Advice for novice speaker builder.

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hello.

I am planning to design (and eventually build) a stereo pair for my small listening area. The layout is such that the speakers will be 7' about, 1.5' feet from the back wall and 10' from the prime listening position. Nine foot ceilings and in one end of a room 25' long.

I am interested in music, mostly classical, small scale with emphasis on voice and the midrange. I wish to have a 3 or 3.5 way speaker system that could cover 9 octaves. My initial thoughts are:

1. the speaker will be a truncated trapezoidal shape with the approximate dimensions 100cm height(before spikes), width 15cm at the top to 30 cm at the bottom. The depth is planned at 30 cm.
2. The primary material will be Appleply plywood by States industries (3/4" or 1".

3. The design will be MTMWW (hopefully) and I wish the crossovers to be housed outside the enclosure.

The initial drivers I have tentative proposed are:

Tweeter: Tang Band 15-1719s (1" ceramic over aluminum dome).
Midrange: Tang Band W3-13645 (paper cone)
MId woofers: Dayton Audio ES180Ti-7 (7" Nomex? cone).

4. The crossovers will be the issue. But I would like to cross over at approximately 600 hz and 4800 hz. The mids would run in paralled as well as the mid-woofers.

5. The woofers I am uncertain about the physical configuration.

a) ported sharing the same enclosure space with twin ports and double are volume.
b) Ported as above but with the .5 technique of crossing the lower woofer at a lower frequency.
c) the top woofer sealed (in its own chamber) and the bottom woofer ported with the .5 as above but of course a smaller chamber than with both. ( I have never heard of this configuration and it just may be silly). Hope springs eternal.

Thank you for any advice or direction and all comments are gladly and gleefully acknowledged.

Artofwar.
 
Nine octaves would be 40 Hz - 20 kHz. Not having done any simulation, however my general impression is that your truncated trapezoid with 30 cm sides at the base, and two 7" mid-bass drivers wouldn't be able to produce f3 = 40 Hz.

Quickly looking thru the PE 2013 catalog, I couldn't spot Dayton ES180Ti-7. If you don't have simulation software, and you were to list the specs of that driver, there are probably people here who could tell you what is best to do in the bass range.

Regards,
cT
 
Hi,

Walk before you can run would be a good idea.

$30 tweeter + 2 $20 each midranges is OK but
then you add two 7" bass units at $140 each ...!!

YMMV but I suggest you consider building these :
https://sites.google.com/site/undefinition/diy-sunflowers

An externally hosted image should be here but it was not working when we last tested it.


Read the good FAQ's on that the site, and shortcut years of
building speakers, messing it up, and learning the hard way.
https://sites.google.com/site/undefinition/diy-mfaq
https://sites.google.com/site/undefinition/diy-mfaq2
https://sites.google.com/site/undefinition/diy-faqs-provendesigns

rgds, sreten.

The esoteric Dayton :
http://www.parts-express.com/dayton-audio-es180ti-8-7-esoteric-series-woofer-8-ohm--295-397
 
Last edited:
Thank both of you for your replies.
Serena the speaker you showed looked very interesting and I will read up on it.

The cost of the speaker drivers are quite disproportional towards the woofers 4 @ $133 x4=$532. But I did not start there. I started by looking at very competent and expensive midranges like scanspeak and Visaton. It is very difficult to find a good midrange driver which starts in the 150-200 hz range.

I am hoping that the woofer mid x-over @ 500-600 hz would work out although it is in the vocal range. This makes it easier for this small midranges to operate more comfortably and pushes the tweeter x-over into the 4200-4800 hz range.
But this tweeter needs this (apparently).

The Dayton Audio woofers require 1.14 cu. ft. for ported (f3=35) and much less sealed f3=69). This design will provide these volumes.

I only chose to double the woofers for sensitivity, using 8 ohms, 6.5 or 7 inch and neo motors for requirements.
I am prepared to consider other alternatives.
Thank you for your responses!
Artofwar
 
artofwar, welcome to the world of speaker building :)

This project you have listed here is very ambitious and may become overwhelming quickly. Nailing a 3-way design requires a lot of time, and a lot of measurement. Ultra premium drivers like Scan-Speak and Visaton mids are only good if you already know how to extract the most performance from them. It is also impossible to choose a crossover point without having measured the drivers or having said measurements. The crossover points depend on driver behavior in real application, not the frequency responses provided by manufacturers.

