Hi All,
since reading about Purifi drivers, I'm thinking about using them in an open baffle system.
Those drivers show fantastic test results from @HiFiCompass , erinsaudiocorner.com etc.
Most of the parameters should be very nice for using them in OB systems: high xmax, very low distortion, even during high excursion, very low intermodulation distortion ...
Only downside I would see is the low sensitivity of some drivers, especially the X-Versions with the highest xmax. Of cause this is logical, but depending on your open baffle design, the sensitivity on the lower end can be - due to the dipole cancellation - even much lower than the value in the specs.
Now I'm asking myself, which could be the optimal Purifi driver for such an system for the midrange.
Actually, there is of cause a dedicated midrange driver, which looks very good, however, from the specs, in terms of distortion etc. the NAA-Versions of the PTT6.5 are on the same level.
So which driver would be the best choice then?
This is my setup at the moment:
Acoustic Elegance Dipole18 20-80Hz
BDDesign BD15 (Beyma OEM) 80-800Hz
Scanspeak 12MU 800-2.100Hz
Mundorf 25D1.1 2.100-20.000Hz
All crossovers, room correction etc. with Acourate FIR filters.
The current system is still in a kind of long-term beta test phase, so somehow never completely finished.
Why this choice of drivers? Because I already had them available Especially the 15 inch might not be the optimal solution, but basically it works great in the used frequency range, keeping a good directivity for the overall system.
The whole system works very well at the moment, however, for non-technical reasons (WAF among other things) I would like to downgrade a little bit. 3-Way would be fine, perhaps later I will add a subwoofer for the lowest octave.
So plan is to replace the BD15 and 12 MU, Change the xover frequencies and use a Purifi instead.
The Dipole18 should be able to play much higher than in my current setup, the Mundorf can be used a little bit lower, than the above mentioned cutoff.
I think the Purifi could play somewhere from 300 or 600 Hz up to 2000Hz
I would be interested in
Thanks and best regards
Matthias
(On the picture the speaker is not at the listening position, as we are rearranging the living room at the moment.)
since reading about Purifi drivers, I'm thinking about using them in an open baffle system.
Those drivers show fantastic test results from @HiFiCompass , erinsaudiocorner.com etc.
Most of the parameters should be very nice for using them in OB systems: high xmax, very low distortion, even during high excursion, very low intermodulation distortion ...
Only downside I would see is the low sensitivity of some drivers, especially the X-Versions with the highest xmax. Of cause this is logical, but depending on your open baffle design, the sensitivity on the lower end can be - due to the dipole cancellation - even much lower than the value in the specs.
Now I'm asking myself, which could be the optimal Purifi driver for such an system for the midrange.
Actually, there is of cause a dedicated midrange driver, which looks very good, however, from the specs, in terms of distortion etc. the NAA-Versions of the PTT6.5 are on the same level.
So which driver would be the best choice then?
This is my setup at the moment:
Acoustic Elegance Dipole18 20-80Hz
BDDesign BD15 (Beyma OEM) 80-800Hz
Scanspeak 12MU 800-2.100Hz
Mundorf 25D1.1 2.100-20.000Hz
All crossovers, room correction etc. with Acourate FIR filters.
The current system is still in a kind of long-term beta test phase, so somehow never completely finished.
Why this choice of drivers? Because I already had them available
Especially the 15 inch might not be the optimal solution, but basically it works great in the used frequency range, keeping a good directivity for the overall system.The whole system works very well at the moment, however, for non-technical reasons (WAF among other things
) I would like to downgrade a little bit. 3-Way would be fine, perhaps later I will add a subwoofer for the lowest octave.So plan is to replace the BD15 and 12 MU, Change the xover frequencies and use a Purifi instead.
The Dipole18 should be able to play much higher than in my current setup, the Mundorf can be used a little bit lower, than the above mentioned cutoff.
I think the Purifi could play somewhere from 300 or 600 Hz up to 2000Hz
I would be interested in
- your suggestions
- your (listening) experience (especially PTT6.5M and PTT6.5X-NAA)
- feedback in general
- any valuable comments
Thanks and best regards
Matthias
(On the picture the speaker is not at the listening position, as we are rearranging the living room at the moment.)
