER use an MMG for anything as they are made as 2 different speakers that are tuned totally different from each other.
It's like having an 8 inch woofer on the right MMG and a 10 inch woofer on the left MMG and together they ought to sound as a sum but it does not.

Go with the other design and leave the MMG for the beginners and kids.

Cheers!


The one who succeeded was the one who didn't know it was impossible.

 

RE: DIY speaker design comments, posted on March 29, 2014 at 05:11:30
AkuAnkka
Audiophile

Posts: 167
Joined: February 3, 2013
 Thanks J for the comment, good to know. Do you know if older SMG-x is any better? Or any other smaller Maggie?

 

RE: DIY speaker design comments, posted on March 29, 2014 at 05:35:07
JLindborg
Audiophile

Posts: 1037
Location: Uppsala
Joined: April 26, 2010
 I am not sure about that.
Off with the socks or take a bright light and look through the fabric to see if the tuning buttons are symmetrically placed between left and right speakers.
If not, then You will NOT have a symmetrical sound from them. I will be unbalanced.

Cheers!


The one who succeeded was the one who didn't know it was impossible.

 

RE: DIY speaker design comments, posted on March 29, 2014 at 06:16:06
Roger Gustavsson
Audiophile

Posts: 2054
Location: Huskvarna
Joined: February 12, 2010
 Yes, the bass of the 2.5 is a better choice. Maybe 8 mid drivers could be even better? I would build the subwoofers in separate enclosures. It use to be better to have the bass coming from more spots. Some useful reading on multiple basses:
http://www.harman.com/EN-US/OurCompany/Innovation/Pages/WhitePapers.aspx

 

RE: DIY speaker design comments, posted on March 29, 2014 at 06:36:02
Davey
 Total nonsense.

If you're going to declare Maggie speakers which utilize the different tuning button locations as non-serious, then you better not stop at MMG's. :)

Dave.

 

RE: DIY speaker design comments, posted on March 29, 2014 at 06:52:19
JLindborg
Audiophile

Posts: 1037
Location: Uppsala
Joined: April 26, 2010
 I am not really sure that I understand what You are talking about.....

I actually know a great deal how different planars work and why.
I have never ever said that tuning buttons and its location is non-serious.
They are EXTREMELY serious!!!

What I HAVE said is that if You give different tuning on left and right, You WILL have 2 different sounding speakers. One character on the left and one character on the right.

If one does not care about this then go ahead and buy it (if You feel that You have the money to throw at it).
The one who succeeded was the one who didn't know it was impossible.

 

RE: DIY speaker design comments, posted on March 29, 2014 at 07:05:31
louie3
 Don't know if this is any help, but the tuning discs on my SMGcs are in the same place on both panels...about 8 inches from the top and bottom on both.

(No sox fer me.)

 

RE: DIY speaker design comments, posted on March 29, 2014 at 07:21:18
Davey
 What's different about left and right MMG's? Only the tuning button locations, yes?
You just got through saying that "if you are serous about sound you shouldn't use MMG's because they are two different speakers." Did I extrapolate incorrectly?

The tuning is not totally different, it's slightly different. The (multiple) resonances from both left and right actually overlap each other with this configuration.

Your 8 inch and 10 inch woofer analogy is piss-poor. Here's another poor one, but it's much better. The MMG's are like two identical 10 inch woofers except one has a small piece of chewing gum attached to the cone. This changes the bass tuning frequency just slightly, but differences are inaudible unless program material just happens to line up (very rarely does this happen) with a resonance of either panel for a significant time. For non-bass frequencies the tuning difference is a non-issue.

It's difficult to blind test this difference because you can't exactly co-locate the two speakers, but here's a fairly worthwhile test. Put your MMG's near the center of the room and play pink noise alternating between the two speakers. Have an assistant randomly switch locations of the two speakers and see if you can identify either speaker consistently in an un-sighted test.

Regardless, if MMG's are beginners/kids speakers because of this design aspect then other models are as well. Would you agree?

Cheers,

Dave.

 

RE: DIY speaker design comments, posted on March 29, 2014 at 07:46:06
AkuAnkka
Audiophile

Posts: 167
Joined: February 3, 2013
 Thanks Roger, I'm sure 2.5 driver version would give better basses.

But, how about the mids & highs; I *think* that model A would be a better choice due narrower frame and less frame distortions, but what's your opinion on this? Should the difference be notable?