I am of the opinion (very firm opinion) that the best way to begin in loudspeaker design and building is with a verified design. The reason is thus: most people do not have the measurement equipment, money, experience or time required to create a very complex design, much less as a first project. When you start by learning and understanding an existing design, you appreciate the nuances of the hobby but yet know that the net result is going to be excellent.

I suggest that, unless you have already purchased these components, that you look on Tech Talk or Zaph Audio's page for simpler projects. Unlike many things, a poor loudspeaker design is always much worse than the sum of its parts. The converse, however, is also true: it's possible to create the best sound you've yet heard without breaking the bank. If you would like suggestions, I would be more than happy to help you.

The place to start, in my mind, is with the following question: what exactly are you looking to get from a speaker design? Absolute 2 channel fidelity? Home theatre versatility? I think you should consider that first and then go from there. Hopefully this helps some. :)
 
Thank you Lemans23.

I think your advice is very sound. My plan is based on the Little Princess mtmww design. I am not in a hurry to rush toward oblivion. I am listening very carefully to what is being said.

I am not a ready, shoot , aim sort of person.
I realize that the final outcome is very room dependent and serious measurements will be required. It is a process of iteration, compromise and competence but not perfection.

To address your question about what I am looking for:
Stereo fidelity yes but I am not sure about absolute. I have listened to many speakers over the years of all types and all have pros and cons. I listen to opera, chamber music, choral works, jazz, blues, folk, etc. the highest priority is the human voice.

The speakers that I liked the best were where the speakers were not driven too hard e.g. Verity speakers, focal Utopias etc. there is a certain effortlessness to this sound.

I would prefer quality over quantity. For example I would rather emphasize phase, low distortions and vocal acuity over high spl, bass extension or extreme resolution.
For example if this system were to be sealed with an f3=69 hz that would be fine with me.

I do not know what spl level to target sitting 9-10 feet from the speakers but it doesn't seem to be too high. There is 3' behind this position and there are 3 seats in a row.

To me it seems this modest 3-way floor stander design will fit in my space and drivers can be recommended that operate in their respective frequency bands and integrate well ( play well together).

I am perhaps not the best at describing this but perhaps this discourse will help.

Again thank you for helping me.
 
I see the 2 criteria of MTMWW and quality vs quantity as slightly contradictory.

You can spend x amount of dollars for 2 mids and 2 woofers or you can spend the same amount of money for a higher quality single mid and single woofer. Doubling up drivers is generally for the purpose of increasing the SPL capabilities (ie. quantity), although doubling also means a little less distortion and less effort and thus why I said "slightly contradictory".

Since I don't actually listen to much of the same music that you do I could be wrong here, but I would think that for your tastes you should emphasize the midrange and treble (or at least equivalent quality across all 3 drivers). I'd say your driver selection does the opposite right now.

So start here: how loud do you need to go? --> TMW or TMWW or MTMWW? --> then what does your budget allow for each of the speaker's 3 'ways'? --> pick your midrange 1st and then the T and W to match.

And now contradicting what I've just said above :rolleyes:, although there isn't any real feedback on the new ES180's yet, they do appear to be a very high quality driver and you may want to stick with those if your budget allows.
 
4. The crossovers will be the issue. But I would like to cross over at approximately 600 hz and 4800 hz. The mids would run in paralled as well as the mid-woofers.

5. The woofers I am uncertain about the physical configuration.

a) ported sharing the same enclosure space with twin ports and double are volume.
b) Ported as above but with the .5 technique of crossing the lower woofer at a lower frequency.
c) the top woofer sealed (in its own chamber) and the bottom woofer ported with the .5 as above but of course a smaller chamber than with both. ( I have never heard of this configuration and it just may be silly). Hope springs eternal.

Why not use drivers used in a proven design, such that when your crossover design fails miserably (which it will) you can convert to the proven design.

