2-Way Design - Doing it right?

0 Members and 1 Guest are viewing this topic. Read 2185 times.

OrionX76

  • Newbie
  • Posts: 1
2-Way Design - Doing it right?
« on: 13 Oct 2020, 01:57 pm »
Hello,

This will be my first post here. So, first a quick introduction.

I recently took the deep dive into speaker design after learning that something called REW was FREE! I am a passionate about science and music. I hold a degree in music and I regularly have the best seat in the house while singing with the Billings Symphony. I also have backgrounds in optics (ran an optical lab for 6 years) and brewing (just sold my brewery, still available as a beer advisor/ drinking buddy).

I love science and art. This hobby is a perfect blend of both.

I am currently building a 2-way bookshelf speaker using the following drivers:

Tweeter - Tymphany (Peerless) XT25TG30-04
Woofer - Dayton Audio RS150P-8A

After chasing down the cause of some anomalous measurements on another forum I have finally convinced myself that I am taking good off axis measurements. At this point I have built two test boxes both full mockups with bracing, damping and rounded baffle edges. My goal here was to read the native performance of the speaker before crossover rather than relying on manufacturer's response charts and simulated diffraction.

First question: Is this recommended/ beneficial or should I be starting with manufacturer's specs and using the simulated baffle effects?

I have been learning 3 different crossover programs for various strengths and features:
-XSim
-PCD (Passive Crossover Designer by Jeff Bagby)
-VituixCAD

Anyone out there an expert with VituixCAD yet?

I could use some advise on crossover design. I am running into the problem of having to pull the trigger on one design while having probably 3 different ways I could go. I have been really trying to take into account the off axis response as is recommended by Dr. Floyd Toole. I am using my off axis measurements in VituixCAD to help inform the crossover design. Anyone with experience here would really help out. I have included some measurements if you would like to check it out. I can supply the ZMA's and other files if you'd like to help.

Wrapping this up cause I got to get to work!




Cheers!

richidoo

Re: 2-Way Design - Doing it right?
« Reply #1 on: 13 Oct 2020, 07:32 pm »
This will be my first post here. So, first a quick introduction.

Welcome to AudioCircle!!

Quote
I am currently building a 2-way bookshelf speaker using the following drivers:
Tweeter - Tymphany (Peerless) XT25TG30-04
Woofer - Dayton Audio RS150P-8A

Good choices!

Quote
First question: Is this recommended/ beneficial or should I be starting with manufacturer's specs and using the simulated baffle effects?

Measuring your drivers yourself, in the actual cabinet, after break in is vastly superior to designing a speaker using the published data from the mfg. Those data might be 20 years old and the mfg process "refined" in that time so that the data are not accurate anymore. Every speaker guy knows that the real data never exactly match the published data. You need the most accurate data possible so that your crossover design works in the real world. The math in the sim assumes T/S specs are accurate.

Quote
Anyone out there an expert with VituixCAD yet?

Not tried VituixCAD yet, but have used PCD and XSim.

Quote
I could use some advise on crossover design.
Couldn't we all...   ;)

Quote
I am running into the problem of having to pull the trigger on one design while having probably 3 different ways I could go. I have been really trying to take into account the off axis response as is recommended by Dr. Floyd Toole. I am using my off axis measurements in VituixCAD to help inform the crossover design. Anyone with experience here would really help out. I have included some measurements if you would like to check it out. I can supply the ZMA's and other files if you'd like to help.

Do all three ways. None of them will be the final answer as all sim designs require tweaking by ear. What you learn from doing all three will allow you to do the final voicing to your own perfection, and give you a good parts bin to pull from.  Follow your ears first, then your heart, then check the measurements but only to save time and not as a performance goal above and beyond what ears and heart like.

Endo2112

  • Jr. Member
  • Posts: 203
  • Particularist
Re: 2-Way Design - Doing it right?
« Reply #2 on: 14 Oct 2020, 04:04 am »
Welcome as well, I've been in the beer business for 30 years myself and your right this hobby is the perfect blend, just like brewing.

Cheers,

Don

JohnR

Re: 2-Way Design - Doing it right?
« Reply #3 on: 14 Oct 2020, 05:27 am »
I have been really trying to take into account the off axis response as is recommended by Dr. Floyd Toole.
.

Hi, you may want to address the tweeter in the 3-4k area before the actual crossover. I'd guess it's a cabinet effect although some tweeters seem to be more sensitive to it than others.

JLM

  • Full Member
  • Posts: 10654
  • The elephant normally IS the room
Re: 2-Way Design - Doing it right?
« Reply #4 on: 14 Oct 2020, 11:53 am »
Hope it goes well for you.  Designing your own gear is a daunting proposition.  Be very aware of the proud papa syndrome.  There is no perfect loudspeaker, although something like the $10,000/pair USD Genelec 8361A is tempting:  https://www.genelec.com/8361

Companies like JBL, Focal, and Dynaudio have been at it for decades with vast engineering departments and in-house manufacturing facilities, so how can you compete?  They have anechoic chambers and trained listening panels doing double blind testing.  Even younger firms like Amphion or Buchardt have years of experience and testing facilities.  Most companies, led by stockholders and marketing departments fail miserably to build a "good" loudspeaker. 

