junk@ur-tut.UUCP (Jan Vandenbrande) (03/07/86)
.... I believe many of us have thought about designing speakers because we ran out of money, thought we good do a better job, like to tool around etc. I was inspired by a talk from one of the engineers from the KEF Co. A very interesting talk, lasting about 2.5 hours, with transfer functions, filter design, enclosure design etc. I was rather amased how envolved it got. Still not yet discouraged I kind of looked around how much it would cost in hardware alone to make my own speakers. It added up to more than what you would buy in a store. Of course I would have used nicer wood, better cables, a filter of my own design (I mean a crossover netwerk, which is nothing more than a filter) good quality equipment stuff etc, etc. But the danger is that you don't know what the final result will be until you connect the speaker to the amplifier. Add to this the amount of work and design that goes into making all this. If you don't have access to a whole range of instruments (pressure mikes [or whatever they are called], netwerk analysers, spectrum analysers, ossiloscopes, signal generators, anechoic rooms, some machine where you can do circuit simulation [SPICE for ex.], and a whole lot more) you are really walking in total darkness, and making wild guesses. The result can be better than what you buy in a store, but it could be much worse too. I am not saying that it is totally impossible. I am sure alot of people tried this and had satisfactory results. But if you want superior results it might end up a bit involved (and may be a lot of fun too). One trick you could use is to "immitate" a speaker you like. This way you already have a good idea of geometric set up and how to make the enclosure which is far from trivial if you were to design this from scratch. Don't forget that these speaker co.'s usually did a certain amount of research and have alot of know-how you don't have. Good luck, keep us posted when you have results: Jan.
tommie@psivax.UUCP (Tom Levin) (03/11/86)
In article <33@ur-tut.UUCP> junk@ur-tut.UUCP (Jan Vandenbrande) writes: >.... > I believe many of us have thought about designing speakers >because we ran out of money, thought we good do a better job, like >to tool around etc. > (....) >Still not yet discouraged I kind of looked around how much it would >cost in hardware alone to make my own speakers. It added up to more >than what you would buy in a store. I must disagree. I built a pair of two-way monitors using the best of everything I could find (i.e. Audax bextrene mids, Audax tweeters, mylar caps, "Live Wire", etc.) and the total cost to me (including cabinets and veneer) was under $200.00 a pair! These monitors compare quite favorably (actually better in my opinion) to the Rogers LS35A's. A good friend of mine designed these little beauties (he has also designed amps, preamps, subwoofers, etc.). They have a 3rd-order Butterworth crossover, Thiel-Small cabinet alignments, rounded front edges (reduces diffraction), 3/4" high density particle board, fuses for both the mids and tweets, etc., etc. etc. The bottom line is that they sound EXCELLENT! I am incredibly happy with them! The best part is that *I* built them and I did a more careful job than just about any other speaker manufacturer would do (my opinion). >(....) But the danger is that you don't know what the final >result will be until you connect the speaker to the amplifier. That's why when you are designing you need to have the crossover outside the cabinet so you can fiddle the theoretical values to where they sound the best (at least to your ears). >Add to this the amount of work and design that goes into making all this. >If you don't have access to a whole range of instruments (pressure >mikes [or whatever they are called], netwerk analysers, spectrum analysers, >ossiloscopes, signal generators, anechoic rooms, some machine >where you can do circuit simulation [SPICE for ex.], and a whole lot more) >you are really walking in total darkness, and making wild guesses. Geeeez! What a pessimist ;-). Seriously though, it is a blast to make your own speakers and I wouldn't worry all that much about all them there fancy electronic equipment. Even if you only use the theoretical cabinet volume and "text-book" crossover circuit, the result will probably be better than 99% of the mid-fi junk out there (have you ever looked at the inside of a Fisher loudspeaker????). >The result can be better than what you buy in a store, but it could be >much worse too. Unless you are talking about GOOD speakers (read "Kef", "Rogers", "Linn", etc.) I seriously doubt you could do much worse. >I am not saying that it is totally impossible. I am sure alot of people >tried this and had satisfactory results. But if you want superior results >it might end up a bit involved (and may be a lot of fun too). Building my own speakers was the most rewarding project of my life. >One trick you could use is to "immitate" a speaker you like. Not such a bad idea actually. My friend who designed my monitors did just that when he designed the subwoofer for the satellites. He "immitated" the Audio-Pro B-250 subwoofer (a self amplified beast with 2 Philips 6-1/2" woofers mounted in a "push-pull" configuration). It is probably better than the Audio-Pro model (It has a much more powerful amp of his own design with high quality components) and only cost about $300.00 total (compared to over $1000.00 list for the A.P). >Don't forget that these speaker co.'s usually did a certain amount of >research and have alot of know-how you don't have. ummmm........maybe....naw! (1/2 :-)) To summarize my very biased and opinionated feelings: Most people would be astounded at the lack of science put into mass market speakers (and other stereo gear too). I can't believe the prices on some of these 3 way monsters (you know, the ones with the 15" woofers!). I have looked inside and seen (with my own eyes) 1st order crossovers, or worse yet, straight wire!!!! Also, flimsy, resonating cabinets, inferior speakers, etc. Well, enough rambling. I'm sure this article will create a lot of replies (but hopefully not too many flames). Let the speaker builing begin! -- __________________________________________________________________________ Tom Levin {ttidca|sdcrdcf|logico|scgvaxd|bellcore}!psivax!tommie "Turn it up!!!"
mohan@topaz.RUTGERS.EDU (Sunil Mohan) (03/12/86)
The latest issue of STEREOPHILE has an article on building a speaker,
designed by the author of that article, at home. It handles 50w with a
response of 60 - 18kHz. The projected cost for a pair is between $200
and $300. The article is followed by a list of places to buy material from,
and also a very favourable review of the finished product. It supposedly
compares in quality to speakers costing in the market upto four times its
price.
--
Sunil
UUCP: ...{harvard, seismo, ut-sally, sri-iu, ihnp4!packard}!topaz!mohan
ARPA: Mohan@RUTGERSjunk@ur-tut.UUCP (03/19/86)
... This is in replie to the replie to my previous message: In article <1055@psivax.UUCP> tommie@psivax.UUCP (Tom Levin) writes: >... >cabinets and veneer) was under $200.00 a pair! These monitors compare >quite favorably (actually better in my opinion) to the Rogers LS35A's. >A good friend of mine designed these little beauties (he has also >designed amps, preamps, subwoofers, etc.). They have a 3rd-order >Butterworth crossover, Thiel-Small cabinet alignments, rounded front >edges (reduces diffraction), 3/4" high density particle board, >fuses for both the mids and tweets, etc., etc. etc. The bottom line >is that they sound EXCELLENT! I am incredibly happy with them! >The best part is that *I* built them and I did a more careful job >than just about any other speaker manufacturer would do (my opinion). I am very glad that you obtained such a good result. There is indeed nothing like a job well done. But please tell us how your friend started the initial design. >>(....) But the danger is that you don't know what the final >>result will be until you connect the speaker to the amplifier. > >That's why when you are designing you need to have the crossover outside >the cabinet so you can fiddle the theoretical values to where they sound >the best (at least to your ears). Several comments: 1] You were already able to listen to similar speakers made by your friend. So you more or less knew the result. 2] Yes, finetuning the filters is a very good idea. But you fine tuned by ear I presume.Is this reliable? I guess you tune it to the sound you like best, which is of course what you ultimately want. 3] Unfortunately I must disagree that tuning your filters is the only parameter. How about annoying enclosure resonance? Can that be eliminated by tuning the filter? >>Add to this the amount of work and design that goes into making all this. >>If you don't have access to a whole range of instruments (pressure >> .... >>you are really walking in total darkness, and making wild guesses. > >Geeeez! What a pessimist ;-). Seriously though, it is a blast to make >your own speakers and I wouldn't worry all that much about all them >there fancy electronic equipment. Even if you only use the theoretical >cabinet volume and "text-book" crossover circuit, the result will >probably be better than 99% of the mid-fi junk out there (have you >ever looked at the inside of a Fisher loudspeaker????). Yes I have looked inside a Fischer loadspeaker, and I must agree that it really isn't too difficult to do much better than that. >>The result can be better than what you buy in a store, but it could be >>much worse too. > >Unless you are talking about GOOD speakers (read "Kef", "Rogers", "Linn", etc.) >I seriously doubt you could do much worse. Actually yes, I did have GOOD speakers in mind. >>One trick you could use is to "immitate" a speaker you like. > >Not such a bad idea actually. My friend who designed my monitors did >just that when he designed the subwoofer for the satellites. He >"immitated" the Audio-Pro B-250 subwoofer (a self amplified beast >with 2 Philips 6-1/2" woofers mounted in a "push-pull" configuration). >It is probably better than the Audio-Pro model (It has a much more powerful >amp of his own design with high quality components) and only cost about >$300.00 total (compared to over $1000.00 list for the A.P). > >>Don't forget that these speaker co.'s usually did a certain amount of >>research and have alot of know-how you don't have. > >ummmm........maybe....naw! (1/2 :-)) The "better" co's probably do (I know KEF does), but others, hmmm, hmmm! > [...] Well but do tell me a few more things. How did you design your Butterworth filter? From scratch or "Text book solution?". Did you actually measure if the filter worked correctly? How about the phase during cross over? Butterworth filters are pretty uniform in phase characteristic, but at crossover it is not unlikely to have two speakers working out of phase. Did you consider a phase correcting (all pass) filter to correct this? Does it matter? And what is your final result? How flat is your responce and what is the impedance of your final result along the frequency range? How about the efficiency? Do all these wonderfull characteristics matter at all to what sound quality you get? Well if you have other usefull tricks, please do share them Jan.