 

Hey Dave, tell him why they exist., posted on March 29, 2014 at 08:05:07
wazoo
Audiophile

Posts: 4062
Location: Middle GA
Joined: December 6, 2006
 The whole point is to prevent the accentuation of combining identical L/R speaker resonances, which can get ugly. The asymmetrical tuning buttons aren't a bane; they're an absolute boon. The result is a smoother combined frequency response. Frankly, I think it's a stroke of engineering brilliance.

 

RE: Hey Dave, tell him why they exist., posted on March 29, 2014 at 08:41:20
Davey



For those that are interested, this is a good illustration of the tuning dot scheme for MMG's. (Green speaker 1, yellow speaker 2.)

It's easy to see the two panel resonances of the single dot speaker transducer (green) and the four panel resonances of the three dot transducer (yellow.)

Cheers,

Dave.




 

RE: DIY speaker design comments, posted on March 29, 2014 at 08:59:29
hemholtz
Audiophile

Posts: 1157
Location: Indiana
Joined: February 21, 2003
 i did a similar project with rd50's and tympani bass panels. i cross the gr-research subs to the first bass panel at 40hz 24 db per octave. after that it's very similar to what you describe. overall, the top sounds about the same as deqx'd 3.xr's. they seem to play louder though.


keep the subs separate. they'll shake the hell out of everything. i have 4 in a large room and it seems to be plenty. the newer hx series amps (hypex module based)look interesting, but 2 a370's seems plenty for 4 ob subs on my system. don't know how 3 woofers per channel would work out electrically. if you run the low output from the deqx through the sub amp then to the midwoofer amp, you can still use the deqx parametric eq for both the sub and mid woofer. i would hook it up that way, then when you make your correction filters you would just need to set the sub output outside the frequency limit of the correction and just eq it by hand (ear?)




build the ob sub "box" plenty heavy. those are some nice sounding woofers for the very bottom. not sure what they would do up to 100hz though. one of these days i should try it just for fun.

 

excellent illustration, posted on March 29, 2014 at 09:27:40
wazoo
Audiophile

Posts: 4062
Location: Middle GA
Joined: December 6, 2006
 I don't remember seeing that before. If that doesn't convince us of the judiciousness of the asymmetrical tuning dots, I don't know what can! It's even more critical when one considers that most typical rooms have modal issues in the neighborhood of 70Hz.

 

RE: DIY speaker design comments, posted on March 29, 2014 at 10:13:52
Roger Gustavsson
Audiophile

Posts: 2054
Location: Huskvarna
Joined: February 12, 2010
 AkuAnkka wrote:
"Thanks Roger, I'm sure 2.5 driver version would give better basses.

But, how about the mids & highs; I *think* that model A would be a better choice due narrower frame and less frame distortions, but what's your opinion on this? Should the difference be notable?"

Well, the line of Neo 8 need some baffle width. The one with the 2.5 bass driver is very much like a 3-series Magnepan. No, it is better! The separate midrange is really an upgrade. The 2.5 bass driver is slightly more "massive" than a Tympani mid bass panel, maybe the Neo 8 will put a light on this, a discontinuity? On the other hand, there will probably be some noticable step from the 2.5 basses to your servo subs. So far, I have always prefered no subs.

 

RE: excellent illustration, posted on March 29, 2014 at 10:24:50
Davey



I got all kinds of 'em.

Here's one that shows woofer element coupling onto the tweeter. The woofer being actively driven but the tweeter being monitored with an analyzer.
A decent indication of the "mechanical" crossover between electrical elements on the same transducer surface. It's sort of like a psuedo IM distortion measurement.

Dave.

 

RE: DIY speaker design comments, posted on March 29, 2014 at 11:12:49
AkuAnkka
Audiophile

Posts: 167
Joined: February 3, 2013
 Well it really should be way better than 3- series Magnepan; if it would not, I would just buy 3.6's ;=)

2.5 has a bigger bass than 3.x, and NEO8 should easily be better than 3.x's mid range. And I on my own experience; if there is a sub that matches Maggies well, it is GR servos on OB. I have tried dozens of subs, and nothing comes even close, IMO.

 

RE: DIY speaker design comments, posted on March 29, 2014 at 11:15:53
Satie
Audiophile

Posts: 5426
Joined: July 6, 2002
 At 400 hz or so XO to the 2.5 midbass you don't need to worry much about baffle loading of the Neo8 line. It will only make a small difference. The 2.5 itself would make enough of a baffle anyway.

The MMG choice is a nono in my book - height alignment and IMD from the subs makes it a bad idea.

The 2.5 midbass should be acoustically isolated by damping the space between the frame and the driver frame. I would do it with Vmax or equivalent on both surfaces (wood frame and driver frame) and add a thin viscoelastic foam in between, then press the whole shebang together before fixing the midbass panel to the frame- done so the compression is retained. Alternately, put the 2.5 midbass in a physically separate sub-frame and use AndyR's method to isolate it from the rest of the frame.

 

RE: DIY speaker design comments, posted on March 29, 2014 at 11:15:55
AkuAnkka
Audiophile

Posts: 167
Joined: February 3, 2013
 Thanks helmholz, sounds very interesting.

How would you describe RD50's? I also thought about using it, but it is quite often said that NEO8's should be better on midrange, and 3.x ribbons should be better on highs. But I have never heard any RD's, so I really don't know.



 

RE: DIY speaker design comments, posted on March 29, 2014 at 11:34:53
AkuAnkka
Audiophile

Posts: 167
Joined: February 3, 2013
 Satie, wouldn't it actually be better to screw the 2.5 bass driver to the heavy wooden frame as tight as possible, and then separate the mids & high acoustically from the frame?

 

RE: DIY speaker design comments, posted on March 29, 2014 at 14:14:42
Satie
Audiophile

Posts: 5426
Joined: July 6, 2002
 That might be easier to do and provide you with some mass damping - so long as you get intimate contact between driver and frame.

Which ever way, the midbass is going to do allot of work - so whatever scheme you come up with should separate the mid and tweeter from that high energy.

 

That's how I made my Frankenpans ..., posted on March 29, 2014 at 23:19:13
andyr
Manufacturer

Posts: 12548
Location: Melbourne
Joined: September 2, 2000
 ... which combine a pair of 2.5 bass panels with a pair of T-IVa mid/ribbon panels.

Bass panel is screwed to a separate, braced, hardwood frame from the mid panel & ribbon - so there is minimal vibration transmission between them. When the bass is pumping, you can touch the top of the bass frame and feel the vibrations; touch the mid/ribbon frame ... and there is no vibration. :-))


Regards,

Andy

PS: Next on the plan is to replace the T-IVa mid panel with a string of Neo 8s.

 

RE: That's how I made my Frankenpans ..., posted on March 30, 2014 at 01:52:01
AkuAnkka
Audiophile

Posts: 167
Joined: February 3, 2013
 Ok, thank you very much for all of you for your comments! It seems to be quite clear that plan B with separate subs is a better choice.

One more question; both the ribbon & 2.5 basses are about 1,5m high. 6 x NEO8 would "only" be about 1,2m, so mid area would be a bit shorter than both basses & highs. Does that actually matter? Should I go for 7 or 8 NEO's, or would that just make the wiring too difficult..?

 

RE: That's how I made my Frankenpans ..., posted on March 30, 2014 at 06:58:12
Roger Gustavsson
Audiophile

Posts: 2054
Location: Huskvarna
Joined: February 12, 2010
 AkuAnkka wrote:
"2.5 has a bigger bass than 3.x"

I am not sure if that is really true. The 3-series uses the full width of the bas+mid driver at the bottom. This is also a reason to why the mid can be polluted by the bass while it is doing heavy work.

About the number of Neo 8. To me it would feel better with 8 drivers in a line but it will probably not be of great importance soundwice. In theory, a long line source is more of a true line source. In this case it would better match the Magnepan ribbon tweeter. The mids of MG 20-series are all shorter than the ribbon tweeter.

 

RE: That's how I made my Frankenpans ..., posted on March 30, 2014 at 07:35:43
AkuAnkka
Audiophile

Posts: 167
Joined: February 3, 2013
 But how long is actually long enough to be "true line source", if the range is from 400hz to 4khz?

And, could it actually be possible to leave ~5cm gap between the NEO's, to have 145cm length with just 6 NEO's?

 

RE: That's how I made my Frankenpans ..., posted on March 30, 2014 at 07:42:30
Roger Gustavsson
Audiophile

Posts: 2054
Location: Huskvarna
Joined: February 12, 2010
 To be honest, 1.2 m is too short for tweeters but maybe you will realize that it is okey for the midrange?