Personally I'm a fan of active loudspeaker design (a low voltage crossover feeding one channel of amplification per driver).  Genelec also uses wave guided tweeters to control directivity for more consistent sound room-to-room.  Bryston crudely modifies passive loudspeakers to be driven by external crossover/power amps (that uses their amps of course, adding significant clutter/cost) but points the way for a DIYer to go active. 

I assembled/built kits 40 years ago, but if going for it again would start with the very best, something like this:  https://purifi-audio.com/transducers/ptt6-5w08/  and  https://audiophilestyle.com/ca/reviews/purifi-ptt65-woofer-and-1et400a-amplifier-technology-review-r866/

The best you can do is to climb on the shoulders of others.  Parts Express used to host an annual DIY loudspeaker building competition/show that drew interesting designs from some of the best independent builders.  And don't forget to account for how loudspeakers interact with the room. 

eichlerera1

  • Jr. Member
  • Posts: 237
Re: 2-Way Design - Doing it right?
« Reply #5 on: 14 Oct 2020, 02:28 pm »
I would try two woofers, one tweeter configured in a D'Appolito arrangement...

JLM

  • Full Member
  • Posts: 10654
  • The elephant normally IS the room
Re: 2-Way Design - Doing it right?
« Reply #6 on: 15 Oct 2020, 11:33 am »
I would try two woofers, one tweeter configured in a D'Appolito arrangement...

Why?

True D'Appolito design requires complex crossovers and of course an extra mid/woofer and probably a larger cabinet in order to gain more efficiency.

Heard $15,000/pair MTM design a few years back that had the classic phase cancellation issue found in typical MTM designs.  Went forward to get an image of listening group and saw their heads bobbing up and down hunting for the midpoint between the midrange drivers to avoid the cancellation effect. 

eichlerera1

  • Jr. Member
  • Posts: 237
Re: 2-Way Design - Doing it right?
« Reply #7 on: 15 Oct 2020, 02:54 pm »
Why?

True D'Appolito design requires complex crossovers and of course an extra mid/woofer and probably a larger cabinet in order to gain more efficiency.

Heard $15,000/pair MTM design a few years back that had the classic phase cancellation issue found in typical MTM designs.  Went forward to get an image of listening group and saw their heads bobbing up and down hunting for the midpoint between the midrange drivers to avoid the cancellation effect.

Per Robert Harley: In a D'Appolito arrangement the tweeter is located between the woofers so that it coincides with their acoustic center to simulate a point-source radiation characteristic. In addition, variations in vertical radiation pattern—the lobing effect—are reduced, resulting in a smooth dispersion and less sensitivity to listening height.

I've never had to "Bob my head" while listening to my Hales Signature Twos.
The imaging and soundstage are world class.

I looked at some of the Butterworth Crossover Circuits that the Hales employ and they don't look any more complicated then any other sophisticated Crossover type.
Certainly, one has to read up on the circuits to get a feel for how they work, but nothing ventured, nothing gained.

NeilBlanchard

  • Jr. Member
  • Posts: 38
Re: 2-Way Design - Doing it right?
« Reply #8 on: 15 Oct 2020, 03:57 pm »
Welcome - it looks like you have chosen some excellent drivers. I would highly recommend you get a miniDSP (with a remote) 2x4 HD (and a UMIK-1 if you didn't already have a calibrated mic) - you can try all the crossover points that you could want, and do it very quickly and easily. You can test and compare four crossover points and slopes by clicking between A/B/C/D, and then adjusting them again.

And better still, you can use REW to get custom driver filters to get a very flat speaker response, or use a so-called "room curve" to get the response close to a Harman curve, which we humans tend to prefer. I did this with my Linaeum LT1000 after I replaced the woofers with the 8" version of yours - and the results are frankly amazing. The TLS tweeters in my speakers are crossed over at 440Hz, with Linkwitz Riley 12dB slopes, and with the driver filters, along with a filter for the whole speaker, this is what I measured in REW at 3' on the listening axis:



And this is the waterfall graph:



The default room curve in REW is what I used: +1dB/octave from 200Hz and down, flat from 200-1000Hz, and -0.5dB/octave above 1000Hz. So +3dB at 20Hz and -3dB at 20kHz - this is mildly tweaked from flat, and it really does sound good.

This is a transmission line cabinet, and I can't say enough about how good the bass sounds - and the tweeter is a masterpiece.

I really like the specs on the tweeter you are using, and I hope you can try some lower crossover points. Having 5+ octaves coming from the tweeter is something that you can't appreciate - until you hear it. From the response curves you posted, it looks like about 500-600Hz is quite reasonable matching levels, and REW would make leveling the broad response of the drivers really easy.

By the way, what is the cabinet design and size, that you are using? And the tweeter looks very similar to Roy Allison's design - the center phase plug may have waves propagating into the center, so it acts closer to a point source?


ooheadsoo

Re: 2-Way Design - Doing it right?
« Reply #9 on: 15 Oct 2020, 04:57 pm »
I'm not sure about crossing over a standard tweeter at 500hz.  Most can't take the power and will be producing extreme distortion unless you have a horn/waveguide.  If you use a 2nd or even 4th order slope, you will still be playing your tweeter at significant levels at 150hz.

