Finally installed both W6 speakers, ditched the prototype boxes. Even this is still not finished - the right speaker is still on experimental XO with wago connectors, the left box needs reconditioning and some wool felt.
One wall is curved, another is bent behind the monitor to save space. The top and back are also tilted, no parallel walls, almost no 90 degrees anywhere. The walls are 5 layers - 4 layers of plywood and 1 layer of damping plastic in between. Only some polyfill, not much. No resonances.
Not ready to post a sound review, as I have several tweaks still planned. The only thing I can say the sound is in the ballpark of the best I ever heard in the showrooms or anyone's home, competing with live unamplified music. Imaging is extremely good, and the sound is very detailed and musical.
I strongly recommend those drivers, I have no idea if there is a better coaxial but the idea of a coaxial in a nearfield bookshelf is validated again. I don't think I would master a 2-way without years of experience.
Be careful, this driver requires at least 60W or so, at least the way I did it.
As I was unable to achieve a perfectly flat response with XO alone, this is what I get after spicing it a bit with Equalizer APO software. As explained above, equalizer or DSP alone are unable to suppress driver resonances, for this a physical inductor in series with driver is required. Moreover, I've learned that for example the Mini DSP upsampler has problems with interpolation distortion, so I would suggest to stay away from any DSP unless you know for sure it is very sophisticated. Having said that, apparently the properly configured Equalizer APO is not damaging the sound, and my experience is positive so far. As my XO is not very far from flat only some gentle equalization was needed.
The boxes are sealed, so going down to 45 Hz is not bad. The bass is very tight and fast, no glass shaking punch. I would say gentle is also an appropriate word.
Last edited:
More instructions for noobs by a noob. I only discovered this video today, but it describes exactly the problem I had with TB W6 and the solution is exactly the one I did, except I came to it by my own. Watch it first:
The problem Danny had is exactly the same: TB W6 driver cannot play above 2.5 kHz cleanly, and the tweeter has the resonance frequency at 1 kHz, and can be damaged unless the crossover has a sharp tweeter rolloff from 3 kHz down. There is no space to cross them, you either get a bad sound from the woofer or damage the tweeter, pick your poison.
The main trick Danny did was to sacrifice the speaker sensitivity. I don't remember the exact numbers, but essentially Danny said: it was a bad 93 Db speaker and I converted it into a good 85 Db speaker, because modern amps are super powerful and powering a 85 Db speaker is not only easy, but also affordable. Back in the 90s they couldn't do that, because there was no real affordable HiFi amp capable to power 85 Db at all, so they've being sacrificing sound quality for sensitivity.
If you scroll back to my actual crossover in post #518 it does exactly that - it adds a huge badass almost pound weight 2 mH 14 gauge inductor and then immediately another 0.4 mH one. This knocked down the frequency response by probably 5 Db, and then the notch filter knocked the 3 kHz area by another 5 Db. This probably made the woofer a 85 Db one, it requires a lot of power. If you read this thread nobody before me used a 2 mH inductor. What I got for this:
1. The cone breakup distortion at 3 Hz is completely gone, I cannot hear it, I cannot measure it
2. The low range of this woofer below 500 Hz was not actually affected by the huge inductors, so I've got the woofer to play down to 50 Hz in a tight sealed box - this is not ported
And you know that the way I use it - near the wall on the computer table 2 ft from my ears makes a ported box impossible, the port would degrade a sound quite a bit, being there before.
Now the tweeter becomes a lot louder than the woofer. To suppress it I had to put quite a brutal L-pad. The one pictured is not the one I actually have now - I have 3 Ohm L-pad parallel to driver and 1.2 Ohm in series. This made the tweeter circuit 2 Ohm, again, quite impossible in the 90s. But it opened the tweeter quite a bit - I don't have a huge resistor in the signal path, the sound is still quite effortless, a lot of air and imaging and instrument separations are great. So I still recommend this driver to noobs, but be aware of all those instructions.
The problem Danny had is exactly the same: TB W6 driver cannot play above 2.5 kHz cleanly, and the tweeter has the resonance frequency at 1 kHz, and can be damaged unless the crossover has a sharp tweeter rolloff from 3 kHz down. There is no space to cross them, you either get a bad sound from the woofer or damage the tweeter, pick your poison.
The main trick Danny did was to sacrifice the speaker sensitivity. I don't remember the exact numbers, but essentially Danny said: it was a bad 93 Db speaker and I converted it into a good 85 Db speaker, because modern amps are super powerful and powering a 85 Db speaker is not only easy, but also affordable. Back in the 90s they couldn't do that, because there was no real affordable HiFi amp capable to power 85 Db at all, so they've being sacrificing sound quality for sensitivity.
If you scroll back to my actual crossover in post #518 it does exactly that - it adds a huge badass almost pound weight 2 mH 14 gauge inductor and then immediately another 0.4 mH one. This knocked down the frequency response by probably 5 Db, and then the notch filter knocked the 3 kHz area by another 5 Db. This probably made the woofer a 85 Db one, it requires a lot of power. If you read this thread nobody before me used a 2 mH inductor. What I got for this:
1. The cone breakup distortion at 3 Hz is completely gone, I cannot hear it, I cannot measure it
2. The low range of this woofer below 500 Hz was not actually affected by the huge inductors, so I've got the woofer to play down to 50 Hz in a tight sealed box - this is not ported
And you know that the way I use it - near the wall on the computer table 2 ft from my ears makes a ported box impossible, the port would degrade a sound quite a bit, being there before.
Now the tweeter becomes a lot louder than the woofer. To suppress it I had to put quite a brutal L-pad. The one pictured is not the one I actually have now - I have 3 Ohm L-pad parallel to driver and 1.2 Ohm in series. This made the tweeter circuit 2 Ohm, again, quite impossible in the 90s. But it opened the tweeter quite a bit - I don't have a huge resistor in the signal path, the sound is still quite effortless, a lot of air and imaging and instrument separations are great. So I still recommend this driver to noobs, but be aware of all those instructions.
Last edited: