I am focusing more these days more aspects of speaker design, like verifying box tuning.
Here is a DATS sweep of the box as built.
Fb appears to be right at 40Hz, which is where I wanted it - WinISD model showed about 38Hz Fb, but I don't care about 2 Hz. A slightly more concerning aspect is the bumps in the impedance. Probably resonances in the TL plenum. Hornresp shows I can smooth them down with more stuffing. We shall see once I've measured FR.
But Chahly - Stahkist don't want speakers that look good, Stahkist wants speakers that sound good!
Hmm, one of Scott Hinsons criticisms for TLs is that the terminus "leaks" a lot of midrange. I never paid much attention to this even though Hornresp shows port output way into the midrange.
This is the Hornresp sim:
And here is the near miked port in the box as built:
Oh well, the box is all put together, so I can't fix this, but next build I will know....
But Chahly - Stahkist don't want speakers that look good, Stahkist wants speakers that sound good!
Continuing in my exploration of terminus "mid" leakage.
The total line length is about 150cm, and that should be about 250Hz,
which is where the first "bump" in port output occurs.
I first tried some pyramid acoustic foam I had in stock at the origin.
Pretty much no change in a terminus measurement.
I tried a little open cell foam in the exit, and as expected, pretty much killed the bass.
So then tried a bunch of Dow703 compressed fiberglass at the origin
Interesting - it affected Fb magnitude, but it also shifted Fb a little lower.
Comparing the absolute magnitude at Fb versus the leakage bumps,
The bumps at 250 and 500 are a little diminished (2-3 dB)
Looks like that is the best I can do at this time, since the box is glued up.
I've got several more TL experiments planned once I finish these.
But Chahly - Stahkist don't want speakers that look good, Stahkist wants speakers that sound good!
@rjj45 said:
Continuing in my exploration of terminus "mid" leakage.
I've got several more TL experiments planned once I finish these.
The 'leakage' as show from ~200 upward in your graphs look much better behaved with the fiberglass. With the extension down towards 28Hz for the line design I'd be very happy (but then would consider using my miniDSP to trim down the peaks).
@rjj45 said:
Continuing in my exploration of terminus "mid" leakage.
I've got several more TL experiments planned once I finish these.
The 'leakage' as show from ~200 upward in your graphs look much better behaved with the fiberglass. With the extension down towards 28Hz for the line design I'd be very happy (but then would consider using my miniDSP to trim down the peaks).
Thanks buddy -
I popped in a test crossover this afternoon and played some music. Bass is pretty impressive for a 176mm driver, but I already knew that from other Rival 176 builds.
A miniDSP would not help in this instance, the frequency response from the drivers is nice and flat. The problem is what happens to the back side radiation and out the terminus.
Coming up, I've got an identical box, but with a removable side so I can alter the plenum behavior - I'm thjinking that perhaps lining the plenum with 1/2 inch foam will kill the leakage and resonance.
But Chahly - Stahkist don't want speakers that look good, Stahkist wants speakers that sound good!
Interesting set of measurements, Don! Looks like it is simply the longest line dimension that keeps reflecting back no matter what you do with reasonable amounts of stuffing along the line or elsewhere. That was my conclusion as well with my dual chamber experimental transmission line project. I tried very hard to solve the problem, cut came up empty. Good luck at trying to find a solution!
I like your foam idea and the removeable sides. I have been thinking lately that I should make a test box with removeable sides and document the results of what I've discovered so far.
@rjj45 said:
The problem is what happens to the back side radiation and out the terminus.
Point taken (line response). But then there is the total / net room response (listening area) of the combined outputs. Guessing much smoother than just the line - and impressive low end extension from the line, with a bit less steep drop than a ported system ?
@rjj45 said:
The problem is what happens to the back side radiation and out the terminus.
Point taken (line response). But then there is the total / net room response (listening area) of the combined outputs. Guessing much smoother than just the line - and impressive low end extension from the line, with a bit less steep drop than a ported system ?
Yeah - that terminus will be on the back side, so much of the >200Hz sound will be mitigated naturally.
Bass on TLs is something I'm still investigating. At best, it usually measures (standard sweep) just like the ported box model, but to me (and others), it just sounds better. May be woofer damping by the stuffing or tone burst behavior. But I like the sound. Can go deeper than a ported box without unloading.
But Chahly - Stahkist don't want speakers that look good, Stahkist wants speakers that sound good!
Jeez - you never know when life will get in the way. After the last update here, we bought a house, then had months of contractor work. Finally got back to finishing the crossover on these.
First, the unfinished, untrimmed baffle on the unfinished box.
and fast forward to the somewhat finished design during measurements today:
Birch baffle finished with water based poly - and teak veneer on the carcass:
and 2 others.
But Chahly - Stahkist don't want speakers that look good, Stahkist wants speakers that sound good!
I literally wasted months trying to use ARTA to make dual channel measurements for input to Vituicad. I had continued problems with ARTA (never figured that out), and just dropped back to Omnimic and knocked it out in short order.
Originally i had the perception that the bass was just a little light. Modified my first test crossover to give about 4 dB or BSC and everything came together.
Four basic curves:
I worked at tuning a nice reverse null, and in the process created a classic BBC dip:
I actually really like this sound, prevents any glaring in the midrange
But Chahly - Stahkist don't want speakers that look good, Stahkist wants speakers that sound good!
Thanks guys. I had not used the XT25 beast mode or a waveguide before, and I like both.
The Rival, of course is powerful and detailed.
I always feel nonplussed when imaging cosmetic choices, glad y'all like them.
But Chahly - Stahkist don't want speakers that look good, Stahkist wants speakers that sound good!
Comments
I am focusing more these days more aspects of speaker design, like verifying box tuning.
Here is a DATS sweep of the box as built.
Fb appears to be right at 40Hz, which is where I wanted it - WinISD model showed about 38Hz Fb, but I don't care about 2 Hz. A slightly more concerning aspect is the bumps in the impedance. Probably resonances in the TL plenum. Hornresp shows I can smooth them down with more stuffing. We shall see once I've measured FR.
Hmm, one of Scott Hinsons criticisms for TLs is that the terminus "leaks" a lot of midrange. I never paid much attention to this even though Hornresp shows port output way into the midrange.
This is the Hornresp sim:
And here is the near miked port in the box as built:
Oh well, the box is all put together, so I can't fix this, but next build I will know....
Continuing in my exploration of terminus "mid" leakage.
The total line length is about 150cm, and that should be about 250Hz,
which is where the first "bump" in port output occurs.
I first tried some pyramid acoustic foam I had in stock at the origin.
Pretty much no change in a terminus measurement.
I tried a little open cell foam in the exit, and as expected, pretty much killed the bass.
So then tried a bunch of Dow703 compressed fiberglass at the origin
Interesting - it affected Fb magnitude, but it also shifted Fb a little lower.
Comparing the absolute magnitude at Fb versus the leakage bumps,
The bumps at 250 and 500 are a little diminished (2-3 dB)
Looks like that is the best I can do at this time, since the box is glued up.
I've got several more TL experiments planned once I finish these.
The 'leakage' as show from ~200 upward in your graphs look much better behaved with the fiberglass. With the extension down towards 28Hz for the line design I'd be very happy (but then would consider using my miniDSP to trim down the peaks).
Thanks buddy -
I popped in a test crossover this afternoon and played some music. Bass is pretty impressive for a 176mm driver, but I already knew that from other Rival 176 builds.
A miniDSP would not help in this instance, the frequency response from the drivers is nice and flat. The problem is what happens to the back side radiation and out the terminus.
Coming up, I've got an identical box, but with a removable side so I can alter the plenum behavior - I'm thjinking that perhaps lining the plenum with 1/2 inch foam will kill the leakage and resonance.
Interesting set of measurements, Don! Looks like it is simply the longest line dimension that keeps reflecting back no matter what you do with reasonable amounts of stuffing along the line or elsewhere. That was my conclusion as well with my dual chamber experimental transmission line project. I tried very hard to solve the problem, cut came up empty. Good luck at trying to find a solution!
I like your foam idea and the removeable sides. I have been thinking lately that I should make a test box with removeable sides and document the results of what I've discovered so far.
Point taken (line response). But then there is the total / net room response (listening area) of the combined outputs. Guessing much smoother than just the line - and impressive low end extension from the line, with a bit less steep drop than a ported system ?
Yeah - that terminus will be on the back side, so much of the >200Hz sound will be mitigated naturally.
Bass on TLs is something I'm still investigating. At best, it usually measures (standard sweep) just like the ported box model, but to me (and others), it just sounds better. May be woofer damping by the stuffing or tone burst behavior. But I like the sound. Can go deeper than a ported box without unloading.
Jeez - you never know when life will get in the way. After the last update here, we bought a house, then had months of contractor work. Finally got back to finishing the crossover on these.
First, the unfinished, untrimmed baffle on the unfinished box.
and fast forward to the somewhat finished design during measurements today:
Birch baffle finished with water based poly - and teak veneer on the carcass:
I literally wasted months trying to use ARTA to make dual channel measurements for input to Vituicad. I had continued problems with ARTA (never figured that out), and just dropped back to Omnimic and knocked it out in short order.
Originally i had the perception that the bass was just a little light. Modified my first test crossover to give about 4 dB or BSC and everything came together.
Four basic curves:
I worked at tuning a nice reverse null, and in the process created a classic BBC dip:
I actually really like this sound, prevents any glaring in the midrange
Crossover came together pretty quickly.
Normalized horizontal polars:
Impressions:
Very smooth, very clean, very detailed. All my torture tracks show no issues.
Those look fantastic! Great work!
Sharp units!!
Really nice job! Looks cool too. Bonus!
Thanks guys. I had not used the XT25 beast mode or a waveguide before, and I like both.
The Rival, of course is powerful and detailed.
I always feel nonplussed when imaging cosmetic choices, glad y'all like them.
Love that tweeter. Much better than the OG XT25, which is no slouch.