Purifi Drivers for Open Baffle

Wow,

@CharlieLaub
these measurements are really helpful, thanks a lot!
So directivity on the front side doesn't seem to be a problem up to 2kHz. That's great!
But I would have expected a more similar response between front and rear in this frequency range. I even have expected, that compared to my post #9 (https://www.diyaudio.com/community/threads/purifi-drivers-for-open-baffle.390740/post-7134793), which shows front vs. rear for the smaller Scanspeak 12MU it would be less an issue with a bigger driver, but this was obviously a false assumption.

The 1.3kHz peak should be the normal dipole peak, or not?
Not a problem with active EQ I assume.
 
The 1.3k Hz peak is indeed the dipole peak. It's often higher Q than what is predicted by models. This might be due to the geometry of the cone, the influence of the basket, or what not. To the rear it is very likely the basket and motor that start mucking up the response above the dipole peak. I have seen this sort of thing many, many times for nude drivers of all different diameters.

With a few different types of equalization via DSP you can make the entire response very flat up until breakup. This also has the effect of linearizing the phase within the passband, and because of that I keep on going to at least 1.5 octaves below where I plan to cross over the driver to the woofer. This is a typically a little steeper than 6dB/oct when you get well below the dipole peak and closer to Fs, but since I use e.g. 4th order crossover filters the crossover kills that relatively quickly. I have a paper coming out in audioXpress in (fingers crossed) the April issue that described a system that I build recently and goes into detail about how I measure, EQ the drivers flat, apply the crossover filters, where I cross over and why.
 
The 1.3k Hz peak is indeed the dipole peak. It's often higher Q than what is predicted by models.
I also made the experience, that Edge calculates a slightly higher peak, than what I measure.

To the rear it is very likely the basket and motor that start mucking up the response above the dipole peak. I have seen this sort of thing many, many times for nude drivers of all different diameters.
Unfortunately, open baffles are still somehow an exotic use case. Hence, measurements from the rear side are very rare. I'm afraid, it will be very difficult to convince guys like @HiFiCompass or Erin to introduce rear measurements into their reviews 😎

May be it is really the open structure of my 12MU, which makes it look good up to 2.7kHz in terms of front/rear similarity. Will a Scanspeak 18WU be at the end the really deal for an open baffle, if you are going for a 3-way?
It has a similar size like the PTT6.5, very high excursion as well and the open structure of basket and magnet like the 12MU.

Anybody here who can do a front rear comparison for this driver? 🙂
 
  • Like
Reactions: Juhazi
Unfortunately my project was delayed while I figured out how to configure the bass section, and then winter hit. Since I need to make outdoor measurements I have to wait until the weather permits this, which around here will not be until April at least. This doesn't impact how I feel about the PTT6.5M. It's superior IMD performance puts it head and shoulders above other drivers as a wide band midrange in my nude OB designs.
Thank you @CharlieLaub and @Matty!

@charliel
How did you end up configuring your bass section? I also settled on sealed subs and I am now revisiting options for my midbass. Currently Faital 18hp1010 in U-cab, sounds a little mild on attack, and exploring other options. Xo at 80 and 300Hz.
 
I am experimenting with a compound dipole arrangement for the woofer section. I did some measurements on two nude drivers at various separation distances sometime last year or the year before (posted the data somewhere here on DIYaudio) and have decided to give it a try for real. But quite a few measurements are needed before I arrive at the final design for that part of the project.
 
Sorry OP for the tangent. I won't continue for long. I would actually welcome your input too in light of your AE Dipole18.

@CharlieLaub
Could you please elaborate on the compound dipole woofers?
I remember your 15" dipole midbass. What is your compound design solving for vs the 15"?

I've been trying to decide between a AE LO15 for midbass and XRK's SLOB using multiple 6". Maybe your compound design becomes a third alternative to consider? 🙂
 
I have OB speakers with PTT6.5W08-01A in MTM. Some measurements and images can be seen here: https://www.diyaudio.com/community/threads/amt-pro-4-as-ob-tweeter.366280/post-6754661
Thanks for sharing this.

Can I ask you why you opted for MTM?

I also do MTM but with B&C 8PE21, and I was after achieving an efficiency high enough to use SET amps. But lately I'm reconsidering SET for mids and then could live with one mid, efficiency-wise.

How did you like the sound of PTT6.5M and AMT? I have AMT too, Beyma in my case.
 
Sorry OP for the tangent. I won't continue for long. I would actually welcome your input too in light of your AE Dipole18
No worries. I‘m happy to have a vibrant discussion here 😉
And as we already found out, a Purifi open baffle needs Non-Purifi drivers as well, as Purifi drivers will not be able to cover the whole frequency range in a reasonable way. So no problem to talk about other drivers here.
What exactly do you want to know about the Dipole 18?
 
Thanks for sharing this.

Can I ask you why you opted for MTM?

