Attention! HH got those distortion measurements completely wrong. The correct ones are published in HH 3/2013 on page 80. The correct distortion is roughly 10 times higher than in your posted diagram.
Rudolf
I've had no opportunity this week to do some more measurements. I hope I can do it this week.
But I really need to decide what mid I'm going to use. The more I look, the more I find. For example, the Visaton B80 looks also very nice. It's not very cheap and performance is probably comparable with the Dayton RS100 (with cost about half). I also looked at Audax HM100Z0. It has high efficiency, and is supposed to be one of the best mid's out there. But sadly, I could not find any objective tests (HH 2001/1 has one, but the issue is not available anymore). Objective information about the PA units I found is also tricky...
So looking at all this, I guess I have two options left:
- MTM with Dayton RS100-8
- MT with Visaton Ti100
The second option is about twice as expensive. But looking at Zaphs site, it shows that the Ti100 has about 6 tot 10 dB less distortion in the usable range, which means a single Ti100 will probably beat two RS100's. So It might be worth the money?
Mean while, my wife and I have been listening to the current experimental setup a lot. One OB, and my old BR speaker to complete the stereo image
Even now, the soundstage is amazing and it really is a joy to listen to it.
But I really need to decide what mid I'm going to use. The more I look, the more I find. For example, the Visaton B80 looks also very nice. It's not very cheap and performance is probably comparable with the Dayton RS100 (with cost about half). I also looked at Audax HM100Z0. It has high efficiency, and is supposed to be one of the best mid's out there. But sadly, I could not find any objective tests (HH 2001/1 has one, but the issue is not available anymore). Objective information about the PA units I found is also tricky...
So looking at all this, I guess I have two options left:
- MTM with Dayton RS100-8
- MT with Visaton Ti100
The second option is about twice as expensive. But looking at Zaphs site, it shows that the Ti100 has about 6 tot 10 dB less distortion in the usable range, which means a single Ti100 will probably beat two RS100's. So It might be worth the money?
Mean while, my wife and I have been listening to the current experimental setup a lot. One OB, and my old BR speaker to complete the stereo image
Even now, the soundstage is amazing and it really is a joy to listen to it.
So today I finally had the time to do some more tweaking, and also do off-axis measurements. Here they are:
As you can see, the NEO is very pronounced off-axis, with a of-axis bump between 2 and 3 Khz.
I also had some difficulty finding the correct x-over point between tweeter en woofers. The woofers are now at 2200 LR48, and the tweeter at 1200 LR48 I tried lots of stuff, but couldn't get it right.. The strange thing is, inverting the tweeter doesn't change a thing.. So any idea's on this are very welcome!
As you can see, the NEO is very pronounced off-axis, with a of-axis bump between 2 and 3 Khz.
I also had some difficulty finding the correct x-over point between tweeter en woofers. The woofers are now at 2200 LR48, and the tweeter at 1200 LR48
I tried lots of stuff, but couldn't get it right.. The strange thing is, inverting the tweeter doesn't change a thing.. So any idea's on this are very welcome!Attachments
You are just facing the inevitable problem, when the woofer loses it's dipole directivity too early, because of the width of the baffle. The Tweeter doesn't have directivity that low, the result is what you measured (inconsistent directivity between 2 and 4 kHz)
The polarity invert test doesn't have any effect because slopes are not symmetric enough and propably also time alignment (adjust delay in minidsp) isn't correct. First you must look at acoustic responses of M and T as an overlay and adjust slopes/xo so that they make "by book" overlap, then invert polarity and adjust delays. it is very sensitive, usually a difference of just 0,02 - 0,06ms is enough. You must try which driver needs delay (the one that has it's acoustic center closer to the mic)
I guess that you noticed that I have decided to try Audax AM100Z0 for upper mids! I haven't received them yet, but they should perform rather well for my intended range. They have a problem of ringing (CSD) at 5kHz but I cross lower-
The polarity invert test doesn't have any effect because slopes are not symmetric enough and propably also time alignment (adjust delay in minidsp) isn't correct. First you must look at acoustic responses of M and T as an overlay and adjust slopes/xo so that they make "by book" overlap, then invert polarity and adjust delays. it is very sensitive, usually a difference of just 0,02 - 0,06ms is enough. You must try which driver needs delay (the one that has it's acoustic center closer to the mic)
I guess that you noticed that I have decided to try Audax AM100Z0 for upper mids! I haven't received them yet, but they should perform rather well for my intended range. They have a problem of ringing (CSD) at 5kHz but I cross lower-
Last edited:
Ah yes, I should have thought about time alignment.. I'll try that.
Did you find any measurements for the HM100Z0? Funny that we both found that one, I didn't know you were also looking at it until I read your thread about the AINOgradient
Currently I have my mind set on the Ti100 though, if I can get it for a nice price.
Did you find any measurements for the HM100Z0? Funny that we both found that one, I didn't know you were also looking at it until I read your thread about the AINOgradient
Currently I have my mind set on the Ti100 though, if I can get it for a nice price.
About xos, first you must measure each driver individually and then do an overlay. The acoustic xo is most important, but naturally in minidsp both lower and upper must have very nearly same Fs and acoustic slopes must be symmetric. But there are exceptions, eg. linear phase approach usually with odd order or unsymmetric slopes. But LR12, LR24, LR48 should show nulling when polarity is reversed for other driver. (LR12 needs inverted polarity for normal operation)
Examples of measured slopes and phase behaviour
Examples of measured slopes and phase behaviour
Attachments
Last edited:
All cone speakers have a heart-like directivity pattern,based on the angle of cone and diameter. My AINOs have 3" (75mm) cones as upper mids and thy show this problem at 2kHz, where "beaming" begins. This behaviour makes a bump in directivity. I wish that the Audax beheaves better, it has a bullet-shaped dust cap (or phase plug) that sould help a little. The real solution is a planar driver.
The first pic is from here http://www.klippel.de/fileadmin/kli...ure/Papers/KLIPPEL_Sound_Radiation_Poster.pdf
and the other is a from a speaker driver's datasheet, pay attention to response at 60¤ Cone geometry related problems start at 1/4 of diameter relative frequency. The 130mm wide cone at 600Hz. 75mm cone at 2kHz.
The first pic is from here http://www.klippel.de/fileadmin/kli...ure/Papers/KLIPPEL_Sound_Radiation_Poster.pdf
and the other is a from a speaker driver's datasheet, pay attention to response at 60¤ Cone geometry related problems start at 1/4 of diameter relative frequency. The 130mm wide cone at 600Hz. 75mm cone at 2kHz.
Attachments
Yes it is possible! But I don't believe it sounds very good! If we just feed several drivers with same signal we get huge diffraction and phase problems. A single wide range speaker is worlds better, but again it has different problems!
As a matter of fact I have not had a possibility to seriously listen to a well designed speaker with planar drivers. Just at hifi shows and shops.
My growing consiousness of factors influencing loudspeaker basics, measurements of off-axis behaviour is making my diy-life more difficult. At this point i find that we have:
Some small room designs that are generally sane (but all of them have flaws!)
- unity horn, corner horn
- well done 3-4 -way speakers monopole, cardioid or dipole no matter which but all of them are different
- Geddes/Econowave style 12-15"W + horn+compression driver
- 4-5" + waveguided tweeter minimonitor (or coaxial) + multisub system
I dislike
- single fullrange no matter monopole or dipole
- a set of horns or any other widely spread system
I see potential in but with serious problems
- full range planars, stacked planars (ESL, Magnepan..)
- Keele style CBT
As a matter of fact I have not had a possibility to seriously listen to a well designed speaker with planar drivers. Just at hifi shows and shops.
My growing consiousness of factors influencing loudspeaker basics, measurements of off-axis behaviour is making my diy-life more difficult. At this point i find that we have:
Some small room designs that are generally sane (but all of them have flaws!)
- unity horn, corner horn
- well done 3-4 -way speakers monopole, cardioid or dipole no matter which but all of them are different
- Geddes/Econowave style 12-15"W + horn+compression driver
- 4-5" + waveguided tweeter minimonitor (or coaxial) + multisub system
I dislike
- single fullrange no matter monopole or dipole
- a set of horns or any other widely spread system
I see potential in but with serious problems
- full range planars, stacked planars (ESL, Magnepan..)
- Keele style CBT
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.