The 6.5M is ideal for a nude midrange like you show in the image above, from about 350Hz and up to about 2kHz. The 4 Ohm version is not too bad in terms of voltage sensitivity, at around 90dB @ 2.83V I think. The low end of the midrange passband is where the power demand is highest. If you are concerned about sensitivity then move the crossover point up a bit from 350Hz or use a small/narrow baffle around the driver no wider than about 8-9 inches.
I have a pair of these drivers and will be building a nude/baffled hybrid OB that is similar to what you show, the main difference being that my woofer section is at the floor in a 0.5m x 0.5m baffle and the mid and tweeter are separated from that and up at about 1m. I use closed box subwoofers below about 80Hz so that the woofer panel does not have to handle the lowest frequencies. This gives excellent bass and midbass but also the impact of a sealed subwoofer.
I have a pair of these drivers and will be building a nude/baffled hybrid OB that is similar to what you show, the main difference being that my woofer section is at the floor in a 0.5m x 0.5m baffle and the mid and tweeter are separated from that and up at about 1m. I use closed box subwoofers below about 80Hz so that the woofer panel does not have to handle the lowest frequencies. This gives excellent bass and midbass but also the impact of a sealed subwoofer.
The whole system works very well at the moment, however, for non-technical reasons (WAF among other things
So you have identified the system is fine and the wife is the issue.
You have two options.
1. Change the wife
2. Change the speakers
For any respectable DIY'er there is really no choice here. Welcome to freedom.
Bad man jokes aside your wife is a saint putting up with your baffles. What you might do is ask her what she wants and ask her to help you design new baffles so she has "buy-in". What does she like? Wood grain look or high gloss paint finish? Would she like a slimmer front dimension? Height? Square or rounded top?
If you do this with a little skill you can keep your current 4-way and the money you save buying new drivers spend on the baffle and a big present for your wife.
A beautiful baffle could hold all four drivers. Sealed to the floor it will shorter. It should be slimmed and tapered. The AE 18 should almost touch the sides and above the baffle tapered in to just hold the 15'. The top may be a tapered and rounded. Small side baffles will provide stability, a small U-frame for the 18's to do bass while providing a frame for grill cloth on the rear to tidy up the looks.
Last edited:
Hi.
FYI, I have an active 3 way project with purifi 6.5M (12inch sealed woofer TBD in its own box and 6.5M and bliesma t25b with 6 inch augerpro WG above on a approx 36cmx24cm rounded box) (Joachim Gherard Beo style)
Because I just have a daugther my project is going really slowly but I am planning to build a prototype of the top box in the next months and do a lot of measurements. Lets wait for the measurements but I expect the 6.5M range to be 230Hz to 2200Hz. Based on Xmax and power limitations, 6.5M max SPL (1m 2Pi) is around 112db at 200Hz and decrease linearly to 90db at 50Hz. Hence a XO around 230Hz seems feasible to me.
Thats my first big project with little experience so I may be in the wrong path.
Regards
FYI, I have an active 3 way project with purifi 6.5M (12inch sealed woofer TBD in its own box and 6.5M and bliesma t25b with 6 inch augerpro WG above on a approx 36cmx24cm rounded box) (Joachim Gherard Beo style)
Because I just have a daugther my project is going really slowly but I am planning to build a prototype of the top box in the next months and do a lot of measurements. Lets wait for the measurements but I expect the 6.5M range to be 230Hz to 2200Hz. Based on Xmax and power limitations, 6.5M max SPL (1m 2Pi) is around 112db at 200Hz and decrease linearly to 90db at 50Hz. Hence a XO around 230Hz seems feasible to me.
Thats my first big project with little experience so I may be in the wrong path
.Regards
Hi @CharlieLaub,
thank you very much for this feedback.
Obviously what you plan is very similar to my ideas
So the idea should not be too bad.
The good thing with software like Acourate is, that I can configure different crossovers in parallel and I can even switch between them during listening or when I change the type of music. Hence, I can take my time and try out.
Did you already try out the PTT6.5M in any configuration? Any impressions yet?