NeilBlanchard

  • Jr. Member
  • Posts: 38
Re: 2-Way Design - Doing it right?
« Reply #10 on: 15 Oct 2020, 05:08 pm »
I looked a little closer at the Peerless tweeter you are using, and it is the dual radiator rings that are what makes this driver unique. Instead of trying to act like a piston, the soft rings each have waves propagating in a concentric pattern, and so they act closely similar to a point source. And they prove more surface area than a 1" dome, so they can drive more air. There is a lot to like with this tweeter.

Tekton Design has used this driver (I think) in some of their speakers. Having a 0.4 gram driver for most of the midrange - and seamlessly covering all the treble; is amazing. I am very tempted to pair this tweeter with a Seas U22REX/P-SL 8" Curv Cone Woofer (only 19.8 grams), in a transmission line cabinet ... (This Seas woofer is the only 8" woofer that I have found that is lower mass than the Dayton Audio RS225-P 8A, which is 23.6 grams.) And by the way, Tekton has also used the Seas 8" Curv Cone driver, in one of their speakers, as well.

The Dayton Audio 6" woofer that you are using is just 6.9 grams Mms. So the total moving mass of the two drivers is just 7.3 grams. That is pretty phenomenal.

NeilBlanchard

  • Jr. Member
  • Posts: 38
Re: 2-Way Design - Doing it right?
« Reply #11 on: 15 Oct 2020, 05:11 pm »
I'm not sure about crossing over a standard tweeter at 500hz.  Most can't take the power and will be producing extreme distortion unless you have a horn/waveguide.  If you use a 2nd or even 4th order slope, you will still be playing your tweeter at significant levels at 150hz.

This particular Peerless tweeter is not "standard" - it has a free air resonance of 436Hz, which is far lower than all but a very few tweeters, as far as I know. (The lowest I know of is the latest driver designed by Paul Paddock, called the WBLS (wideband line source) than is crossed over at 100Hz (!!). The MC Audiotech Forty-10 uses 10 of these - and crosses them over to a pair of 18" woofers, in an open baffle cabinet.)

ooheadsoo

Re: 2-Way Design - Doing it right?
« Reply #12 on: 15 Oct 2020, 05:43 pm »
Sure, but it's still a small 1.4ish inch radiator with 0.5mm xmax.  I don't think it's going to be too happy taking all that power down that low.  Well, I'm no speaker designer, so don't mind me.
« Last Edit: 15 Oct 2020, 08:43 pm by ooheadsoo »

NeilBlanchard

  • Jr. Member
  • Posts: 38
Re: 2-Way Design - Doing it right?
« Reply #13 on: 15 Oct 2020, 10:17 pm »
That is true; it is a small driver, and that is why Tekton uses an array of 7 tweeters for the midrange. That is the nice thing about an active crossover - you can try different setups, and listen. Try 1kHz and then 900Hz, and then 800Hz, and see how it goes. The tweeter has a fairly smooth roll off all on its own, so adding a 2nd order, or 3rd order might work, or try a lower crossover, but with a steeper slope.

Same for the woofer; possibly it can be the louder driver up to 1kHz, and then have a steeper slope, while the tweeter has a very low slope.

ooheadsoo

Re: 2-Way Design - Doing it right?
« Reply #14 on: 15 Oct 2020, 11:15 pm »
My understanding is that an assymetrical (acoustic) slope will not sum flat, if that's a design goal.  Are those Tektons 3 ways?  I can't imagine how else they would work without massive combing issues.

Fwiw, I have a spl spreadsheet that tells me that a single tweeter will hit xmax at about 80db at 500hz, if one were interested in xmax at 500hz.  I wonder how much distortion there would be.

NeilBlanchard

  • Jr. Member
  • Posts: 38
Re: 2-Way Design - Doing it right?
« Reply #15 on: 16 Oct 2020, 03:58 pm »
As I understand it, the models with 7 tweeters in a hexagon with one in the middle - have only the center unit go all the way up, and the 6 are acting together as a midrange. They are close enough together, that they act like a ring, concentric with the tweeter, so the comb filtering is minimized?

Someday, I'd like to listen to these.

Maybe a crossover at 800-1000Hz, would be the sweet spot for the OP's project? Listening would let you hear how the drivers, and their off axis responses blend the best together.

mcgsxr

Re: 2-Way Design - Doing it right?
« Reply #16 on: 16 Oct 2020, 07:38 pm »
+1 on the MiniDSP as the Xover for exploratory purposes - as long as you have enough amp channels.

I had some fun building a set of DIY actives around 3 years ago.  I didn't want to spend the $ for the MiniDSP and struggled using pro sound active Xovers instead.

100% what JLM said for me - I ended up not loving them, and selling off the various parts and returning to traditional manufactured speakers.

That said, you should do what you want to - I had a ton of fun with it, and learned a lot along the way (mostly refining cabinet building skills and planning said build).