looking for 2way design -planar/isodynamic or fullrange +woofer

[SoneNelson]

I am designing a custom 2 way tower from scratch and am at the point that I need some advice, This is my first Home Theater project. I have custom built dozens+ of subwoofer enclosures for cars, but custom crossovers and matching tweeter's with woofers/midranges is unfamiliar territory, I have over a month of researching a couple hours every night but still have a lot to learn.

I have read a lot of diy threads and I don't want to go with a proven design because 1; This is fun for me 2; I was not able to find a design using the parts I wanted to use and would like to be able to make future designs of my choosing 3; The experience gained and having a speaker I can put my name on is priceless to me.

my current design plan is pending but I plan on it being a simple 2 way tower with with a 8" woofer and the widest range planar/isodynamic tweeter. As of now my top choice is the BG NEO8 with a Peerless or SEAS woofer. crossed as low as possible 500Hz-1kHz. I am planing on a sealed enclosure with separate tweeter chamber for tight fast bass. Separate subwoofers will be used to supplement the LFE's.

>EDITED<

these will be the first pair for a HT 7.1 setup. I will be using subs and later low frequency transducers but would like these to be as full range as they can be without sacrificing clarity/ causing distortion.
My original goal was studio/pro quality, low distortion, high sensitivity, full range towers good for Home Theater and some Music. but since I believe music is harder to faithfully reproduce then HD HT sound these will be a experiment to suit my taste for ultra high definition, resolution and fidelity in a HT environment.
by budget is about $1000 for the first pair

- Random Questions -


Would a sealed enclosure really make a difference in distortion in the 85hz-1000hz range vs a port design?

What difference does it make if I placed the port lower? I prefer forward or down firing because they may be closer to walls than I would prefer.

Any suggestion comments questions are welcome and very much appreciated!

[peter_m]

Do you have any measuring equipment?

"...budget is about $1000/pair not including "extras"..."

Just a thought: ZRT

Peter

[Dave Bullet]

Let me be honest but in a helping sort of way.

I hope you are not using text book formulas for crossover component values. That doesn't work in the real world. If you are dead keen on proceeding and cannot measure or use modeling software on measured results with that sort of money, you must be either rich or mad (or both).

Cheap and nasty caps = much better sound than fancy caps with a hit and miss design.

[---k---]

I understand you wanting to building your own. I was motivated by that myself, but I teamed w/ CJD. We fear that you'll get very subpar results with the approach you're taking. Your results may sound decent, but for that much money, they should sound amazing. It pains us to see people getting subpar results. We're trying to be helpful, even if it doesn't seem like it.

It looks like your attempting to use X-over-3 to design the crossover. That appears to be a very very simplistic crossover, with just text book filters. Is it using measurements?

Designing a decent crossovers without your own measurements is possible, if someone else has done the measurements. I suggest you read this thread at PE:

http://techtalk.parts-express.com/showthread.php?t=210439

If you're up to that, go forth. If you're not wanting to do the SPL tracing and then modeling in Speaker Workshop or Jeff B. PCD, then I really really recommend you build a proven design.

Like Dave said, but slightly different: An experienced designer can make poor performing drivers sound much better than a novice designer with expensive drivers.

Good luck.

[penngray]

3 sites same post??

Diyaudio.com
avsforum.com
and here....he is a busy guy

I didnt check PEs forum, he probably posted there too

[JonMarsh]

Don't you just hate that sort of "shot gunning"? :W

Seriously, everyone above is posting with the best of intentions. I notice you haven't included any baffle step compensation in your crossover model. IF you don't know what baffle step compensation is, put the power saw and the computer down, and go read some of the listed reference threads and some of the design discussions in the proven designs section of this forum and some of the other designs around the web, like Zaph Audio. When you're pretty sure you understand what's going on with these, and if you're ready to pop for some basic measurement stuff (at least an ECM8000 mic, a mic preamp/UB/Firewire interface, etc, then consider designing your own.

I've been doing this for a few years, my first designs in 1973, and the other moderator and I counted up the total number of designs I've done since then (up to 2006), it's about 52 altogether. I've finished one more since then, and am working on another now, with two more in process (i.e., drivers on hand, cabinets planned, etc). I'm a slow learner, I guess... :W

Don't waste the money you're thinking about spending. If you want to go on your own, spend more time educating yourself and get some basic measuring gear and a package like LspCAD or Sound Easy, or Arta to save bucks. If you want some speakers to listen to soon, try a proven design.

BTW, I'd suggesting selecting drivers that have published distortion performance measurements by some one- that might help you with your driver choices. Hint: The HiVi RT8II is not very clean. You can do much better for $86 each. Neo3 PDR is an alternative to consider if you want a planar driver. You can do better for $30 each. Of course, if measured performance isn't important....

One last suggestion- there are many things we can hear about speaker systems related to the quality that aren't always easy to measure, but pretty much anything you can measure you can hear easily- this includes driver resonances, SPL non-linearity, and distortion.

Good luck with your project!

BTW

[SoneNelson]

3 sites same post??

Diyaudio.com
avsforum.com
and here....he is a busy guy

I didnt check PEs forum, he probably posted there too

LOL... This is my first time posting in audio forums so I figured it best to feel out the different forums(but didn't post in PE), some are more helpful and more active than others, as I found out when I first picked a computer forum. I do prefer to have only one site to check and participate in, And so far this has been the most helpful and knowledgeable forum.

I am and have been researching the suggestions made so far I do plan on buying measuring equipment and have been researching proven designs to try and understand how they were designed(reverse engineer). I'm not in a rush I assume I have at LEAST another month of heavy research before putting a design into production.

I thought from my research that the HIVI tweeter was a pretty good choice, the rt2c and rt8II used in these speakers.

Swan M1 reference class minimonitor $729

http://www.theaudioinsider.com/product_info.php/loudspeakers/swan-high-end/p/swan-m1/cPath/21_23/products_id/28

RT8II Isodynamic Tweeter

http://www.swanspeaker.com/product/htm/view.asp?id=18

Ed Momkus reports the feat that is the $15,000 Coincident Speaker Technology Total Victory IV speakers

http://www.dagogo.com/CoincidentTotalVictoryIV.html

Reimer Tetons GS

http://www.6moons.com/audioreviews/roadtour10/roadtour10.html

B99

http://images.google.com/images?q=Reference-%20line%20loudspeaker%20B99&oe=utf-8&rls=org.mozilla:en-US:official&client=firefox-a&um=1&ie=UTF-8&sa=N&hl=en&tab=wi

but I can understand about going with drivers that have published distortion levels, minimum distortion is one of my design goals and I am open to any suggestions. (looking at other tweeters now)

I am exploring a 2.5 way design now, my modeling with bassbox/Xover pro looks much better with 2 SEAS 7" woofers. but with a different tweeter I may go back to a 2.0.

I have been pretty busy but will post more in the next couple days...

[Dave Bullet]

At the very least, if you proceed with designing your own take an impedance measurement sweep at the very least. You'll want to make sure you haven't created a short in the crossover and minimum impedance is at least higher than rated for your amplifier, otherwise you could destroy the output transistors and worse leaving you with an expensive fix up job.

PS: Start with a 2-way MT. Yes it is fun pondering 2.5 ways, 3+ ways, bipole or dipole tweeter designs, open baffle etc... but you are much more likely to design something that will sound acceptable with a simple old 2 way MT.

You should also have a good handle on the following topics:
Baffle step & compensation
Diffraction ripple
Power response
Relative phase
Electrical vs acoustic rolloff
Impedance
Group delay
Power handling - thermal and excursion
Lobing

PS: The above aren't in any particular order. Some could argue that some of the above aren't required to have a "decent" sounding speaker.

Cheers,
Dave.

[JonMarsh]

At the very least, if you proceed with designing your own take an impedance measurement sweep at the very least. You'll want to make sure you haven't created a short in the crossover and minimum impedance is at least higher than rated for your amplifier, otherwise you could destroy the output transistors and worse leaving you with an expensive fix up job.

PS: Start with a 2-way MT. Yes it is fun pondering 2.5 ways, 3+ ways, bipole or dipole tweeter designs, open baffle etc... but you are much more likely to design something that will sound acceptable with a simple old 2 way MT.

You should also have a good handle on the following topics:
Baffle step & compensation
Diffraction ripple
Power response
Relative phase
Electrical vs acoustic rolloff
Impedance
Group delay
Power handling - thermal and excursion
Lobing

PS: The above aren't in any particular order. Some could argue that some of the above aren't required to have a "decent" sounding speaker.

Cheers,
Dave.

As I'm sure you know, sometimes accidents of fortune can happen- "Beginner's luck"? But when you have a lot of variables to juggle, as your list so aptly conveys, the luck factor becomes less and less probable.

There is certainly a range between what some folks would rightly call "decent" sound, and that which is realizable with care from an optimum set of drivers.

Last, I would point out that though a driver having been used in a commercial design is sometimes a reliable indication of quality, it's also good to do a sanity check- for example, Swans is a division of the same company that makes the HiVi tweeter- or the other way around. It just says they eat their own dogfood.

With regards to audio reviewers, I respect and appreciate John Atkinson at Stereophile because of the measurement competency and depth he displays in his reviews; the other magazines focussing on subjective reviews I tend to be a little more skeptical of. Just compare his review of the Avalon Indra with some of the other ones online. Now, the Indra has some good qualities, but from the Sterephile review you can readily identify the weaknesses.