Or you can use the proven design as a benchmark.

It seems that you picked the criteria based on "art", "taste" or "magic". No, it doesn't work like that! There are specific reasons why designers do "this and that".

The crossover frequency (fx) mostly dictated by the chosen drivers. For tweeter fx, it is usually not far from the tweeter's resonance (1.5 to 2x depends on slope steepness). The woofer fx can be 100Hz to 500Hz depends on the woofer-mid match (trial is done during design for each possible fx).

Deciding whether woofers will be series, parallel, same box, or different tuning frequency also done during design based on situations.

Three-way tend to create too low speaker impedance. So you don't parallel 4 ohm woofers but put them in series. Parallel woofers are usually opted to increase efficiency (SPL) to match mid and tweeter.

Parallel mid is usually opted to limit displacement, especially with brittle cone material that tend to break up at high SPL (ceramic, poly-glass, diamond, etc).

Two-and-a-half-way (woofers with different tuning) is rare because it requires that one woofer is efficient enough and the second woofer is only slightly anchoring the bottom end (similar like a port does)
 
Thank you Lemans23.

I think your advice is very sound. My plan is based on the Little Princess mtmww design. I am not in a hurry to rush toward oblivion. I am listening very carefully to what is being said.

I am not a ready, shoot , aim sort of person.
I realize that the final outcome is very room dependent and serious measurements will be required. It is a process of iteration, compromise and competence but not perfection.

To address your question about what I am looking for:
Stereo fidelity yes but I am not sure about absolute. I have listened to many speakers over the years of all types and all have pros and cons. I listen to opera, chamber music, choral works, jazz, blues, folk, etc. the highest priority is the human voice.

The speakers that I liked the best were where the speakers were not driven too hard e.g. Verity speakers, focal Utopias etc. there is a certain effortlessness to this sound.

I would prefer quality over quantity. For example I would rather emphasize phase, low distortions and vocal acuity over high spl, bass extension or extreme resolution.
For example if this system were to be sealed with an f3=69 hz that would be fine with me.

I do not know what spl level to target sitting 9-10 feet from the speakers but it doesn't seem to be too high. There is 3' behind this position and there are 3 seats in a row.

To me it seems this modest 3-way floor stander design will fit in my space and drivers can be recommended that operate in their respective frequency bands and integrate well ( play well together).

I am perhaps not the best at describing this but perhaps this discourse will help.

Again thank you for helping me.
A properly executed 2-way speaker will always beat a poorly executed 3 way. When a 3-way is done right, however, the benefits may often be worth the extra driver cost. I do not like half-way designs because as Jay said, the theoretical advantage is outweighed by the practical limitations. We won't get into that however. :)

The sound you're speaking of is due to low distortion. When drivers are properly integrated, are of high quality, and are able to reproduce their passband comfortably, the power sent to the speaker is turned into glorious sound. When you need to push the speaker to create the sound you want and it cannot handle it, you induce distortion which is likely the strain you mention.

Much like the human body, two components rule all others in loudspeaker design. The brain, #1, is the crossover. You could have Scan-Speak Illuminator woofers and Seas Millenium tweeters, but if you make a poor crossover, the speaker is just an expensive thing to look at. The heart, #2, is the tweeter. People more often than not will spend boatloads on midranges and woofers, and then put a decent tweeter into the mix.

The tweeter reproduces the most crucial frequencies for musical accuracy, that is, the harmonic structure called timbre. If the tweeter creates distortion or is forced to play either too low or too loudly, then your Seas Excel magnesium woofers are worthless also. If the tweeter is not integrated well in the crossover, however, then even a Morel Supreme tweeter is a waste of money.

Now: do not mistake me saying this as advising spending lots of money on a tweeter. Rather, I am saying that if you take some cost from the ultra premium woofer budget and allocate it to a superior tweeter, then you're going to have a nicer sound.

So here is my advice: set yourself a budget for this project such that if it does not turn out the way you want, it will not be considered a complete waste. This could be Dayton Audio classic series woofers and a Vifa DX25 tweeter just to throw out some value options. Design an enclosure for the woofers, and measure the tweeter response IN THE BAFFLE. This is crucial to crossover design. Spend the remaining 80% of your time with a good crossover simulation program (one which accepts real measurements, not X-Over Shop) that lets you easily place components. You will need to match the tweeter level to the woofer level, not the other way around. Knocking down the tweeter level is fine. Running the tweeter flat out with 4 woofers is not.

Buy some x-over components, but don't spend a huge amount. You do not need expensive caps and inductors. Build your first prototype crossover, and listen to it with a reference speaker. Your brain will mask the DIY's flaws because you made it. You expect it to be great, therefore it will be great regardless of its faults. You must have a tried and true reference to combat this. It happens to absolutely everyone who makes their own speaker without anything to compare it to. You may even begin to think that the DIY design is better than "anything in a hi-fi store". Don't fall into this trap.

After you've listened to it extensively, determine what is missing compared to that sound you described. Is there an edge to it? Midrange emphasis? Sloppy bass? Once you get to this point, report back.

Don't forget what I said about the components: expensive components don't promise that the sound will be good. It just means that the components are expensive. Best of luck. :)
 
I am interested in music, mostly classical, small scale with emphasis on voice and the
midrange. I wish to have a 3 or 3.5 way speaker system that could cover 9 octaves.
Artofwar.

Hi,

My suggestion is post #3 is a good as your going to
get, at a reasonable price and well documented.

It is something different in terms of voices and the midrange, and
suits the above outline to a tee, it would be very churlish to ignore.

Note that it is the 3rd generation of a design affected by
years of seriously developing developing speakers, and
you get all that experience for free if you so choose.

Perfect, no, nothing is. Highy developed yes.
No fuss bass tuning recommended.

rgds, sreten.
 
Last edited:
Gentlemen: Lemans23 and Sreten:

Again thank you for your responses. I agree with your assessment of the effortless sound that comes from drivers that are not over driven, have competent builds and eventually have through iteration and measurement have smooth crossovers.

My budget for the drivers alone is approximately $1,000. This does not include any enclosure costs, x-over costs nor measuring equipment.
I fully understand the importance of the tweeter. I have heard the soft dome air circ but it is out of my budget.

Because of your responses I am considering different midrange drivers. Also I will seal the woofers. Do you think the following configuration in is an improvement?

Woofers: as before dayton esoteric 7" in parallel
Midranges: vifa 4" mtm in parallel.
Tweeters: tang band 1" titanium dome.
Artofwar
 
GOAL: Truncated Trapezoidal (discussion examples)

I would suggest a TMW if a wide'ish cabinet is acceptable to you. One driver per range.
I would suggest a TMWW if you want a slightly narrower cabinet. A two box TM and WW allows using the TM for computer/bookcase speakers. Extra WWs can be duplicated later to balance room bass modes.

MTM designs often suffer from lobing, since short spacings are demanded between the M-T and M-M. If the T/S parameters vary between the two M units, you get distortion.

Look for a well reviewed, fully specified designs that you can successfully copy and learn from. Use the genius engineers that have spent months selecting compatible drivers, building cabinets, taking measurements, building and rebuilding crossovers, plus going through the review process.

Designs using SB Acoustics drivers often provide high performance at a
modest price.

http://www.diyaudio.com/forums/multi-way/212738-sb-acoustics-3-ways.html
 

Attachments

  • Examples.jpg
    Examples.jpg
    65.2 KB · Views: 359
  • SB Acoustics.jpg
    SB Acoustics.jpg
    129.9 KB · Views: 358
There is an interesting comment I often see thrown about on forums. It's the notion that hi-fi companies can't afford the high end components that the consumer can, or likewise, a DIY will have higher quality bits in it. When you build a speaker system as a manufacturer using Seas or Scan-Speak drivers as examples, you get those drivers at build cost + internal margin. This is a mere fraction of what any consumer would even imagine. The total sum cost of the total loudspeaker from the manufacturer's side might be the cost of the DIYer buying a pair of tweeters for an extremely high end speaker. The pricing comes in the form of endless hours of engineering, marketing, shipping, and all other things. I just want to make it plain that hi-fi companies do not "cut corners due to cost" when it comes to their products. Were that the case, they wouldn't hire the engineers whose last job might have been designing Morel tweeters (just as a real life example).

The problem is when audiophile companies (note the difference) try to sell you a set of Scan Speak revelator mids and tweeters in a nice looking speaker costing $40,000. That's when it gets silly. Take my word that there are companies out there charging 10-20k for speakers because the driver cost alone is simply in the 1000s. The Sonus Faber Aria is a perfect example of this. That said, again, the speaker is priced at something like $100k. It does not mean, however, that the guts and engineering are not the best in the world.

Food for thought. :)
 
It will help if you link to the drivers you are considering - I'm not sure which of the TB titanium tweeters you're thinking of. This one?

Just fyi, the Vifa NE123-8's are N/A right now and there are only 2 of the ES180-8's available from Solen at the moment.

Beyond that, a quick glance looks likes those drivers will work well together in terms of SPL and xo point matching.

If I was going MTM with 4" mids, I might consider a ribbon tweeter instead, like:
CSS RT2
Fountek NeoX1
Fountek Neo CD3

For what you want, I would still perhaps consider a single higher quality 5" or 6" mid although those Vifa NE123's are indeed supposed to be quite good. Like:
the new 5" Satori (although also N/A at the moment)
6.5" Satori
5.5" Scan Revelator
BG Neo10 Planar
You will still get enough effortless SPL's from these when they're just used as mids and crossed over to the ES180's somewhere around 300/400Hz.

Unless you are going to run with a sub, I would want an F3 in the 40Hz neighborhood so I'm not exactly sure why you are changing from ported to sealed?

And a couple of other threads you might find helpful:
Favorite tweeters at different price points
What-mid-and-tweeter-for-high-end-3-way
 
Hi,

I have 40+ years of experience and my advice you can take or leave.

However I am used to to newbies and rambling lists of drivers
to consider, and pedantic sticking to very unsuitable drivers,
and principles, with no real clue along the lines of how they will
work together and why, other than spending a lot must work.

I'll guarantee if you plough your own furrow with expensive
drivers, and don't give up after your first expensive mistake,
everything I'm advising will make perfect sense, but not now.

No way should the first pair of speakers you build be an
attempt to also build the last you build, and any attempt
to build your perfect speaker based on your conjecture.

So far your sticking to all the mistakes novices make,
you asked for advice, your best bet is to simply take it.

IMO the Sunflowers Redux ticks all your boxes to a tee.
So IMO its a nobrainer, what you want on a plate.*


Forget about spending at least $1000 on drivers, you
won't do better on your first attempt with anything.

Might sound harse, it is intended to be, for all the
will in the world novices don't lay golden eggs,
and in the DIY world golden eggs are really
good designs that perform a balancing act
to produce a great sum of the parts.

rgds, sreten.

* All design is a compromise, balancing stuff, and PC
is quite clear about why you may want the Sunflowers.
PC is no mug, and if its not obvious whar you get
involves very litle compromise of other factors.
 
Last edited:
<snip>
IMO the Sunflowers Redux ticks all [the OP's] boxes to a tee.
<snip>
Seems like an interesting (if tallish) speaker which appears not to be too demanding of woodworking skills. I suspect it would have a great mid-range performance?

The Vifa XT25TG30-04 appears to be available in the UK for about £30. However, I haven't been able to locate a UK source for the Dayton Reference RS225-4 or RS125-4 - any suggestions other than importing from the US?
 
Did you build a Sunflower Redux then?

If so, did you buy the drivers and crossover components from Parts Express in the US and how complex was the cabinet and crossover construction process?

No, I didn't build the Redux but some other (and parallel drivers is never my goal). If you know speaker design and familiar with drivers, you can understand a lot of things from looking at the design/crossover. Because when you work with a certain driver or pair of drivers, if you are really a speaker builder, you may have built/listen all possible combinations, and may come to "conclusion" similar to others regarding how to make use of the drivers.

Also, when I suggested novices to build proven design, it is not just because they are novices. I myself will build ALL available designs I can find using the same drivers before building my own. Having a benchmark (side by side comparison) is really helpful.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.