Hi @kazap,
And yes, I'm happy that my OBs are basically accepted. But it was dearly bought. Most design decisions in the house I had to leave to her
Yes, I will probably be able to design a baffle for the 4 existing drivers with a smaller optical footprint, but somehow I also like the idea to change the system and integrate a Purifi driver. It is not all about the WAF.
On the other hand, may be I will try to find a nice, slim modern design for the existing drivers in parallel and then decide. I'm not in a hurry.
A tapered baffle however could be a nice solution for the PTT6.5 as well. I did several simulations with The Edge and tapered baffles normally reduce the dipole dips and are smoothing the frequency response, if done right.
It was somehow obvious, that this comment will be made in this thread, so no worries
And yes, I'm happy that my OBs are basically accepted. But it was dearly bought. Most design decisions in the house I had to leave to her
Yes, I will probably be able to design a baffle for the 4 existing drivers with a smaller optical footprint, but somehow I also like the idea to change the system and integrate a Purifi driver. It is not all about the WAF.
On the other hand, may be I will try to find a nice, slim modern design for the existing drivers in parallel and then decide. I'm not in a hurry.
A tapered baffle however could be a nice solution for the PTT6.5 as well. I did several simulations with The Edge and tapered baffles normally reduce the dipole dips and are smoothing the frequency response, if done right.
Not yet, so I can't give you any sort of listening impressions. In any case that will be strongly influenced by how the driver is used, exactly.
I only did some measurements to check out the response pattern off axis and to the rear. It suits my needs well, and I trust the performance measurements I have seen on line made by other DIYers and by Purifi, which all indicate the driver is excellent.
The nice thing wth dipoles is that in dipole range, edge diffraction doesn't happen at all! As well baffle shape and edge rounding etc. don't mean a thing. Very smooth on- and off-axis response is lost above dipole peak, when also narrowing of radiation starts to happen.
Purifi 6.5 has small cone, radiating width is only 140mm, frame width is 165mm. In attachment is Edge sim, we see that dipole pattern is lost at 2kHz, usable range is roughly 2 octaves, depends on xo slopes. This why 4-way is needed, or we are in trouble with radiation pattern or distortion. Purifi's good motor makes it possible to extend low range, to make highpass xo of the midrange driver lower.
Purifi 6.5 has small cone, radiating width is only 140mm, frame width is 165mm. In attachment is Edge sim, we see that dipole pattern is lost at 2kHz, usable range is roughly 2 octaves, depends on xo slopes. This why 4-way is needed, or we are in trouble with radiation pattern or distortion. Purifi's good motor makes it possible to extend low range, to make highpass xo of the midrange driver lower.
Anything worth mentioning about the response patterns?
I recently measured the Scanspeak 12MU at 0° and 180° and both frequency responses looked quite similar up to 2700Hz, so a good dipole behaviour can be assumed here. However, I did not check the rear off axis patterns.
Attachments
Attached are the quickie FR measurements I made, with the driver nude. The middle plot compares the response on axis (0 deg) to the 180 deg response (e.g. "on axis" but as seen from the rear). I made these measurements outdoors, so they are are pretty high resolution.
EQing the peak around 1.1k Hz will be challenging. I have a large AMT that I can cross below 2k Hz, so I think the driver will just work out response wise for me. If you add a baffle of 1" to no more than 2" wider than the driver flange the dipole peak will likely be smoothed a bit, but the pattern may break up earlier. I will try it nude first, which is my preference and keeps the physical size presented to the listener around the M and T as small as possible.
EQing the peak around 1.1k Hz will be challenging. I have a large AMT that I can cross below 2k Hz, so I think the driver will just work out response wise for me. If you add a baffle of 1" to no more than 2" wider than the driver flange the dipole peak will likely be smoothed a bit, but the pattern may break up earlier. I will try it nude first, which is my preference and keeps the physical size presented to the listener around the M and T as small as possible.
Attachments
Last edited:
Since Purifi drivers are part of this topic, I thought I would share a plot that Lars Risbo posted on Facebook recently, that compares the max SPL for their drivers vs frequency, as limited by Xmax and input power. Note the position of the 6.5M above 200Hz? Per-fect.
Attachments
Hi Charlie,
What was your drive level and measuring distance for the above measurements?
I too, have a pair of 6.5M-04 on hand, and an MTM dipole panel already built so I should be able to take some measurements to verify, dependent on the weather.
The curves that Lars present are 2pi on an infinite baffle. In a narrow width panel eg. 8-11", or nude, do you think the 3mm of so x-max of the 6.5M is enough? This is probably a question that needs some empirical measurements and depends on how low/loud one wants to take it. I think with the W version a 3 way open baffle is possible.
Did I get his correctly Matthias that you bought that open baffle speaker?
Before you embark on this project, do you have some measurement equipment and familiarity in dual channel analog microphone measurements?
Acourate/Dirac etc are useful for taking in-room measurements of a completed loudspeaker. Y
ou will need a non-USB based microphone and get started with ARTA, REW, Omnimic, or Clio to take quasi-anechoic measurements to properly measure your drivers and incorporate your crossovers.
With your interest in trying different configuration of OB and needs to design for WAF imperatives, I wonder if you and your good wife would like a modular design (so you can easily swap driver modules) with aesthetically pleasing covers curved covers. Inspiration from the brilliant designers at the Pure Audio Project:
Hi @tktran303 ,
no, i built them completely by my own.
Actually, it is meanwhile the 5th iteration or so of my OB experiments.
With the drivers I have, a 4-way turned out to be the best solution so far.
I have a measuring microphone, Acourate, ARTA, 'ARTA measuring box' etc., so I think, I'm well equipped with all I need.
Uli from Acourate is very helpful, when there are any questions or problems with his software.
And he is full of ideas, how to optimize crossovers, room correction and everything you can do with Acourate
Exactly.
I'm not sure which frequency range at which SPL can be covered by the PTT6.5M, when using a small baffle or a nude driver.
'Normal SPLs' should be possible at 300Hz.
However, maybe it will be helpful to have some flexibility with the woofer in the upper range to adjust the crossover frequency to the personal needs.
Or take the PTT6.5X04-NAA instead and we are again at my initial question
What is normal is like "how long is a piece of string"
For instance, the 15" and 18" drivers is too big or too small, depending on what your needs are. How did you choose your crossover points of 80Hz for those two? How did you choose the 800Hz crossover point between the 15" and the 4"?
For instance, the 15" and 18" drivers is too big or too small, depending on what your needs are. How did you choose your crossover points of 80Hz for those two? How did you choose the 800Hz crossover point between the 15" and the 4"?
Last edited:
The 80Hz has been chosen, because - though they have been physically integrated into the speaker - I have used the Dipole18 as a double mono subwoofer, not as left and right woofer of the stereo part. As my room is very asymmetrical, it is difficult to get a similar low frequency response for a left and right stereo woofer. Using two subwoofers instead smoothens the response a bit and improves of cause the channel equality. Actually I have used a third Dipole18 as a third subwoofer for a while.
The 800Hz have been chosen, because I made a rough off axis measurement of the BD15, which showed, that it still has quite good dipole directivity there and does not start beaming in the current baffle. The BD15 is actually a very very good midwoofer, which besides from beaming issues due to its size shows a very nice frequency response much higher than the 800Hz.
800Hz as high pass for the 12MU seems to me a good choice as well. And has been chosen by a couple of other users as well.
My idea was to benefit from the midrange capabilities of the BD15 as high as possible.
Limiting factor is - as already mentioned- the beaming due its size.
If you want to go 3 way and keep the Dipole 18, one approach would be to-
Find the optimal working range of your Dipole 18-
Model the excursion limited SPL of the Dipole 18 in the layout in which you intend to use it. eg. Rectangular baffled, or nude.
Take baffled or nude impedance measurements and nearfield FR measurements of the Dipole 18- this will give a rough guide to how high it can go. The nearfield measurement (~1cm) from the cone, at the midpoint of the driver, should be accurate to about 250Hz in a 18" driver with. If there are no bumps in the zoomed impedance or the large peaks/dips in the nearfield measurement from 10Hz to 250Hz, then your crossover point should be OK out to 250Hz.
Take a farfield measurement at least 3 times the distance of your cone diameter, or twice the width of your baffle. 1m may not be enough. So you may need to get outdoors and off the ground. Use the longest gate you can practically organise eg. 5-10ms. This will give you insight into the driver's behavior from 250Hz upwards.