And if you read a lot of Stereophile reviews, you'll come to realize there are a lot of commercial speaker designs that are not all that from measured performance, and that a solid DIY design that dot's the i's and crosses the T's on on the details may substantially out perform commercial speakers costing much, much more.

In the end, many constructors are looking for a specific "sound", and engineering that "sound" takes some knowledge and skill. But the truly sublime result, in my opinion, is when you can engineer a system that contributes as little of it's own sound as possible, and just gets out of the way of the music and original recording. Even an MT two way with careful engineering and choices can approach that, particularly if the driver and crossover budget is liberal; the real genius is being able to approach that on a shoe string. That's what marks a gifted and competent commercial designer.

[djn04]

Do you have any measuring equipment?

"...budget is about $1000/pair not including "extras"..."

Just a thought: ZRT

Peter

Ditto. can't recommend the ZRT's enough.

[fjhuerta]

X-Over 3 is a toy. Its modeling engine cannot take into consideration real impedance - it needs shunts and filters to aproximate a flat one, and even then, it won't consider BSC or diffraction. It's pretty much useless for serious crossover work (and, given your budget, you are building a pretty serious speaker).

[SoneNelson]

So many topics to google search so few hours in the day... I have been jumping all over from suggested topics, random threads, diy designs, google searches, driver reviews, etc.

Thanks for the suggestions, I feel like this a project and hobby I am capable of undertaking, but I was getting a little ahead myself thinking a had a design worth posting I am still in the how do design a quality speaker phase.

I am looking for better software, I think I saw the LEAP 5 program referenced on a speaker Brands homepage that looked good. What software is best, what are people here using, free or not? I am about to google this now but suggestion from actually people are great.

I feel that some real world experience would be ideal, so I am thinking about building a simple 2 way design and tweaking/experimenting with it using proper measuring tools and better software. I think this would be a very helpful experiment at this point.

For a first experiment I am thinking the NEO8 with a 8" peerless exclusive/PPB woofer. The NEO8 as a mid/tweeter transducer will (I think) create my ideal sound for HT and I'm sure I could reuse both of these in different experiments like a dipole or line array since I have seen both of these used in DIY projects.

For a first experiment I am thinking NEO8 in monopole but I have not been able to find much data on it in this setup only as Dipole/open baffle. with an 8" peerless in a sealed enclosure.

I will be back for more when I run out of ideas to research.

[ch83575]

SoneNelson,

I am with you on wanting to design from scratch. As a matter of fact I have been doing just that for about 3 years now... without my first success. If you are really going to start from scratch you really will need have a well-rounded knowledge of the design factors Dave listed and a measurement setup. Even then, it aint easy. But, that said, I don't think building a published design prepares you for doing your own design very much. It gives you experience building, but not designing.

I might suggest something different: build a published design (a simple 2way would be best), used your measurement rig to rev-engineer it and see how it ticks (I know that all of the published designs show you things like the FR, but that is not the same as measuring it yourself), then try your hand at re-designing the crossover yourself. Javier just had a post about re-doing the crossover on Zaph's 3 way . Maybe that's a good compromise. I know the frustration of ending up with drivers that I do not have the expertise to make work, and basing your first design on at least the drivers and cabinet size of a published design would at least give you a fallback.

So there is the compromise scenario. It's a no-lose solution. Or, if you persevere with starting from scratch do so with the knowledge that it is a long road with lots of learning and probably some wasted money... but it is a lot of fun!

-Chad

[Dave Bullet]

SoneNelson,

I am with you on wanting to design from scratch. As a matter of fact I have been doing just that for about 3 years now... without my first success. If you are really going to start from scratch you really will need have a well-rounded knowledge of the design factors Dave listed and a measurement setup. Even then, it aint easy. But, that said, I don't think building a published design prepares you for doing your own design very much. It gives you experience building, but not designing.

I might suggest something different: build a published design (a simple 2way would be best), used your measurement rig to rev-engineer it and see how it ticks (I know that all of the published designs show you things like the FR, but that is not the same as measuring it yourself), then try your hand at re-designing the crossover yourself. Javier just had a post about re-doing the crossover on Zaph's 3 way . Maybe that's a good compromise. I know the frustration of ending up with drivers that I do not have the expertise to make work, and basing your first design on at least the drivers and cabinet size of a published design would at least give you a fallback.

So there is the compromise scenario. It's a no-lose solution. Or, if you persevere with starting from scratch do so with the knowledge that it is a long road with lots of learning and probably some wasted money... but it is a lot of fun!

-Chad

+1 - good suggestion.

Choose some drivers and cabinet of an existing design, then give your own crossover a whirl. (Read: With cabinet being the same - the BAFFLE dimensions and driver placement on the baffle the same - this affects diffraction and baffle step and is an integral part of crossover design).

If your own crossover doesn't work out - you can put the original designers in and still be satisfied.

[SoneNelson]

I agree...

I have been looking for a proven design to experiment with but have not found one that really jumps out at me, I do think the minilith design using the NEO3/NE08/8" SEAS is interesting http://www.htguide.com/forum/showthread.php4?t=26798
but after reading the thread this design does not seam to be the most thoroughly documented and I a 2way design would be best to start with.

I got a little side tracked from my studies and spent most of the day reading about the Manger bending wave driver/transducer, I don't have the experience to invest in a pair (just yet) but from everything I have read a Manger 2way design would be my ideal setup.

Does anyone know of a proven or well documented 2way design for a planar/isodynamic/ribbon or bending wave driver? or a 2way NEO8 design?

[ch83575]

I agree...

I have been looking for a proven design to experiment with but have not found one that really jumps out at me, I do think the minilith design using the NEO3/NE08/8" SEAS is interesting http://www.htguide.com/forum/showthread.php4?t=26798
but after reading the thread this design does not seam to be the most thoroughly documented and I a 2way design would be best to start with.

I got a little side tracked from my studies and spent most of the day reading about the Manger bending wave driver/transducer, I don't have the experience to invest in a pair (just yet) but from everything I have read a Manger 2way design would be my ideal setup.

Does anyone know of a proven or well documented 2way design for a planar/isodynamic/ribbon or bending wave driver? or a 2way NEO8 design?

I know that once again this is probably not exactly what you want to hear, but here goes: If I were able to make my first driver purchase over again, I would chose to mimic the Zaph ZR71. They are great drivers, the design is well respected and most importantly it is reasonable to think that a beginner can effectively design a crossover that will sound good for them. If you pick something esoteric you will really be out on your own, you won't know where it crosses well, how far you can push, or, for that matter if you really will even enjoy how it sounds in the first place.

I recommend the ZR71 because I have never heard of somebody specifically not liking the sound of a paper woofer and soft dome. Some people find their true joy with hard domes and cones, but I have never heard somebody actually dog the tried and true paper / soft dome combo. So, you will probably be satisfied with the sound of the ZR71, and the drivers are pretty well behaved, so you can probably come up with your own crossover without much trouble. If you were to choose a design like the NatP, you might love the sound, but the chances of you being able to do your own crossover and liking it as much are slim.

I just throw that out there because it is what I wish I had done. I had all of the thoughts you are having, and my first woofers that I purchased were slightly exotic Arum Cantus non-woven carbon fiber woofers... and I never could get them to behave in any respectable way. They sit on my shelf while I attempt to finish a design with better behaved woofers.

-Chad

[HareBrained]

SoneNelson, don't let these guys discourage you from pursuing designing and building your speakers. You will learn a lot more doing it yourself.

I had success with my first design. I spent 4 months learning about the software tools and worked through multiple ideas before I was comfortable to outlay the $200. It's been 6+ months since those and I've built a second design and spent a lot of that time learning a lot more. It's only now that I would feel comfortable to spend a grand on a pair but that's a lot of money to me.

Were I starting with a $1000, I would invest in a measurement system first. But only if this were a first-of-many design. Although great success can be had building from a modeled design, the chance of creating a POS is much higher. And know exactly how the system is behaving will make up for the initial expenditure. Potentially, a $500 system measured >> a $1000 system simulated. If this is a once and done thing, build an existing design.

You may want to start smaller. Maybe you should look at designing an inexpensive TM that will become the surrounds when you know more.

Something that helped me greatly was to model a well-documented existing design. There are a lot of variables to consider and in my experience, the software modeling tools can get very close. Here's a good thread to get you started.

If you're intrigued by the Manger, you may like my latest folly, full-range/woofer combo. It's a different way to skin the cat, and one that I find very satisfying.

[SoneNelson]

@ch83575

The SR71 (I think you meant SR?) is my favorite design from the zaph site, I like SEAS and the paper woofers.

@HareBrained

This will be the first of many speaker designs, I will be building a 7.1 and a 2ch stereo for myself, How many attempts it will take before I am satisfied is likely to be a lot, but I won't mind if it takes a lifetime :B This Is Fun ...plus I would like to build a couple speakers for my friends and family, or maybe just pass off the old designs as I upgrade. so finding measuring gear with reasonable price and performance is a must.

I am thinking about building a pair of 2way monitors for my PC first to replace the logitech z-2300's I'm using. but at the same time it won't take much to improve on the Sony 'fullrange' 2x3" speakers in my current HT system.

I have also been thinking that as much as the Manger interests me that a fullrange +woofer design would be great experience to get me started. I looked at the Tang band fullrange used on the zaph site and they seem to fall off above 8kHz, but some other fullranges like the fostex's looked better, is your project posted? I could only find the "woofer + full-range 2-way?" thread. I would be very interested to here more about your design.