 

RE: That's how I made my Frankenpans ..., posted on March 30, 2014 at 12:06:20
Satie
Audiophile

Posts: 5426
Joined: July 6, 2002
 You don't want to introduce greater gaps in the Neo8 line if you want to operate them to 4khz. There is already a gap in the radiating surface of about 3" for the driver flanges etc. That will start having measurable comb filtering effects above 5khz - but not yet audible since the direct output of the drivers is extremely directional with hardly any vertical dispersion. If you add another 2" to the gaps then it will become audible.

The fact of the output being so directional is what allows the line array to cancel out the resonance peak at 12khz. Since the resonance output does have a vertical component - it is not direct output, but reflected from the magnet board to the diaphragm and back at least once before exiting the driver.

 

RE: DIY speaker design comments, posted on March 30, 2014 at 12:55:40
russ69
Audiophile

Posts: 951
Joined: December 13, 2009
 I'm not going to crap on your design. I think it's a worthwhile project. However if it was me, I'd get the high/mid panels sorted out then work on the lower end. I have a set of Infinity woofer columns that I think work really great. If I was to envision an improvement I think a vertical array of 4 or 6 eight inch drivers per side would do the trick. Many people criticize the Infinity and Genesis woofer designs but the things they do right is way ahead of the things they do wrong.

 

RE: DIY speaker design comments, posted on March 30, 2014 at 20:42:23
Satie
Audiophile

Posts: 5426
Joined: July 6, 2002





Well here is the RD75 waterfall plot and the same for the Neo8 from the OEM and from an unknown DIYer

The scales are different, so don't be taken aback by the appearance of the RD75 plot. It actually goes down 30 db below the signal level, while the Neo8 values are cut off at -18 db or so.

Using the -18db cutoff for the RD75, the floor would be hit at about 1.4ms from 200hz to 2+khz and after the 2khz resonance the driver is damped -18db within 1ms to about 8-9khz, then finally 0.7ms for the top octave.
This is still way longer than the Neo8 which clears up to -18db at 0.4 ms. It is simply the faster driver.

 

RE: DIY speaker design comments, posted on March 30, 2014 at 21:18:03
hemholtz
Audiophile

Posts: 1157
Location: Indiana
Joined: February 21, 2003
 i don't get what you're implying with the decay measurements. milliseconds are room acoustics to me not driver limitations. are you saying the rd's ring? as far as i know, the construction is similar. i would question the amp driving them if there were an audible difference.

 

RE: DIY speaker design comments, posted on March 30, 2014 at 21:50:50
Satie
Audiophile

Posts: 5426
Joined: July 6, 2002
 Its an energy storage issue, you get similar slowdown in decay by horn loading the Neo8.

We had a discussion about segmented drivers vs. long planars after the new Kingsound review from Doug Schroeder (sp). Part of it is a measurement problem where the aperature needs to be small to block incoming waves from above and below the measuring point at a delay. But I believe this was done in this case.

What I still didn't find is the THD vs freq curves for the RD series like this one from Zaph for the Neo3PDR and his THd sweeps. I had those on a computer that died before the scheduled backup, and they have been pulled from the net so I can't replace them.

 

RE: That's how I made my Frankenpans ..., posted on March 31, 2014 at 10:06:19
AkuAnkka
Audiophile

Posts: 167
Joined: February 3, 2013
 Ok, thanks. 7 would actually be enough, but is it anyhow possible to wire 7 NEO8's to get a reasonable impedance?

 

RE: That's how I made my Frankenpans ..., posted on March 31, 2014 at 10:43:44
Satie
Audiophile

Posts: 5426
Joined: July 6, 2002
 7.5 ohms with 3 parallel pairs and one parallel set of Neo8 and 3.75 ohm resistor.

Quite frankly I would not bother with a single additional driver. I would go from 6 to 8 or stick with 6.

 

RE: DIY speaker design comments, posted on March 31, 2014 at 18:59:59
EmailTim
 Satie,

Have you tried the Way-Back-Machine? If you know the website, it may have been cached on the Way-Back-Machine.

HTH,
Tim

 

RE: DIY speaker design comments, posted on March 31, 2014 at 19:38:45
Satie
Audiophile

Posts: 5426
Joined: July 6, 2002
 Thanks
Forgot about them
will give it a try

 

Page processed in 0.037 seconds.

[ Top Top of Page ] [ Contact Us ] [ Support/Wish List ] [ Copyright © 2024, Audio Asylum, all rights reserved ]
[ General ] [ Speakers ] [ Tubes ] [ Vinyl ] [ Digital ] [ Hi-Rez ] [ Video Asylum ] [ Cables ] [ Tweaks/DIY ] [ Music ] [ Films ]