Once you find the upper limit of the Dipole 18, then it will be easier to decide whether your midrange needs 3mm (M), 6mm (W) or 9mm (X) x-max whilst functioning in dipole.
I haven't seen any measurements of the Dipole 18.
I haven't measured the NAA or M yet in dipole. Both drivers are on hand and baffle already built, and I intend to when the weather permits. And report back
Find the optimal working range of your Dipole 18-
Model the excursion limited SPL of the Dipole 18 in the layout in which you intend to use it. eg. Rectangular baffled, or nude.
Take baffled or nude impedance measurements and nearfield FR measurements of the Dipole 18- this will give a rough guide to how high it can go. The nearfield measurement (~1cm) from the cone, at the midpoint of the driver, should be accurate to about 250Hz in a 18" driver with. If there are no bumps in the zoomed impedance or the large peaks/dips in the nearfield measurement from 10Hz to 250Hz, then your crossover point should be OK out to 250Hz.
Take a farfield measurement at least 3 times the distance of your cone diameter, or twice the width of your baffle. 1m may not be enough. So you may need to get outdoors and off the ground. Use the longest gate you can practically organise eg. 5-10ms. This will give you insight into the driver's behavior from 250Hz upwards.
Once you find the upper limit of the Dipole 18, then it will be easier to decide whether your midrange needs 3mm (M), 6mm (W) or 9mm (X) x-max whilst functioning in dipole.
I haven't seen any measurements of the Dipole 18.
I haven't measured the NAA or M yet in dipole. Both drivers are on hand and baffle already built, and I intend to when the weather permits. And report back
I tried to get a 3-way nude OB to work with an 18" woofer. It would have to cover the range 80Hz to 350Hz. The low end is more of a function of baffle. More baffle makes for more low end capability. But I could not really find any 18" driver that would work well up to 350Hz. Even some 15" drivers did not sound good that high. I think this is due to cone resonances. If you look at the impedance plots for 18 and 15 drivers you often see a glitch around 250-300Hz. This is probably the first cone resonance mode. The only driver I was satisfied with was the JBL 2268H, and I had to cross it at 325 Hz. So I have mostly given up on that idea now (but see below for a solution).
The other thing that limits how high a large driver can be used is the off axis response, which starts pinching at some point. Juhazi posted a plot of this behavior above in post #8. This gets worse the larger the driver or baffle.
I came up with another solution: use multiple smaller drivers in a baffle that is about as big as, or slightly bigger than, the nude 18". One successful example of this uses four 8" low distortion drivers, and covers 80-350Hz nicely. Sounds effortless. The 80Hz cutoff means excursion requirements are modest and even 5mm Xmax is enough. The drivers do not need to be "exotic" or wildly expensive, you just need to pay attention to where distortion starts to climb, e.g. around 100Hz, and choose the best one. Online data such as from HifiCompass or Erins Audio Corner is invaluable for this purpose.
OTOH, choose a really high performance, high cost driver for the midrange! This is where I think the Purifi 6.5M really shines because of its low IMD and HD.
Diffraction effect of a naked 6.5" driver (Sd 133cm^2, OD 176mm) (ie. no baffle)
Simulated response of baffle-less PTT 6.5M-04. 2.83V drive level @ 1m.
Diffraction effect on 6.5" driver on 19cm x 1.2m baffle, driver mounted 40 from top, centered along width. Open/dipole baffle. Mic & tweeter 10cm from top edge (reference axis)
Simulated response PTT 6.5M-04 19cm x 120cm open/dipole baffle; 2.83V @ 1m. 40cm from top edge, centre along width of baffle. Mic 10cm from top edge, on axis with tweeter.
Simulated response of baffle-less PTT 6.5M-04. 2.83V drive level @ 1m.
Diffraction effect on 6.5" driver on 19cm x 1.2m baffle, driver mounted 40 from top, centered along width. Open/dipole baffle. Mic & tweeter 10cm from top edge (reference axis)
Simulated response PTT 6.5M-04 19cm x 120cm open/dipole baffle; 2.83V @ 1m. 40cm from top edge, centre along width of baffle. Mic 10cm from top edge, on axis with tweeter.
Last edited: