No, you did nothing wrong in showing measurements, to explain your point of view
But we really dont have to agree
Just recently I heard a scientist say that scientific proofs was merely a way to explain things we dont understand... well, something like that
How I choose a driver? ... well, I look at it, and well ... listen to it
most are disappointing, few are not ... but my rule is "if it doesnt work make it work"! ... but one thing is sure, if I dont like the look of a driver I wont waste time on it
btw, Lots of good stuff on Zaphs ... but I still dont find any distorsion graphs on finished designs ... is that not possible ?
But we really dont have to agree
Just recently I heard a scientist say that scientific proofs was merely a way to explain things we dont understand... well, something like that
How I choose a driver? ... well, I look at it, and well ... listen to it
most are disappointing, few are not ... but my rule is "if it doesnt work make it work"! ... but one thing is sure, if I dont like the look of a driver I wont waste time on it btw, Lots of good stuff on Zaphs ... but I still dont find any distorsion graphs on finished designs ... is that not possible ?
norcad:
as you already have drivers I wouldn't think of st. else... its not all about distortion numbers.
I would do a MTM system with two CA18 at least for front and center speakers. As Tweeter I would use modificated seas new 29TFF/W it is great tweeter and definitely best of all textil domes in seas prestige series(I testest all of them)
Thinking most economicly I would buy 3x29Tff/W and exchange magnet systems with those 27tffc or 27tdfc you have. You gain 29TFFC/W (tested- outperforms 27tdfc by a mile) and rest you can use/sell as 27TFF or 27TDF. Also you can play with damping material in entering of chamber(I used soft isover with great succes) You wrote you have arta so I hope you built its measuring box so you should be able to measure impedance with phase making you capable of doing such modifications to TW (you can play and match those TW perfectly)- definitely worth it!
if you do that you will have no problem crossing low. But I wouldn't go way low down to 1.4khz. TW can handle it without trouble but I would sacrifice distortion a bit to have "bigger sound" (my personal taste, you can try what suits you best) I would try 1.6-1.7khz LR4 accustic slopes
for rears I would use TM ca18+27tdfc(tweeked -exchanged seas damping for isover) crossed around 1.8khz (if money isn't problem->make rears same as front)
As sub I can recommend only L26rfxp. it will go realy deep if you can give it more volume than 50 liters. If you can't try to think of few L22rn4xp. these go deep in small BR enclosures (around 40liters each)
note: I have or had at home all mentioned speakers except for ca18 and I tried/measured/tweeked them. When you will be designing crossovers I would sacrifice "straight" response curve to better driveres phase concours.
You have tools so start to try my recomendations.
ps. I know my english is not the best...sorry for that but I hope you will understand what I mean.
as you already have drivers I wouldn't think of st. else... its not all about distortion numbers.
I would do a MTM system with two CA18 at least for front and center speakers. As Tweeter I would use modificated seas new 29TFF/W it is great tweeter and definitely best of all textil domes in seas prestige series(I testest all of them)
Thinking most economicly I would buy 3x29Tff/W and exchange magnet systems with those 27tffc or 27tdfc you have. You gain 29TFFC/W (tested- outperforms 27tdfc by a mile) and rest you can use/sell as 27TFF or 27TDF. Also you can play with damping material in entering of chamber(I used soft isover with great succes) You wrote you have arta so I hope you built its measuring box so you should be able to measure impedance with phase making you capable of doing such modifications to TW (you can play and match those TW perfectly)- definitely worth it!
if you do that you will have no problem crossing low. But I wouldn't go way low down to 1.4khz. TW can handle it without trouble but I would sacrifice distortion a bit to have "bigger sound" (my personal taste, you can try what suits you best) I would try 1.6-1.7khz LR4 accustic slopes
for rears I would use TM ca18+27tdfc(tweeked -exchanged seas damping for isover) crossed around 1.8khz (if money isn't problem->make rears same as front)
As sub I can recommend only L26rfxp. it will go realy deep if you can give it more volume than 50 liters. If you can't try to think of few L22rn4xp. these go deep in small BR enclosures (around 40liters each)
note: I have or had at home all mentioned speakers except for ca18 and I tried/measured/tweeked them. When you will be designing crossovers I would sacrifice "straight" response curve to better driveres phase concours.
You have tools so start to try my recomendations.
ps. I know my english is not the best...sorry for that but I hope you will understand what I mean.
Norcad,
My post re a L26/L15 and either 27TBFC/G or 27TDFC/G design might help (well - regarding L26 and tweeter parts anyway)....
http://www.diyaudio.com/forums/showthread.php?s=&threadid=106791
I am a little worried about 7mm xmax limits on the L26, so will go ported around 90L and implement a line level highpass around 30Hz to prevent unloading (I'll let you know exact FR one I design the circuit / post the excursion graph on the above link).
The above simulation has 240Hz and 1725Hz xo points - which would suit these drivers well by all accounts.
Cheers,
David.
My post re a L26/L15 and either 27TBFC/G or 27TDFC/G design might help (well - regarding L26 and tweeter parts anyway)....
http://www.diyaudio.com/forums/showthread.php?s=&threadid=106791
I am a little worried about 7mm xmax limits on the L26, so will go ported around 90L and implement a line level highpass around 30Hz to prevent unloading (I'll let you know exact FR one I design the circuit / post the excursion graph on the above link).
The above simulation has 240Hz and 1725Hz xo points - which would suit these drivers well by all accounts.
Cheers,
David.
JAY_WJ: I'm going to try lower Xo, and measure it! Very interesting how you designing crossover, many good points in your guide, but I will stick to measurements, and simulate from there.
BTW: Why are you so interested in measurement (exp dist), when you dont have measurments on your own design? Maybe you also have something to hide?
I havent tested so many drivers, and dont know if theres a big difference in soundquality reg to distortion, or if its just your "golden hifi ears" that here it. In many cases there can be rather bad measurements of a speaker, but still sound rather good (in my ears). I lack some experience in this matter, I think.
BTW: Why are you so interested in measurement (exp dist), when you dont have measurments on your own design? Maybe you also have something to hide?
I havent tested so many drivers, and dont know if theres a big difference in soundquality reg to distortion, or if its just your "golden hifi ears" that here it. In many cases there can be rather bad measurements of a speaker, but still sound rather good (in my ears). I lack some experience in this matter, I think.
kyselym: How is the 29TFF outperforming 27TFFC/27TDFC? Distortion? Top octave?
The tweak is to take the frontplate, diagraph and voice coil from the 29TFF and put it on the magnet system from 27TDFC, have I understood you right?
What is isover?
Maybe the 29TFF sounds good, but it looks bad
I have not build the measure box for the Arta yet. I build the Wallin jig for Speaker Workshop, but did not get the calibrating to work. My impedance measurements did not get right.
Why can you only recommend L26? Because its best, or because you havent tested the CA26?
PS Your english are better than mine, so I understand it perfectly
The tweak is to take the frontplate, diagraph and voice coil from the 29TFF and put it on the magnet system from 27TDFC, have I understood you right?
What is isover?
Maybe the 29TFF sounds good, but it looks bad
I have not build the measure box for the Arta yet. I build the Wallin jig for Speaker Workshop, but did not get the calibrating to work. My impedance measurements did not get right.
Why can you only recommend L26? Because its best, or because you havent tested the CA26?
PS Your english are better than mine, so I understand it perfectly
Dave Bullet: My first idea was also an all aluminium Seas, similar to yours! After a while I understood that those drivers isnt the right choice for a newbie like me.
I have tested the L26 in 50L closed, (qtc 0,6) and had some high hopes when I simulated the freq resp. But I was very dissapointed because it reached x-max at once. It only needed about 5w@30Hz to get the 7mm linear excursion.
I'll tried vented also, in the same enclosure, and that was much better. But still it had some problems around 40-45Hz
The CA26 is better in that enclosure, play louder, and isnt reaching x-max, but dont go that low in frequency.
In my finish speaker I will have about 60L for the bass I think. If i build something bigger, my wife will throw them out (and maybe me also)
I have tested the L26 in 50L closed, (qtc 0,6) and had some high hopes when I simulated the freq resp. But I was very dissapointed because it reached x-max at once. It only needed about 5w@30Hz to get the 7mm linear excursion.
I'll tried vented also, in the same enclosure, and that was much better. But still it had some problems around 40-45Hz
The CA26 is better in that enclosure, play louder, and isnt reaching x-max, but dont go that low in frequency.
In my finish speaker I will have about 60L for the bass I think. If i build something bigger, my wife will throw them out (and maybe me also)
Norcad said:Why are you so interested in measurement (exp dist), when you dont have measurments on your own design? Maybe you also have something to hide?
Good question. I don't hide anything! Perhaps, it seems to you that my method of designing crossovers doesn't rely on measurements. But it's the other way around in some sense. The main reason why this method is possible is that I can rely on nice measurements done by other extremely serious hobbyists such as Zaph and Mark K. Based on their distortion measurements I can choose drivers and crossover points, and more. And thanks to Zaph's consistent and meticulous infinite-baffle measuremets in each of his driver test groups, I can achieve consistency among different designs. Indeed, without their measurements, my method is completly infeasible. See what I mean?
Also, as for the voicing process I described in my page, I don't claim any "golden hifi ear." I simply use my ear to fine-tune the speakers, which is done by any serious speaker designer. This is totally different from what some hi-fi nuts say to claim that only they have a "golden ear" to judge good sounding speakers. Voicing is an important step in speaker building. And a good trained ear is helpful in this process. That's what I state in my page.
When designing speakers and xo, there are so many parameters that can affect the result. I dont say that your method isnt right, I dont say that your speakers doesnt sound good.
But there are several software and measurements from others you have to trust that are correct. And like you say, you have to use them carefully and correct.
Thats why I bought some measure equioment. (and thought that it would be fast an easy))
I have learned that measuring, and getting the correct results from simpel measure equipment, isnt that easy either.
I have measured 4 complete well known speakers, and 12 different drivers, and compared the result with measurements from others. At the end I was able to get a good match on the freq measurements, but my impedance measurements are not accurate.
I think that in the end, it will be expereince that makes thing easier and with better result. This will be my first total DIY speakers, and I hope that when they are finished, they sound OK. And after a couple of years with tweeking, they will sound great!
But there are several software and measurements from others you have to trust that are correct. And like you say, you have to use them carefully and correct.
Thats why I bought some measure equioment. (and thought that it would be fast an easy))
I have learned that measuring, and getting the correct results from simpel measure equipment, isnt that easy either.
I have measured 4 complete well known speakers, and 12 different drivers, and compared the result with measurements from others. At the end I was able to get a good match on the freq measurements, but my impedance measurements are not accurate.
I think that in the end, it will be expereince that makes thing easier and with better result. This will be my first total DIY speakers, and I hope that when they are finished, they sound OK. And after a couple of years with tweeking, they will sound great!
Good approach... just believe in yourself and what you do
But dont wait too long, or the drivers will be "rusty" before you get it done
Just get started, and you can have fun tweaking fore the next 5years, and at the same time learn some of the secrets, but expect some degree of frustration from time to time...well, a lot actually
But dont wait too long, or the drivers will be "rusty" before you get it done
Just get started, and you can have fun tweaking fore the next 5years, and at the same time learn some of the secrets, but expect some degree of frustration from time to time...well, a lot actually
Well, I dont believe in my self, and what I do now, but I believe in experience!
Thats why I like this forum, where there are several people with more or less experience, who can give me help in, hopefully, the right direction.
Its fun to learn some new things, but my brain are getting old, so it takes a while.......
Thats why I like this forum, where there are several people with more or less experience, who can give me help in, hopefully, the right direction.
Its fun to learn some new things, but my brain are getting old, so it takes a while.......
norcad: I test driver first and if it is ok than you can modify it, so you don't have to worry about waranty, there is nothing to screw up. This proces is reversable because only thing you do is: you unmount front plate(also with voice coil; those 4 screws) the rest is magnet(+chamber) and damping in entering of chamber= black foam material originaly from seas. this you can easily remove and give your own - better. I use isover DOMO 14 ( http://www.tepelna-izolace.cz/data/mod_eshop/84/mo/main/large-isover-domo.jpg ) if you use it only for tweeters you can use "thinner"-> DOMO 10 or 12
I use arta measuring box for impedance and phase with big succes and I can guid you to measure properly(but after reading short help you can do it on your own I think). Its box is very simple...
I have to admit that I didn't like the look of 29tff firstly...but when I kept it in my hand and looked at it in baffle I began to like it. 29TFFC/W is great speaker because it has waveguid, chamber and biggest sd among all seas tweeters. this helps it very much under 2kHz. With proper damping you clean all mess the driver produces and gain very neutral and exact sound. Also it has very good top octave. I'm not able to produce distortion graphs like zaph so no proof here but according to not only my ears its at least the same as 27tdfc (but I'm conviced its much better)
if you really don't want to try 29TFFC/W because of looks. Stick to 27tdfc and tweek it with damping. According to it's Sd I wouldn't go such low, as zaph, in your design. I would stay slightly under 2Khz. Not sure of sound of CA18 but L18 and P18 would produce awesome sound crossed with this tweeter at this frequency. Personaly I don't think L-series of speakers is difficult to crossover for beginer...it just needs RLC compensations If you waveguid the 27tdfc like zaph did in one of his designs I would go a bit lower with crossing. As I said earlier...you can go even now with it down to 1.5khz but such a system wouldn't sound "big" because of small area of 27tdfc for such low frequences.
If you decide to go for different drivers my pick would be P18+29TFFC/W in MTM or if you would like smaller diameter go for L15 and 27tbfcg or 29tafcW
It seems you don't have much volume for sub so I suggest to look for st else than CA26 and L26.
yes I never had ca26 in hands so can't compare but does not seem to be better for such purpose.
I suggest to build arta box and try tweekes I suggested. If you can't sell ca18 without much loss I would keep them. As distortion is not the best it will still play good I think
ps. all this discusion about crossover points is just recomendations...Maybe you will be unable to cross like this properly(I mean with good frequency response and phase alignment at once)
I use arta measuring box for impedance and phase with big succes and I can guid you to measure properly(but after reading short help you can do it on your own I think). Its box is very simple...
I have to admit that I didn't like the look of 29tff firstly...but when I kept it in my hand and looked at it in baffle I began to like it. 29TFFC/W is great speaker because it has waveguid, chamber and biggest sd among all seas tweeters. this helps it very much under 2kHz. With proper damping you clean all mess the driver produces and gain very neutral and exact sound. Also it has very good top octave. I'm not able to produce distortion graphs like zaph
so no proof here but according to not only my ears its at least the same as 27tdfc (but I'm conviced its much better) if you really don't want to try 29TFFC/W because of looks. Stick to 27tdfc and tweek it with damping. According to it's Sd I wouldn't go such low, as zaph, in your design. I would stay slightly under 2Khz. Not sure of sound of CA18 but L18 and P18 would produce awesome sound crossed with this tweeter at this frequency. Personaly I don't think L-series of speakers is difficult to crossover for beginer...it just needs RLC compensations
If you waveguid the 27tdfc like zaph did in one of his designs I would go a bit lower with crossing. As I said earlier...you can go even now with it down to 1.5khz but such a system wouldn't sound "big" because of small area of 27tdfc for such low frequences.If you decide to go for different drivers my pick would be P18+29TFFC/W in MTM or if you would like smaller diameter go for L15 and 27tbfcg or 29tafcW
It seems you don't have much volume for sub so I suggest to look for st else than CA26 and L26.
yes I never had ca26 in hands so can't compare but does not seem to be better for such purpose.
I suggest to build arta box and try tweekes I suggested. If you can't sell ca18 without much loss I would keep them. As distortion is not the best it will still play good I think
ps. all this discusion about crossover points is just recomendations...Maybe you will be unable to cross like this properly(I mean with good frequency response and phase alignment at once)
kyselym said:
Thanks for the good information, again, kyselym.
But I have trouble understaning what you mean by "big sound." Would you explain what it means technically or more specifically?
If sound at a certain fixed frequency is generated from different sources that have different sd's and different cone/dome shapes, it may have different off-axis behaviors in a room. But at a fixed listening position, if it is at the same level of SPL and has exactly the same hamonic distortion patterns, the sound shouldn't be different by any means. That's what I think.
Seas standard 1" tweeters (TDFC and TBFC/G) have excellent low-end distortion performance. If power handling isn't an issue (it's not an issue, actually), I don't see any problem crossing them low like at 1.5 kHz.
Any other factors I need to consider to understand what you mean by "big sound?"
-Jay
Norcad said:
I understand that. You want to use your own measurement instead of trusting others' plus simulations. It perfectly makes sense. One more thing to consider in your case is that the baffle you want to use is not conventional. A current baffle diffraction simulator may not generate an accurate prediction. I'd use my own measurement, too, if I were you.
But even if you obtain accurate measurements, they are not the only thing you need to consider to achieve the goal of having "good-sounding" speakers. As you said in some of your last posts, you'll need some experience. I think what I said in my page is appropriate to quote here:
"... ... even if you have very accurate, measured system FRs both on-axis and off-axis, usually you don't have ability to translate the measurements into actual listening experience unless you are a professional or an experienced DIYer who has voiced speakers many, many times via listening and measurement feedback. Look for a flat on-axis, or listening window average response? Or a smooth power response with a certain target curve? Useful standards. But there's no speaker that has perfectly flat or perfectly smooth responses. Every speaker has its unique response with dips and bumps at idiosyncratic frequencies. Also its nonlinear distortion characteristics are unique. This is a fundamental reason why voicing is important and required. It's not optional. Even professionals do not simply interpret measurement and finalize a project. Voicing via listening is a must."
-Jay
my technical english and abilities to describe sound are pretty bad but I will try...
in midrange area it same like in bass area...if you have bigger cone area you gain bigger accustic pressure.
if I should describe sound I would call it more stressed if you let small driver play low. It is better to let bigger one play a bit higher.
I think it is about achieving good accustic pressures at your listening positions. For example if you want to listen to orchestra at live perfomace volume levels... I don't think you would like it with 2way system where 27tdfc is crossed at 1.5khz. maybe now it is understoodable to get my point? or maybe its more about of my(and few others) personal taste...you are free to try I can only say that I had chance to compare similar 2way with bigger system with bigger cone areas and I liked it more.
anyway if I listen to my 2way with L18+27tbfcg with crossover point around 1.7khz(if I remember well) it still sounds very,very well even at higher(bit higher than normal-not concert) volume levels.
ps.I don't want to say in any way that zaph's designs are bad(if somebody got this feeling). They are pretty good!
in midrange area it same like in bass area...if you have bigger cone area you gain bigger accustic pressure.
if I should describe sound I would call it more stressed if you let small driver play low. It is better to let bigger one play a bit higher.
I think it is about achieving good accustic pressures at your listening positions. For example if you want to listen to orchestra at live perfomace volume levels... I don't think you would like it with 2way system where 27tdfc is crossed at 1.5khz. maybe now it is understoodable to get my point?
or maybe its more about of my(and few others) personal taste...you are free to try I can only say that I had chance to compare similar 2way with bigger system with bigger cone areas and I liked it more.anyway if I listen to my 2way with L18+27tbfcg with crossover point around 1.7khz(if I remember well) it still sounds very,very well
even at higher(bit higher than normal-not concert) volume levels.ps.I don't want to say in any way that zaph's designs are bad(if somebody got this feeling). They are pretty good!
I know what you mean. But if going beyond max capable output is the problem, then it should be observable in distortion measurements. According to Zaph and Mark K's measurements, the Seas tweeters' low-end distortions (e.g., at 1 kHz) is better than most midwoofers even at high SPL. As I said in a post earlier, if you use them crossed at 1.5 kHz with LR4 slopes, distortion or compression due to high SPL is not a matter. Zaph also talked about this in his BAMTM design page.
I think it's your personal preference.
I think it's your personal preference.
kyselum, you make perfectly sense
Jay, "I am all ears"
My experience match with kyselum regarding higher xo point ... actually I like it much higher
remember that a driver dont just play inside its passband, but should also work equally well outside the passband, some say the stopband
Also I am convinced that small components are way better than big ones ... besides I dont like to "shut down" a driver, but prefer to let it run loose, and instead use corrections to make it possible
To get a coherent sound I work from the principle that a mid and a tweeter should play with the same sound nature on both sides of the xo point ... as should the woofer sound like the mid
With a very low xo point it will be difficult to make the tweeter and mid sound the same
Jay, "I am all ears"
My experience match with kyselum regarding higher xo point ... actually I like it much higher
remember that a driver dont just play inside its passband, but should also work equally well outside the passband, some say the stopband
Also I am convinced that small components are way better than big ones ... besides I dont like to "shut down" a driver, but prefer to let it run loose, and instead use corrections to make it possible
To get a coherent sound I work from the principle that a mid and a tweeter should play with the same sound nature on both sides of the xo point ... as should the woofer sound like the mid
With a very low xo point it will be difficult to make the tweeter and mid sound the same
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.