Building the best loudspeakers

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


The 18W8531G00 is highly recommended. It is obviously better than the 18W8545 I previously worked with. The T330D is no doubt much better than the Seas Millenium I used.


I am thinking of using 8545 and you say 8531 is better at the same time more expensive. It is better by big or small margin, and in what way it is better? Please explain.
 
Ceibal said:
Is x over 3 pro any good? Has anyone used it?


..modeling without measurement capability (for actually building a loudspeaker), is of little worth.

The current favorite seems to be SoundEasy. It would of course also require numerous other things:

http://www.zaphaudio.com/setup.html



Anyway, doing without the design work..

This is your best bet in stand mounted MTM designs:

http://www.zaphaudio.com/BAMTM.html

As long as the baffle is the same exact dimensions then you should get a very similar upper freq. response.

This particular cabinet is the curved version of the 1 cubic foot pre-fab partsexpress box - with the same baffle dimensions.. so it should work for the sealed design without any problems.
 
ScottG said:



..modeling without measurement capability (for actually building a loudspeaker), is of little worth.

The current favorite seems to be SoundEasy. It would of course also require numerous other things:

http://www.zaphaudio.com/setup.html



Anyway, doing without the design work..

This is your best bet in stand mounted MTM designs:

http://www.zaphaudio.com/BAMTM.html

As long as the baffle is the same exact dimensions then you should get a very similar upper freq. response.

This particular cabinet is the curved version of the 1 cubic foot pre-fab partsexpress box - with the same baffle dimensions.. so it should work for the sealed design without any problems.


I really appreciate the help. I know some of these questions are stupid but I am just learning. I did buy Daytons woofer tester knowing I would need some way of testing the speakers once I got them. Will this work or did I just wast my money?
 
Ceibal said:



I really appreciate the help. I know some of these questions are stupid but I am just learning. I did buy Daytons woofer tester knowing I would need some way of testing the speakers once I got them. Will this work or did I just wast my money?

Not a waste.. if you really get into this hobby, then a quick way of testing the driver's TS parameters, (not its impulse response derivatives like freq. response), can be very useful. Its also good just to check that those parameters are within spec. quickly - so you aren't using a damaged or poorly made driver.

If you aren't really into the hobby, then you can always sell it on the secondary market (..prob. less than half the purchase price though).

Again though, IF you aren't throughly committed to the hobby then SERIOUSLY consider a proven design. Again, Troels, Tony Gee, and yes even Zaph have some very good designs.

Note that the major reason I suggested the BAMTM from Zaph is that it "fits" with the MTM you wanted and in the enclosure you specified. It most certainly is NOT a design I'd recommend as a 5K speaker beater.. not that it couldn't, but rather when I think of such a design I tend to look for several outstanding features. The BAMTM doesn't have this - its just a really good stand mount with the one excellent feature of extended bass in a sealed enclosure of a very small size. (EDIT: Actually it has one other outstanding feature: price.) Now IF you are more particular about your crossover components and can measure the values to make sure they match what Zaph spec.s - then you might well be able to enhance the design, but that really depends on if YOU hear a difference between different components. Still though, its a good place to *start*.

My suggestion then really depends on your preference:

1. If you want to design your own loudspeaker and a willing to put in a lot of time, effort, and money - then start with learning how to use SoundEasy. Note that you don't even have to purchase the program to get a "taste" for this.. instead read and re-read the owner's online manual and/or first purchase JohnK's excellent primer on SoundEasy:

http://www.musicanddesign.com/Guide.html

http://www.interdomain.net.au/~bodzio/

2. If you want a design that fits within the MTM and cabinet selection you have made then stick with the BAMTM. After making it, then consider crossover parts upgrade (Goertz foil inductors, capacitors taken for value and performance from TonyGee's Capacitor test, and perhaps winding your own resistors - do a search for that on diyaudio). IMO though, your best bet here is to make use of an outboard crossover (i.e. do not stick the crossover in the cabinet (for a visual, look at the Wilson audio Duette).

i.e the box at the bottom of the stand:

http://www.wilsonaudio.com/images/duette/Enlargements/Duette-Stand-Seamless-Grill.jpg

3. If you want a design that does have several "wow" features, then consider something like:

http://www.humblehomemadehifi.com/Humble_Statement.html

http://www.troelsgravesen.dk/JA8008.htm

http://www.troelsgravesen.dk/AcapellaNEXT.htm

http://www.troelsgravesen.dk/AT-R.htm

http://www.diyaudio.com/forums/showthread.php?s=&threadid=74508

http://www.creativesound.ca/details.php?model=AlphaLS

4. If you want something thats different (and again has "wow" features), but can be more of a "turn-key" solution - then consider the LinkwitzLab Orion or JohnK's Nao.

5. If you want something that pushes the envelope of design then perhaps you should wait until Lynn Olsen finalizes his next design (..though it could be a substantial wait).

With the exception of the JA8008, all will be costly - but all should give you "state of the art" in several respects, (though note that several have significant limitations particularly with respect to low freq. extension and power handling). Hey, its all given-and-take.

Note though that there are LOTS of other solutions as well.. for instance you could skip a lot of the complexity surrounding crossovers by using an active digital crossover (and more amplifiers). (..and of course there are compromises there as well.)

Best thing is to research it, give it a lot more thought - and then make a decision. The one thing I would counsel you on is that IF you aren't into wood working, and you DO want to pursue the hobby, then strongly consider finding some one who does that professionally near you - and paying them to do what you might not be able to.

One final "word" of advice:

If all of this puts you off, (which it certainly wasn't intended to), then seriously consider this:

http://www.diyaudio.com/forums/showthread.php?s=&threadid=76764
 
HiFiNutNut said:
Rabbitz,

The 8531 is planned to be used as a mid and upper base to cover 150Hz to 2000Hz. Below that I have the 8561. I have not built the active XO yet so the 8531 is being run from 20Hz-2000Hz without EQ in an OB stuffed U-frame, with a simulated XO to cross over to the tweeter, in MTM. This arrangement is for running-in the drivers only before taking acoustic measurement.

It sounds very good indeed. Distortion arises when playing loud bass materials but that will be gone once the 8561 comes into the picture. It is not 100% accurate as I can hear the dipole peak making the music a bit chesty but overall gives a very satisfactory sound. I can imagine how good the system sounds when completed.

I am totally happy with the 8531 used as mid or bass driver. Its bass quality is very surprising (good).

By the way, with the drivers and cabinets I have, in theory the XO point should be at 2kHz (1 wavelength between the tweeter and midrange) and should not be higher. However, I have modelled some gentler asymetrical slops for the two drivers and I believe the XO would sound excellent but with a XO point at 2.6kHz. That is too high in theory. I am not sure if I will try that.

Do you use MTM? how does your 8531 sound with 2.4kHz XO? at what slop?

Regards,
Bill

Thanks for the reply Bill.

I just run the single driver in partnership with the S2905 and M22WR and find the sound exceptional with great inner detail and natural tonal quality..... a lot due to the OB. The driver is still however acoustically loaded with some damping material. Power handling is not a problem even with the rated 60W so 1 driver is enough. I'm not huge on the volume control but do like a decent SPL.

Checked my info and the xo is just under 2600Hz (not 2400Hz... sorry my mistake) and 2nd order acoustic slopes with a tad less on the woofer. I did try some low xo points down to 1800Hz and higher orders on the tweeter but didn't sound as good for me.... tended to suck out some of the life. In theory, I'm running them too high but in real life it sounds better for my tastes.

I'm surprised at the amount of bass the 8531 was producing in the OB and is aided by a series resistor which brings the mid level down a tad but helps the bottom end. The odd bass I was referring to was in a vented enclosure and could never get it right but in OB, don't need it as the M22WR does the job. I did try a 22W8851T00 but was bettered by the M22WR in this instance but would have been better with steeper slopes but still not as good IMO. The M22WR is a freak and better than it has a right to be.

Sounds like yours is going to be a cracker. Good luck with it.
 

Attachments

  • rzss.jpg
    rzss.jpg
    53.8 KB · Views: 3,336
planet10 said:


You could also just skip the crossover and go with a 1-way.

dave

..and just how long does it stay a "1-way"?;)

Having also gone that route, I can personal say that each time it progressed to something more complex.

AKA - a 1-way presents some excellent qualities, and a LOT of compromises.

I would say though that it is a good place to start as well.. BUT to get it right "down the road" your back where you started with respect to "commitment" to the hobby and dealing with a crossover. Moreover even if it stays a "1-way" you still have multiple issues to deal with that are not related to designing and implementing a crossover.
 
Ceibel -

You say you didn't see anything amongst suggested designs you liked.

You're also asking if you wasted your money by purchasing a woofer tester and are asking which design software you should be purchasing.

I'm not trying to be offensive here, only helpful ......

If your statements are true and questions relate to your degree of understanding what makes you believe that even if you owned software you could design your own crossovers for drivers of your choice?

These guys have electrical engineering backgrounds, understand theory and are computer geniuses in comparison to most of us to boot. It's not a matter of plugging in some equipment and then feeding in your desired box size and TS parameters of drivers 'you think' would sound good and a little ticket popping out of a printer with a statement reading ' you have just designed a loudspeaker that sounds better than high end 5K units available'. There are several 'freeware' programs available that you could try without making an expenditure of any kind . Based upon the questions you ask I'm thinking your attempting to use them might convince you this stuff isn't learned overnight. If you truly understood all you need to understand to navigate these programs I'm certain you'd find you'd be spending a good deal more on equipment alone than the target price of the speakers you want to beat. Start pricing what good software costs and just how many different programs are involved in all aspects of design. Alot of guy's here are far more than weekend hobbyists.

Very simply, Loudspeaker design is primarily Physics. Guys who took General Math or Algebra will get lost soon after the starting gate. I'm in that group of the lost.

This Study is extensive and attempting to duplicate the sound of a loudspeaker you liked without duplicating that speakers drivers and that designers XO is virtually impossible .

I looked at the suggested designs given and if your statement that none of these were what you were after is true than it appears you have definite drivers or a definite appearance you desire to acheive in mind or both.

Tell the guys here exactly what you want if such is the case. You have truly been shown some excellent designs regards quality of sound. Google Curts Speaker Projects for a few more.

I can't help you as after a year of studying I've realized I'll never be a designer , the maths, theory and application would take me 4 years of full time study with a personal tutor before I could actually begin. Building isn't exactly easy either, you should see some of the garbage that's come out of my workshop to date and I do have some background there as well as a pretty decent workshop. Even MDF gets expensive when you start blowing angle cuts by a degree or 2.

Tell the Guru's what you want, pride cometh before the fall.

Bluto
 
frugal-phile™
Joined 2001
Paid Member
ScottG said:
..and just how long does it stay a "1-way"?;)

If you want a true 11-octave full range speaker it does need a helper woofer or tweeter or both, but the XOs are active at the bottom and simple at the top.

I've been happily going back to a 65-16k FR for the last 5 years. Still working on getting the helpers to help out and not get in the way. Just queued up are a FR as mid-tweeter and a purpose designed active helper woofer. I know the woof is capable of close to 20, and the mid-tweeter (according to the maker) is good to 32k ... will be interesting. The hard part will be the XO, but it will be active so that makes it both easier & harder.

The one thing a good FR does right is the critical mids where 90% of the music is.

dave
 
Bluto said:
Ceibel -

You say you didn't see anything amongst suggested designs you liked.

You're also asking if you wasted your money by purchasing a woofer tester and are asking which design software you should be purchasing.

I'm not trying to be offensive here, only helpful ......

If your statements are true and questions relate to your degree of understanding what makes you believe that even if you owned software you could design your own crossovers for drivers of your choice?

These guys have electrical engineering backgrounds, understand theory and are computer geniuses in comparison to most of us to boot. It's not a matter of plugging in some equipment and then feeding in your desired box size and TS parameters of drivers 'you think' would sound good and a little ticket popping out of a printer with a statement reading ' you have just designed a loudspeaker that sounds better than high end 5K units available'. There are several 'freeware' programs available that you could try without making an expenditure of any kind . Based upon the questions you ask I'm thinking your attempting to use them might convince you this stuff isn't learned overnight. If you truly understood all you need to understand to navigate these programs I'm certain you'd find you'd be spending a good deal more on equipment alone than the target price of the speakers you want to beat. Start pricing what good software costs and just how many different programs are involved in all aspects of design. Alot of guy's here are far more than weekend hobbyists.

Very simply, Loudspeaker design is primarily Physics. Guys who took General Math or Algebra will get lost soon after the starting gate. I'm in that group of the lost.

This Study is extensive and attempting to duplicate the sound of a loudspeaker you liked without duplicating that speakers drivers and that designers XO is virtually impossible .

I looked at the suggested designs given and if your statement that none of these were what you were after is true than it appears you have definite drivers or a definite appearance you desire to acheive in mind or both.

Tell the guys here exactly what you want if such is the case. You have truly been shown some excellent designs regards quality of sound. Google Curts Speaker Projects for a few more.

I can't help you as after a year of studying I've realized I'll never be a designer , the maths, theory and application would take me 4 years of full time study with a personal tutor before I could actually begin. Building isn't exactly easy either, you should see some of the garbage that's come out of my workshop to date and I do have some background there as well as a pretty decent workshop. Even MDF gets expensive when you start blowing angle cuts by a degree or 2.

Tell the Guru's what you want, pride cometh before the fall.

Bluto

I understand where you are coming from. I am doing the necessary reading and as I said before, I am just getting started. However, I do have a bit of help here. A friend of mine is an electrical engineer at Texas Instruments. That is the good part, the bad is I can't seem to get enough of his time to really help at this point.

I have started reading the Speaker Building 201 book and also have the Loudspeaker Cookbook. I learned last night that choosing the woofer is much more complex than looking at a few bits of data and price alone. So, I will be taking some time to figure out which driver to go with first for this box I have chosen, then comes the tweeter. Once I have accomplished that I will work my way into the crossover, hopefully by then I can get my TI buddy to sit down and help me do the math. It appears I have a long way to go before I can even start to build this set of speakers. I am not worried about the money so long as I accomplish my goal.
 
To explain what I am wanting is this: To build a set of speakers that have a high level of detail with out harsh highs. I like the highs to be smooth and natural. The drivers should provide a good level of bass without loss of detail to the music. This is why I am looking at the SEAS Excel drivers. However, I have yet to do the math as to if they will work in the enclosure I have chosen.
 
Ceibal said:
To explain what I am wanting is this: To build a set of speakers that have a high level of detail with out harsh highs. I like the highs to be smooth and natural. The drivers should provide a good level of bass without loss of detail to the music. This is why I am looking at the SEAS Excel drivers. However, I have yet to do the math as to if they will work in the enclosure I have chosen.


The BAMTM should do that.

If you are still bent on designing your own then the drivers for your application that I'd recommend are:

Usher 8945A X 2 in an appropriate bass reflex design (i.e. a lower freq. tunning than "aligned", probably a little lower than what is referred to as an extended bass shelf design).

Neo 3 PDR and a lower freq. crossover..

Note Dennis Murphy's statement on this:

"I had not realized the Neo planar tweeter could be crossed as low as 2000 Hz, and doing so definitely makes for a more enjoyable speaker."

If designed properly, it should be what you are looking for. (i.e. that "5k beater".)
 
ScottG said:



The BAMTM should do that.

If you are still bent on designing your own then the drivers for your application that I'd recommend are:

Usher 8945A X 2 in an appropriate bass reflex design (i.e. a lower freq. tunning than "aligned", probably a little lower than what is referred to as an extended bass shelf design).

Neo 3 PDR and a lower freq. crossover..

Note Dennis Murphy's statement on this:

"I had not realized the Neo planar tweeter could be crossed as low as 2000 Hz, and doing so definitely makes for a more enjoyable speaker."

If designed properly, it should be what you are looking for. (i.e. that "5k beater".)


I am really set on a closed design. Would this woofer work in that setup? Looks to have to low of a QTS according to what I am reading.
 
diyAudio Member
Joined 2004
Ceibal said:



I am really set on a closed design. Would this woofer work in that setup? Looks to have to low of a QTS according to what I am reading.

The efficiency bandwidth product(EBP) is a guideline and because of this you'll find some drivers work well in circumstances that the EBP would lead you to believe otherwise.

My favourite example would be the Peerless XLS10 with its Qtc of 0.19. Despite that it works well, maybe even arguably best, in a sealed enclosure.
 
diyAudio Member
Joined 2004
ScottG said:



The BAMTM should do that.

If you are still bent on designing your own then the drivers for your application that I'd recommend are:

Usher 8945A X 2 in an appropriate bass reflex design (i.e. a lower freq. tunning than "aligned", probably a little lower than what is referred to as an extended bass shelf design).

Neo 3 PDR and a lower freq. crossover..

Note Dennis Murphy's statement on this:

"I had not realized the Neo planar tweeter could be crossed as low as 2000 Hz, and doing so definitely makes for a more enjoyable speaker."

If designed properly, it should be what you are looking for. (i.e. that "5k beater".)

Parts Express have the Usher on special at $89 so that's a good call Scott.

I'm not sure about the Neo 3 PDR but no denying its low distortion. For the tweeter I'd look at the new Seas 27TBCD/GB-DXT or as its more easily remembered part number of H1499. In particular is its impressively linear off axis response and the ability to cross fairly low - I'd imagine 1.5Khz to be no real problem for everyday use. This would be a boon for a 2-way MTM. I'm particularly interested in trying out this tweeter for myself.

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


Plot shows 30 and 60 degrees offaxis:
An externally hosted image should be here but it was not working when we last tested it.
 
ShinOBIWAN said:


The efficiency bandwidth product(EBP) is a guideline and because of this you'll find some drivers work well in circumstances that the EBP would lead you to believe otherwise.

My favourite example would be the Peerless XLS10 with its Qtc of 0.19. Despite that it works well, maybe even arguably best, in a sealed enclosure.


Is there a good way to figure out if a woofer will work best in a sealed enclosure or vented? Or, is it just one of those things you just have to try it and see if it works.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.