So far I've been working round the principle of minimising (more spreading over the curve length) the rate of change of curvature for the profile, using a full wrap around 180° into the cabinet, via a spiral profile. I've done this for the horizontal and vertical profiles with an eliptical mouth. I might try to create a back to back of the spiral so that I can smooth the abrupt curvature transition in the diagonals as it meets the flat baffle, rather than an edge with continued curvature.
Very true, but I am currently trying to do shape optimization in COMSOL based on this model, but with different driver materials/geometry and I have removed the dome shape optimization. While it does seem to work, it seem to be very picky when it comes to both the optimization criteria and the initial shape. I feel like the optimization procedure ends up in a local minima and thus doesn't converge further.
This would suggest that you would also get shallow WG's for 2.5D simulations as there are no baffle edge effects at all!
Yes, I already own the PTT8 and also the BlieSMa T25B-6 Tweeter. Still not sure about the midrange driver. Originally I planed to use the BlieSMa M74B-6 but since it’s End Of Life - and only available for prices around 2000€ - I need to find something comparable or go with the Silk version.
I need a 3 way setup for my studio.
Would run it with an active Hypex Amp (FA-253) @tktran303
I need a 3 way setup for my studio.
Would run it with an active Hypex Amp (FA-253) @tktran303
I have friends who are studio owners and most seem to listen in the near or midfield ie. monitors 0.5m - 2m
I wonder whether a 2 way with waveguided tweeter is preferable, for directivity matching and beamwidth- limiting early reflections and getting more direct sound vs side wall.
This seems to consistent with anecdotal evidence by Sean Olive re. Wide dispersion va narrow dispersion.
But if you already have the T25B I can see the temptation to use a midrange as well.
Consider using @augerpro. ‘S waveguide-
https://www.somasonus.net/bliesma-t25b
10 left in stock (?), though I guess it's close to 2000€ for a pair (is that what you meant?)
M74B-6 - Solen
Personally, I'd also consider the PTT5.25X08-NFA-01 because of:
M74B-6 - Solen
Personally, I'd also consider the PTT5.25X08-NFA-01 because of:
Last edited:
@Paul Ebert PTT5.25X08-NFA-01 has a SPL of 84.7dB. A bit low to match with the PTT8. The 6.5M maybe better.
@Paul Ebert
Agreed. For passive designs sensitivity matters a lot. But for active/DSP driven, much less so. Comparisons should be SPL matched.
Luckily Purifi provides comparisons for all their drivers measured at the same SPL (94dB@1m).
They ought to be commended on this.
In terms of distortion, the PTT8.0 remains un-matched for typical mid-bass frequency range uses eg. 50Hz to 2KHz:
But in terms of wide dispersion, the PTT4.0 is unmatched above 1KHz.
Exhibit A: PTT8.0X04-NAB
Exhibit B: PTT4.0M04-NFC
click on thumbnail then Left and Right cursor keys to compare and contrast harmonic distortion and on and off axis frequency response.
Exhibit A harmonic distortion:
Exhibit B harmonic distortion:
Exhibit A: frequency response:
Exhibit B: frequency response:
Exhibit A: PTT8.0X04-NAB
Exhibit B: PTT4.0M04-NFC
click on thumbnail then Left and Right cursor keys to compare and contrast harmonic distortion and on and off axis frequency response.
PTT5.25 and 6.5 are somewhere in between...
Agreed. For passive designs sensitivity matters a lot. But for active/DSP driven, much less so. Comparisons should be SPL matched.
Luckily Purifi provides comparisons for all their drivers measured at the same SPL (94dB@1m).
They ought to be commended on this.
In terms of distortion, the PTT8.0 remains un-matched for typical mid-bass frequency range uses eg. 50Hz to 2KHz:
But in terms of wide dispersion, the PTT4.0 is unmatched above 1KHz.
Exhibit A: PTT8.0X04-NAB
Exhibit B: PTT4.0M04-NFC
click on thumbnail then Left and Right cursor keys to compare and contrast harmonic distortion and on and off axis frequency response.
Exhibit A harmonic distortion:
Exhibit B harmonic distortion:
Exhibit A: frequency response:
Exhibit B: frequency response:
Exhibit A: PTT8.0X04-NAB
Exhibit B: PTT4.0M04-NFC
click on thumbnail then Left and Right cursor keys to compare and contrast harmonic distortion and on and off axis frequency response.
PTT5.25 and 6.5 are somewhere in between...
Last edited:
@hifijim Apart from the price, why would you go for the M74-A rather than the M74-S?
@tktran303 Thanks for the tip regarding the wave guide.
I'll see if I can find someone in Germany who can finish the whole thing for me with a 3D printer.
If you personally were faced with the choice of building a 3 way system, would you rather go with a Purifi for the mids or via BlieSMa M74-A/BlieSMa M74-S ?
If you are interested in designing an appropriate kit, maybe we could work together on it ?
But I'll say it straight out, I'm new to building speakers and actually have virtually no idea and rely mainly on the know-how of others/experience of others.
@Paul Ebert Shipping + Import Tax etc and yes roughly 2.000 € for a pair.
Crazy if you know that you could buy them for 1300€ a pair just a couple weeks ago.
@tktran303 Thanks for the tip regarding the wave guide.
I'll see if I can find someone in Germany who can finish the whole thing for me with a 3D printer.
If you personally were faced with the choice of building a 3 way system, would you rather go with a Purifi for the mids or via BlieSMa M74-A/BlieSMa M74-S ?
If you are interested in designing an appropriate kit, maybe we could work together on it ?
But I'll say it straight out, I'm new to building speakers and actually have virtually no idea and rely mainly on the know-how of others/experience of others.
@Paul Ebert Shipping + Import Tax etc and yes roughly 2.000 € for a pair.
Crazy if you know that you could buy them for 1300€ a pair just a couple weeks ago.
My thoughts are based on the data presented on hificompass.com. All versions of this driver have low enough distortion that I would not use distortion performance as a decision criteria. All are very high sensitivity and very high SPL capable. The paper and silk versions have resonances at around 6k. The typical upper crossover for this driver will be in the 3k -4k range, and a 6k resonance will be difficult to deal with. The aluminum version has its resonance at about 11k, and this will be significantly easier to work with.
j.
So, I'm not the only one who has observed a significant price increase for these. Unfortunate, that.
Part of the reason I like the PTT5.25X is the absence (reduction?) of a break-up related distortion peak. This peak appears to be much more pronounced in the OEM PTT5.25M:
I wonder why it's so less prominent in the PTT5.25X than, really, all the others?
No problems. Don’t thank me, shout our own @augerpro a coffee or beer:
https://www.gofundme.com/f/waveguides-for-cnc-and-3d-printing
He has painstakingly designed all those wave-guides from scratch using a highly iterative process. No AI was harmed (or used) to do this!
If I was personally faced with the choice of building a 3 way system, would I rather go with a driver A, B or C... well, this is a interesting question but a complex answer.
My design process usually starts with goals like -
"how big (or small) does the system need to be",
"what frequency range does it need to cover",
and "where is the listener(s) and how far away are they listening".
This generally determines my maximum cabinet volume, required SPL and directivity requirements. From there I do some modelling with LF elements. Suppose I am allowed 1 cu ft for woofers vs 4 cu ft vs 14cu ft. They have different solutions. For a given volume of box (Vd), I then look at desired F3/F6/F10. What is the intended use case? Is this for listening of the computer workstation for BGM, Youtube or general entertainment, or for a small bedroom sized 10ft x 10ft studio, a typical living room, a ballroom, nightclub or outdoors festival. After modelling the woofers I look at the HF element. With a waveguide or horn many compression drivers can make it down to 500Hz - 1KHz. If they can't, then I need a filler driver. This is the filler driver, or midrange, which is somewhat of a legacy driver. They are very rare these days.
I generally don't start with what drivers do I want to use, unless there is some contraints eg. cabinets already built etc,
There have been some exceptions to the rule, like the Purifi driver, because as a prototype, I wanted to hear what happens when distortion is pushed really low.
In simulations, the peaks behave the same way as the larger 8.0X, and the recommended passive circuit seems to work the same with a little adjustment. Funny that their peaks are so similar on such different sizes. Long ago when I designed a simple passive x-over for a 5.25X04-NFA based nearfield monitor, it practically snapped into place like their whitepaper.
All peaks seem to be around the 5khz and 7khz region respectively, when they appear. Variations of the 4.0M NLC have their peak either slightly above or below 5khz, the aluminum pushes their peak above 7khz. The 10.0X is right on 5khz, and the 8.0X shows both 5 and 7khz. Some of the 6.5M's and 6.5W's seem to be trying to push that breakup higher and higher. Even my scaly lizard brain is telling me it has to do with the way they design something. If the style of circuit application in their whitepaper could be adapted to other drivers, why aren't they selling the circuit as an add-on?