I also do MTM but with B&C 8PE21, and I was after achieving an efficiency high enough to use SET amps. But lately I'm reconsidering SET for mids and then could live with one mid, efficiency-wise.

How did you like the sound of PTT6.5M and AMT? I have AMT too, Beyma in my case.
One of my reasons was also efficiency, even 2 PTT6.5W08-01A together are actually less sensitive than other speakers I use. Other reason was narrower vertical polar response, what helps to reduce floor reflections, third was that virtually radiation point of tweeter and mids are same, without placing speakers in same point physically. PTT6.5W08-01A sounds good, very natural, also it has no lack of Xmax when compensating some of OB roll-off, in my case lower cut off is at about 160 Hz and compensation at lowest point is about 12 dB. With my modest listening SPL levels I had not seen/measured any distortion increase causd by cone amplitude Xmax reach. PTT6.5W08-01A H-frames have dimensions what create first dipole null at about 2 kHz and high cut off is at 1 kHz.
I like AMT and generally ribbon tweeters. Specially good is that AMTPRO4 tweeter can be cut at 1 kHz, this give good integrity of mid range and high range sound. AMT is also only type of ribbon tweeters what can be used as OB with good results. AMTPRO4 has some problems in the 15+ kHz range and top end is not same easy and open as have real ribbon tweeters, but I can manage that little with some DSP correction.
 
Last edited:
To the rear it is very likely the basket and motor that start mucking up the response above the dipole peak. I have seen this sort of thing many, many times for nude drivers of all different diameters.
Do you have any measurements from other drivers than the above mentioned PTT6.5M for that available?
These would be very valuable and helpful. 😀
Especially drivers, which will go as high as 2000Hz or more and still show good dipole behavior.
 
There is ONE cone type driver that I have found can be used above 2kHz, and that is the SB Acoustics SB17MFC. I have attached the front response (with a nude driver mounting). I can't find a similar multi-axis plot of the responses measured at the rear, but I attached the on-axis comparison of front and rear. To the rear there is only a gentle downslope above 1.5kHz compared to in front, which is no big deal.

As you can see from the front response measurements, the dipole peak is smooth and broad and the usable response extends up to just over 3kHz before you reach the first breakup peak, which isn't all that large. If you use a steep crossover and knock down the breakup peaks using DSP you can cross over above 2kHz.

As Juhazi mentioned, if you really need to cross over higher than 2k then you should probably look at large planars like the GRS PT5010. The only downside is that its directivity is increasing around about 2kHz, but the response on axis is usable up to almost 10k. Not usable below about 600Hz unless you put it in some baffle, and then that ruins the high end.
 

Attachments

  • SB17MFC35-4 nude 0-20-40-60 deg.png
    SB17MFC35-4 nude 0-20-40-60 deg.png
    21.1 KB · Views: 136
  • SB17MFC35-4 nude front vs rear-green.png
    SB17MFC35-4 nude front vs rear-green.png
    17.1 KB · Views: 128
  • Like
Reactions: Matty and Juhazi
If your baffle is huge, maybe. Actually, funny enough a really, really large planar OB can be a great solution. By really large I mean like 3feet or more wide and 5 feet or more tall. Just get a 4x8 sheet of plywood, stiffen the edges with some 2x4s and put it up on some large L-brackets. There ya go. When the baffle is that large, the response starts to not see anythign from the rear and you get something more like an infinite baffle type speaker. This means very little in the way of dipole cancellation, and you can use drivers much as if they are in a closed box. But unfortunately that tends to negatively impact many of the good things about open baffle systems, namely imaging, etc.

IMHO no 3" driver can do any sort of good sounding reproduction at 120Hz, no matter how many of them you use. It will be producing 10 percent distortion, more if you are unlucky. No thanks. Case in point:
https://www.audiosciencereview.com/...research-lgk-2-0-speaker-review-a-joke.34783/
 
Last edited:
  • Like
Reactions: Juhazi
I am experimenting with a compound dipole arrangement for the woofer section. I did some measurements on two nude drivers at various separation distances sometime last year or the year before (posted the data somewhere here on DIYaudio) and have decided to give it a try for real. But quite a few measurements are needed before I arrive at the final design for that part of the project.
When you say compound are you referring to a clamshell configuration?
Edit; sorry posted before reading the rest of the the thread!
 
Last edited:
What exactly do you want to know about the Dipole 18?
Thanks for the comments.

My bad: in reality what I wish I found is someone who has listened to both an AE Dipole or LO15 or 18 and a slot-loaded open baffle (SLOB) like @xrk971
did with 8x6" per side. My Faital 18" in OB lack a bit of slam above 80Hz (sealed subs below that), and XRK reports the SLOB has nice impact for an OB. I'm deciding if I should go through the hassle of importing a couple of LO15 vs going down the SLOB path.

This is what I'm after now. Next will be the midrange and these PTT6.5M look really good for that! Thanks for starting the thread.