Fixed without a cap but volume is a bit high, I don't think I will ever listen that loud so perhaps XO must be optimized for the volume I need.
I opened mine and I just realized I sent you the measurements I did when the W6 were in XPS boxes, hence the jagged impulse.
So, please, don't use those. Sorry.
I'm not going to be home for quite a while, so no luck sending you good files.
Your box is sealed, right? Because I find it strange to see two impedance bumps. Are you sure you sealed the box well? Might be due to air leakage.
So, please, don't use those. Sorry.
I'm not going to be home for quite a while, so no luck sending you good files.
Your box is sealed, right? Because I find it strange to see two impedance bumps. Are you sure you sealed the box well? Might be due to air leakage.
Very likely not sealed well, thanks for pointing that impedance is apparently an amazing trick to check for leakages! I did 3-layer walls, 1/2" + 1/4" + 1/4" MDF, all with hand saw, so it's pretty complex to do at home. I did solid wood baffle and angled bracing. As a bonus the box is very stiff and heavy. And probably leaking.
Thanks @perceval for your help! As you don't have impedance files around and mine look incorrect, I think I will do the 1st round and assemble the XO post #191 as is from cheapest parts and see what happens. When I have a variant I like I'll bump the parts up from the quality batch. A bit easier with caps because I can stack them in parallel to get the desired value.
Mine was a work in progress, and the best sound was from my last XO that I posted.
When doing it half way, and changing along the way, it is usually more expensive than going all in at the first try.
Mine was a work in progress... feel free to follow the progress or just in at the end result.
Not saying it can't be improved. I'm sure it could, but I'm very happy with the way they sound now.
Enjoy the ride! 🙂
When doing it half way, and changing along the way, it is usually more expensive than going all in at the first try.
Mine was a work in progress... feel free to follow the progress or just in at the end result.
Not saying it can't be improved. I'm sure it could, but I'm very happy with the way they sound now.
Enjoy the ride! 🙂
I saw your last XO of course, but my problem it's serial. I have Schiit Aegir amp, which is 20W on 8 Ohm and 40W on 4 Ohm. My current Dynaudio speakers are 6 Ohm and the amp is at the border of sliding onto higher distortions at higher volume. When I dial the knob to shake the room I just hear a lot of bad sounds. The reason I picked the TB driver is because it's 4 Ohm and I want to keep it so.
However I admit I am only guessing your serial XO is above 6 Ohm, it's not an educated knowledge.
However I admit I am only guessing your serial XO is above 6 Ohm, it's not an educated knowledge.
I found some in the back corner of my computer... couldn't tell you from which measurements right now. I'm more organized on my backups, but they are at home. Well, the other place I call home sometimes! 🙂
But... mine is a ported box, so it has the double peak.
And, that XO would be a little dark for some. Myself, I like a downward curve of about 7dBs between peak bass and treble. Especially listening so close.
But... mine is a ported box, so it has the double peak.
And, that XO would be a little dark for some. Myself, I like a downward curve of about 7dBs between peak bass and treble. Especially listening so close.
Thank you! My hearing is black zero at 16.5 kHz and barely whisper audible at max volume 16 kHz. So my goal is to bump the 13 kHz area 6 db up and 16 kHz area 12 db up. I will hear that flat. I think realistically speaking I need separate filters so I can measure speakers separately. I plan to keep the XO externally forever, so I can experiment with it anytime without even moving the speaker from its place near the wall.
My hearing starts dropping at 12kHz, and I still like a downward curve.
The acoustic center is not zero, but pretty close.
One way to measure is to do a woofer sweep, a tweeter sweep, then a sweep with both of them.
Plug in the individual drivers in XSim.
Add another "driver" which is the sweep of both of them.
Change the distance of the tweeter until the FR matches the "both of them" sweep.
The acoustic center is not zero, but pretty close.
One way to measure is to do a woofer sweep, a tweeter sweep, then a sweep with both of them.
Plug in the individual drivers in XSim.
Add another "driver" which is the sweep of both of them.
Change the distance of the tweeter until the FR matches the "both of them" sweep.
The ultimate weirdness of sound perception. While I'm waiting my XO parts from parts express I'm listening a weird combo - one speaker is my TB W6 prototype with $8 3.5 kHz high pass on twitter and nothing on woofer (1st chart), another is Dynaudio Emit 10 (right chart). And I actually find this pair very enjoyable - quite cleaner bass than Dyns alone, a lot more details in the highs, and yet I don't really hear the huge gap in the mids at TB - perhaps the second speaker is filling it enough. The bass is clean but soft as if an elephant sized cat is dancing. The highs are supposed to be fatiguing but they are not, even after several hours. I'm already at 100 hours of this combo and I don't want to switch the proto speaker back to Dyn. From the measurements, it should sound horrible:
BTW I don't really understand why woofer is dropping that much after 1 kHz. You can see the woofer alone in post #422.
BTW I don't really understand why woofer is dropping that much after 1 kHz. You can see the woofer alone in post #422.
Wow thank you! Why is that, because acoustic offset happened to be exactly at those frequencies to make the interference in the opposite phase? Now I need another 100 hours with that...
Phase interference, yes.
I wouldn't get too accustomed to a +10dB plateau from 1kHz and on.
Everything after that will sound too dark.
You need to pad the tweet to bring it down, so the speaker can be more balanced.
I wouldn't get too accustomed to a +10dB plateau from 1kHz and on.
Everything after that will sound too dark.
You need to pad the tweet to bring it down, so the speaker can be more balanced.
Looks like a great improvement (I have had similar phasing problems with my DSP based system), Looks like significant attenuation (10 - 12dB) is still needed on the tweeter, which might then still be too bright. (Look at some examples of the Harman target curves for an idea of what might sound good in your space.)
Ideally you would start a very gradual roll-off around 1kHz and be down 3dB or more at 20kHz. (room and your preferences play a role here)
Ideally you would start a very gradual roll-off around 1kHz and be down 3dB or more at 20kHz. (room and your preferences play a role here)
Last edited: