Exploring Purifi Woofer Speaker Builds
- Thread starter mainframe99
- Start date
More or less, yes. The vertical spacing will have an impact, and the slope of the filters (2nd, 3rd, or 4th order) will have an impact. But you are correct.
Many fine speakers were designed and built before the concept of calculating and using Directivity Index, Predicted In-Room Response, Power Response, and Early Reflections became widely known. A set of good Power & DI curves will make a speaker easier to position in a room, more flexible in where it is located. But if a person is willing to place the speakers optimally, and do some room treatment if needed, the Power & DI performance becomes less relevant.
Nonetheless, with the tools we have today, speaker designers who offer their designs to the public should make the effort to manage the Power & DI performance of the design.
j.
Directivity Index has been my issue with the Purifi drivers from the start. All of this excellent power response, impedance and distortion are great. Why are they advertising an 8" should be used up to 5khz? It makes no sense to me. I don't care if a circuit can correct the breakup modes, there won't be any reflections that high!
@D1sco I don't recall Purifi says PTT8 should be used up to 5kHz. Lars did mention that it is clean to 2kHz. Even then, agreeing with what you said, the driver has become directional well before 2kHz. I am using mine at 400Hz right now matching with a PTT6.5 as a mid. Which honestly sounds very very good.
It is a very very promising combination. There are a few areas of concerns and compromises to be aware of.
You will need to cross the 2 at 1.2-1.5kHz. It is a compromise as the TW29BNWG distortion starts to rise pretty quickly below 1.6kHz or so. And on the other end, the PTT8 gets directional after 1kHz. Even then, you will find the PTT8 beats the TW29BNWG in distortion at the crossover point. The PTT8 is that good.
Your bass alignment will need some thoughts. Ported PTT8 works. A small-ish box around 30L will get down to 33Hz or so. You will have a very long port for that though, more than 1 foot in length even with 3". Purifi demo box has some trickery to minimize the issue that arises with such long ports.
As for passive radiator, which is the popular solution for the PTT6.5, I don't recall seeing anyone tried it with the PTT8 yet.
As the 3rd option, if you choose to do closed box, it will cut off around 50-55Hz. Perfect to transition to sub-woofer in a non-directional crossover frequency.
I have personally tried very close to that combination. TW29BNWG + Dual PTT8 ported. Dual coz I wanted to make sure it will have enough omph plus it cuts the port length in half. There are other issues to deal wish such a center-to-center distance between the TW29BNWG and the 2 PTT8. But honestly it sounds pretty good but obviously not technically perfect.
Both TW29BNWG and PTT8 have fairly flat response, so there won't be much correction needed so designing the crossover passive or active will be a breeze.
Note that the usual 2pi/IB directivity/off axis plots are great if you do inwall but can be misleading if the driver is in a box and measured in 4pi.
The box is causing a transition from omni 4pi in the bass to more directional approaching the 2pi data sheet response at HF. This transition is benefiting the tweeter a lot so that its DI is kept up lower down in frequency than what we would think from looking at the 2pi data sheet curves. The baffle dimensions can be used to optimise this. It has been very interesting to explore comsol simulations of drivers in a box in 4pi space.
For me it looks like there is a current trend to drive the xover lower than it needs to be. This causes more distortion from the tweeter and more diffraction ripple (tweeter being a smaller source with less spatial averaging).
cheers
Lars
The box is causing a transition from omni 4pi in the bass to more directional approaching the 2pi data sheet response at HF. This transition is benefiting the tweeter a lot so that its DI is kept up lower down in frequency than what we would think from looking at the 2pi data sheet curves. The baffle dimensions can be used to optimise this. It has been very interesting to explore comsol simulations of drivers in a box in 4pi space.
For me it looks like there is a current trend to drive the xover lower than it needs to be. This causes more distortion from the tweeter and more diffraction ripple (tweeter being a smaller source with less spatial averaging).
cheers
Lars
it is desirable to have clean response an octave or even two octaves above the XO frequency - depending on how sharp filters are used. The PTT8 can be crossed at say 1.5k and it has extremely low distortion and IMD well above.
How does the midrange from the PTT8 and from the 6.5 differ if used up to 1.2/1.5kHz? In the context of a 3 or 4 way system.
The 6.5 is meant as a midrange and reaches higher without beaming, while the 8 is meant as midwoofer? But if used to 1.3kHz where beaming issues are minimized for the 8, would the 6.5 still sound better?
The 6.5 is meant as a midrange and reaches higher without beaming, while the 8 is meant as midwoofer? But if used to 1.3kHz where beaming issues are minimized for the 8, would the 6.5 still sound better?
Not sure what with? comething I wrote?
indeed. However, the box design and WG design heavily influences the polar. Not enough to just look at 2pi data
I meant how does the PTT8 sound in the 300 to 1300Hz range (approx) vs the PTT6.5 in the same range? My tweeter reaches to 1200Hz with low distortion and approx 80° dispersion so wondering if I would loose midrange quality if I crossed to a PTT8 instead of a 6.5.
I take it you’ve seen @Yevgeniy’s review ?
https://www.diyaudio.com/community/threads/some-speaker-driver-measurements.317632/post-7237652
You can compare and contrast using the comparison feature at his website (registered user only) or open 2 tabs to compare and contrast.
Useful graph here-
Depending on your cabinet design, in-cabinet dispersion and tweeter’s in-cabinet dispersion a crossover from 1000-1800Hz looks like a reasonable start to play with; and depending on phase alignment at crossover region or other factors it might move up or down.
https://www.diyaudio.com/community/threads/some-speaker-driver-measurements.317632/post-7237652
You can compare and contrast using the comparison feature at his website (registered user only) or open 2 tabs to compare and contrast.
Useful graph here-
Depending on your cabinet design, in-cabinet dispersion and tweeter’s in-cabinet dispersion a crossover from 1000-1800Hz looks like a reasonable start to play with; and depending on phase alignment at crossover region or other factors it might move up or down.
Last edited:
