the 8R data sheet should become online very soon in a 0.99 version (just missing the power test)
Hi Lars,
Have you had any more thoughts on giving any info in regard to the acoustic centre for this driver?
My build is progressing, drivers have arrived, tweeter pods are made but can't be shaped to fit close to the spheres until I have an offset figure for the drivers (easy for the tweeter it's a ribbon)
I will soon be in a position to make some measurements for this offset, but to have an approximate physical measurement would give me much more confidence in my acoustic measurements!
Steve
Attachments
here's everyone's favorite older, but smaller brother PTT6.5X-04-NAA in a thoughtfully constructed cabinet-
Mr. NFS shows that the 10K resonance seen on the datasheet has been suppressed:
with just one notch filter filter tuned for that 5K resonance:
Mr. NFS shows that the 10K resonance seen on the datasheet has been suppressed:
with just one notch filter filter tuned for that 5K resonance:
Last edited:
the 1k glitch from Erins NFS is not there.interesting. well done!
hi Steve, what an amazing construction! I have some ideas for measuring the AC but it will take a while to implement. One challenge is that the AC of the mic is also an unknown so it needs to be relative to a refence driver, eg some tweeter/ribbon with pistonic motion far up.
Hello All,
I am past the temptation phase. I am expecting some of the 8 inch 8R woofers to arrive in a couple of weeks or in a month or so.
Some thoughts:
I put the 8 inch 8R Purifi woofer TS/P's into the simulator and see that sealed enclosure models 4 times the volume of the 6-1/2 inch mid-woofer. If you are familiar air-suspension or sealed enclosures they are sensitive to air compression distortion, especially small sealed enclosures. The 8 inch woofer uses a much larger sealed box than the 6-1/2 inch mid-woofer. I will likely test drive the 8 inch woofer is a sealed box to see if there is nasty compression distortion or not. In the end I would like to keep the speakers bookshelf size. I expect to tuck a sub-woofer under the bench.
If you look at the 8R datasheet it shows that the 8R woofer begins to beam a bit before 2khZ. I want to make an effort to control directivity, you know no large discontinuity between woofer and tweeter, a smooth gradual slope with increasing frequency is okay.
A tweeter with a waveguide is expected. I do not want a horn that is dissimilar in size compared to the Purifi woofer.
I have a couple of Compression drivers with waveguides sitting here that I want to test drive:
FatialPro HF 108R compression driver with STH100 Tractix waveguide.
JBL D2430H compression driver with 90X50 constant directivity waveguide.
I have measured these tweeters with a GRAS microphone and APx555 analyzer, if there is an interest I will post some FR and distortion plots.
Thoughts?
Thanks DT
I am past the temptation phase. I am expecting some of the 8 inch 8R woofers to arrive in a couple of weeks or in a month or so.
Some thoughts:
I put the 8 inch 8R Purifi woofer TS/P's into the simulator and see that sealed enclosure models 4 times the volume of the 6-1/2 inch mid-woofer. If you are familiar air-suspension or sealed enclosures they are sensitive to air compression distortion, especially small sealed enclosures. The 8 inch woofer uses a much larger sealed box than the 6-1/2 inch mid-woofer. I will likely test drive the 8 inch woofer is a sealed box to see if there is nasty compression distortion or not. In the end I would like to keep the speakers bookshelf size. I expect to tuck a sub-woofer under the bench.
If you look at the 8R datasheet it shows that the 8R woofer begins to beam a bit before 2khZ. I want to make an effort to control directivity, you know no large discontinuity between woofer and tweeter, a smooth gradual slope with increasing frequency is okay.
A tweeter with a waveguide is expected. I do not want a horn that is dissimilar in size compared to the Purifi woofer.
I have a couple of Compression drivers with waveguides sitting here that I want to test drive:
FatialPro HF 108R compression driver with STH100 Tractix waveguide.
JBL D2430H compression driver with 90X50 constant directivity waveguide.
I have measured these tweeters with a GRAS microphone and APx555 analyzer, if there is an interest I will post some FR and distortion plots.
Thoughts?
Thanks DT
DT, I'm using some Seas 8" aluminium cones, and find they're a good match with 1" compression drivers. My crossover is around 1kHz. Thread here: https://www.diyaudio.com/community/threads/18sound-nsd1095n-as-good-as-it-gets.381895/
IMO, 8"+1"CD is a good combo that works well for a wide range of music.
Enjoy!
Chris
IMO, 8"+1"CD is a good combo that works well for a wide range of music.
Enjoy!
Chris
hi DT,
You bring the linearity of the air suspension up. The stiffness from the box air experienced by the cone is proportional to Sd^2/Vbox. At large excursion, the volume is modulated which modulated the stiffness. However, this adds in parallel with the Kms of the driver it self. Kms nonlinearity drops very fast with frequency/excursion and does not produce much IMD. see examples https://purifi-audio.com/2019/12/12/imd/
A bigger concern is if Sd is modulated (which affects the air stiffness strong let due to the power of two in the equation above). Sd modulation causes amplitude modulation (IMD) of all frequencies reproduced and is what we worry most about.
The high pressure in a small box will exacerbate the Sd modulation of conventional half roll surrounds unlike the Purifi surround that keeps an almost constant Sd even at large excursion and high back pressure.
I hope this helps. Happy building!
cheers
Lars
You bring the linearity of the air suspension up. The stiffness from the box air experienced by the cone is proportional to Sd^2/Vbox. At large excursion, the volume is modulated which modulated the stiffness. However, this adds in parallel with the Kms of the driver it self. Kms nonlinearity drops very fast with frequency/excursion and does not produce much IMD. see examples https://purifi-audio.com/2019/12/12/imd/
A bigger concern is if Sd is modulated (which affects the air stiffness strong let due to the power of two in the equation above). Sd modulation causes amplitude modulation (IMD) of all frequencies reproduced and is what we worry most about.
The high pressure in a small box will exacerbate the Sd modulation of conventional half roll surrounds unlike the Purifi surround that keeps an almost constant Sd even at large excursion and high back pressure.
I hope this helps. Happy building!
cheers
Lars
Does this technically mean that drivers in open baffle do not experience this form of distortion? And to a certain extend, subwoofers in "infinite baffles" as well?
yes, correct (the air stiffness modulation goes away in infinite volume but the Sd modulation persists)
I think nearly the same could be acomplished by using 2 identical drivers with soft rubber surround , one positive and the other one with negative half roll .
So there will be very low effective Sd modulation!
Both surrounds shold be finetuned (slightly different diameters) for maximum effect .
There could be higher modes in the surround I noticed five sections in the ugly HYDRA surround 🙂 ... the idea seem to be taken from the "pentacone" introduced by VIFA , long time ago .
But other manufacturers using complex shaped surrounds too (better visuals)
so I see chances to get rid of the modes too by using some nicier 3D sculptings!
And - the old FOSTEX drivers , with double suspension above and below the air gap , HAD NO SURROUND AT ALL (just a foam ring around the cone edge)!!! (quite big of an air gap dont know why maybe to avoid shuffing)
Recently I saw IMD measurings of the KEF LS50 with TONS OF IMD !!!
(ASR/audiosciencereview.com)
Around 20dB below signal only (translates to 10% distortion) .. must be a result the nearly flat Z - Flex rubber surround not withstanding the positive/negative air pressure inside the box .
I must say I was pretty much shocked that such a high regarded speaker is such a disappointment!
Other more expensive models (Reference Line) with the same coax driver (maybe 5inch instead of 6-7inch) have very small surrounds of different shape and needed a woofer or subwoofer , along with a crossover around 350Hz .
Now back to the drawing board scratching the head and waiting for some enlightments 🙂
bye!
So there will be very low effective Sd modulation!
Both surrounds shold be finetuned (slightly different diameters) for maximum effect .
There could be higher modes in the surround I noticed five sections in the ugly HYDRA surround 🙂 ... the idea seem to be taken from the "pentacone" introduced by VIFA , long time ago .
But other manufacturers using complex shaped surrounds too (better visuals)
so I see chances to get rid of the modes too by using some nicier 3D sculptings!
And - the old FOSTEX drivers , with double suspension above and below the air gap , HAD NO SURROUND AT ALL (just a foam ring around the cone edge)!!! (quite big of an air gap dont know why maybe to avoid shuffing)
Recently I saw IMD measurings of the KEF LS50 with TONS OF IMD !!!
(ASR/audiosciencereview.com)
Around 20dB below signal only (translates to 10% distortion) .. must be a result the nearly flat Z - Flex rubber surround not withstanding the positive/negative air pressure inside the box .
I must say I was pretty much shocked that such a high regarded speaker is such a disappointment!
Other more expensive models (Reference Line) with the same coax driver (maybe 5inch instead of 6-7inch) have very small surrounds of different shape and needed a woofer or subwoofer , along with a crossover around 350Hz .
Now back to the drawing board scratching the head and waiting for some enlightments 🙂
bye!
Hello Lars and All,
Just for clarity I am going into this with the understanding that the Purifi 8 inch 8 Ohm driver is a Ultra Low Distortion driver.
Any enclosures, amplifiers, crossovers or the like used/assembled will be with the target of optimum performance.
Any discussion about the nonlinearity of sealed box air compression has nothing to do with the driver.
Thanks DT
Just for clarity I am going into this with the understanding that the Purifi 8 inch 8 Ohm driver is a Ultra Low Distortion driver.
Any enclosures, amplifiers, crossovers or the like used/assembled will be with the target of optimum performance.
Any discussion about the nonlinearity of sealed box air compression has nothing to do with the driver.
Thanks DT
a push pull configuration cancels all even order distortion including IMD. This will kill a good part of the Sd modulation. The Purifi surround is kind of combining an equal amount of positive and negative half roll in order to cancel the Sd modulation.
Hello All,
To start I will fabricate a sealed enclosure near 1 cubic foot to see how things measure with a microphone.
There is also a 6-1/2 inch Purifi driver installed in a 0.55 cubic foot Denovo sealed enclosure sitting here on my bench. I do not believe that the sealed 0.55 cubic foot Denovo enclosure got a fair audition previously.
I will sort out the harmonic and the multi tone procedures before I post test plots.
Thanks DT
To start I will fabricate a sealed enclosure near 1 cubic foot to see how things measure with a microphone.
There is also a 6-1/2 inch Purifi driver installed in a 0.55 cubic foot Denovo sealed enclosure sitting here on my bench. I do not believe that the sealed 0.55 cubic foot Denovo enclosure got a fair audition previously.
I will sort out the harmonic and the multi tone procedures before I post test plots.
Thanks DT
hello Boden,
We thought we were done some years back when we finally had our new motor ready and assembled a driver with a normal half roll. To our big disappointment the THD and IMD performance was just on or with other good drivers. By borrowing a velocity laser we could measure that the cone motion was linear to -60dB but SPL only linear to around -40dB (at 150Hz if so recall). That was when we had to find the last distortion mechanism namely the surround Sd modulation. This was pointed out in a 1994 paper but no one took notice.
After some month of development we had a surround prototype ready that had taken the Sd modulation way down. Applied in the same cone and using the same motor we now measured the expected -60dB linearity. A moment to celebrate indeed.
cheers
Lars
We thought we were done some years back when we finally had our new motor ready and assembled a driver with a normal half roll. To our big disappointment the THD and IMD performance was just on or with other good drivers. By borrowing a velocity laser we could measure that the cone motion was linear to -60dB but SPL only linear to around -40dB (at 150Hz if so recall). That was when we had to find the last distortion mechanism namely the surround Sd modulation. This was pointed out in a 1994 paper but no one took notice.
After some month of development we had a surround prototype ready that had taken the Sd modulation way down. Applied in the same cone and using the same motor we now measured the expected -60dB linearity. A moment to celebrate indeed.
cheers
Lars
Hello All,
Today I have been test driving the Purifi 6-1/2in 8R driver today in an effort to dial in the APx 500 test procedures. The plan is to do similar testing with the 8in 8R drivers when they arrive here in the Pacific North West.
I set the Purifi 6-1/2in 8R driver up in a sealed 0.55 cubic foot Denovo enclosure and input 2.83 volts in an effort to compare my test results with the data sheet. See the attached test plot below.
My results are mostly in line with the data sheet. For my possible application as a mini 2-way bookshelf on my bench this sealed enclosure speaker looks like it could work well. There will be a sub tucked in on the shelf below the bench.
After previously seeing the red flag caution about about using this driver in a sealed enclosure I am glad to see tame levels of distortion in the below 100hZ range. The little assist in stiffening the LF suspension limits the LF travel of the cone. (speculation)
EDIT:
I should also mention that the input voltage was 2.83 volts as in in the datasheet. the test microphone was placed 6 inches away from the driver to limit the room effects. The dBSPL on the plot is elevated because the microphone is placed near the driver.
Thanks DT
Today I have been test driving the Purifi 6-1/2in 8R driver today in an effort to dial in the APx 500 test procedures. The plan is to do similar testing with the 8in 8R drivers when they arrive here in the Pacific North West.
I set the Purifi 6-1/2in 8R driver up in a sealed 0.55 cubic foot Denovo enclosure and input 2.83 volts in an effort to compare my test results with the data sheet. See the attached test plot below.
My results are mostly in line with the data sheet. For my possible application as a mini 2-way bookshelf on my bench this sealed enclosure speaker looks like it could work well. There will be a sub tucked in on the shelf below the bench.
After previously seeing the red flag caution about about using this driver in a sealed enclosure I am glad to see tame levels of distortion in the below 100hZ range. The little assist in stiffening the LF suspension limits the LF travel of the cone. (speculation)
EDIT:
I should also mention that the input voltage was 2.83 volts as in in the datasheet. the test microphone was placed 6 inches away from the driver to limit the room effects. The dBSPL on the plot is elevated because the microphone is placed near the driver.
Thanks DT
Last edited:
The mono-causal specoolation hyperbolea?
Well of course the engineers need something to chew on for their brain teeth ... but I remember there was a statement in the Klippel papers that the input signal determines to a large degree what you can measure in the output!
So the REAL MUSIC SIGNAL has an asymmetric waveform (typical) that is totally different from the IMD measurement signal , that raises the question how relevant is it really to hunt for better Klippel numbers?
And even beyond that - our ears are still the limit , having their own distortion generation! (at least related to harmonics)
Another factor to consider (there are more) , the deformation of the surround - is it depending of the input level?
And we have the subjective influence of the ELC - Equal Loudness Contours . A set of curves in 10dB steps showing how loud our ears perceive frequencies below 1000Hz . Around 40Hz the subjective loudness of a pure sine signal in anechoic space is typically 30dB(!) more quiet than a 1000Hz tone! (playback level 90dB)
So any (small amount) low-freq distortion will fall behind the detectable horizon!
I remember Bill Waslo (LAUD/PRAXIS etc) has that "DiffMaker" tool , can it be used extract the IMD-distortion component(s)? If so , the spectrum may show where the most energy of the IMD is concentrated . Other rubber types probably better? (room for improvement)
Just a few thoughts from my non-academic hobby brain , sorry!
bye ..
Well of course the engineers need something to chew on for their brain teeth ... but I remember there was a statement in the Klippel papers that the input signal determines to a large degree what you can measure in the output!
So the REAL MUSIC SIGNAL has an asymmetric waveform (typical) that is totally different from the IMD measurement signal , that raises the question how relevant is it really to hunt for better Klippel numbers?
And even beyond that - our ears are still the limit , having their own distortion generation! (at least related to harmonics)
Another factor to consider (there are more) , the deformation of the surround - is it depending of the input level?
And we have the subjective influence of the ELC - Equal Loudness Contours . A set of curves in 10dB steps showing how loud our ears perceive frequencies below 1000Hz . Around 40Hz the subjective loudness of a pure sine signal in anechoic space is typically 30dB(!) more quiet than a 1000Hz tone! (playback level 90dB)
So any (small amount) low-freq distortion will fall behind the detectable horizon!
I remember Bill Waslo (LAUD/PRAXIS etc) has that "DiffMaker" tool , can it be used extract the IMD-distortion component(s)? If so , the spectrum may show where the most energy of the IMD is concentrated . Other rubber types probably better? (room for improvement)
Just a few thoughts from my non-academic hobby brain , sorry!
bye ..
Boden and Marvel,
If you haven't already, pour yourself a coffee and like a good book, set aside 2 hours to watch:
Erins Hardison of www.erinsaudiocorner.com interviewing Lars Risbo:
Recall that over a quarter of century ago young Lars Risbo discovered/invented direct switching "digital" amplifiers following his PhD and founded Toccata Technology. Chip giant Texas Instruments (US) who first licensed their tech (some of us middle aged fogies might remember the TacT Millenium, which used with Toccata's "EQUIBIT" technology, and premiered at the final stages of last century (1998) for ~US$10K ) and then later bought the company, and Peter Lyngdorf changed TacT to Lyngdorf Audio and rest; as they say, is HiStory. Now almost all of TI's line of amplifiers is built upon this technology, and lives in everything from earphones to smartphones to laptops to bluetooth speakers to soundbars to AVRs to cars.
Virtually everything that isn't 20th century class A/B or triodes, or Hypex. 😉
https://www.ti.com/audio-ic/amplifiers/speaker-amplifiers/overview.html
I suspect life was too short to do something boring like sailing around the world for decades in a yacht after the TI buyout, so instead he Lars founded Purifi Transducer Technology with his ex-Toccata partner Claus Neesgaard and brothers from other mothers Bruno Lars (responsible for UcD over at Philips in the late 90s) and Carsten Tinggard (ex-Peerless) and Peter Lyngdorf (DALI, Steinway Lyngdorf, ex-NAD, Snell). Between them probably have over a hundred man-years of experience in the speaker/amplifier industry.
So despite knowing a lot, they continued to learn and continue to grow. And look at original data/research.
Who knew that flappy bit of foam or rubber connecting the cone to the frame was responsible for at least 20dB of distortion??
Perhaps that's why 20 years ago people like John Krutke were ranting about "woofers with a lower Xmax sound better" weren't completely off base...
If you haven't already, pour yourself a coffee and like a good book, set aside 2 hours to watch:
Erins Hardison of www.erinsaudiocorner.com interviewing Lars Risbo:
Recall that over a quarter of century ago young Lars Risbo discovered/invented direct switching "digital" amplifiers following his PhD and founded Toccata Technology. Chip giant Texas Instruments (US) who first licensed their tech (some of us middle aged fogies might remember the TacT Millenium, which used with Toccata's "EQUIBIT" technology, and premiered at the final stages of last century (1998) for ~US$10K ) and then later bought the company, and Peter Lyngdorf changed TacT to Lyngdorf Audio and rest; as they say, is HiStory. Now almost all of TI's line of amplifiers is built upon this technology, and lives in everything from earphones to smartphones to laptops to bluetooth speakers to soundbars to AVRs to cars.
Virtually everything that isn't 20th century class A/B or triodes, or Hypex. 😉
https://www.ti.com/audio-ic/amplifiers/speaker-amplifiers/overview.html
I suspect life was too short to do something boring like sailing around the world for decades in a yacht after the TI buyout, so instead he Lars founded Purifi Transducer Technology with his ex-Toccata partner Claus Neesgaard and brothers from other mothers Bruno Lars (responsible for UcD over at Philips in the late 90s) and Carsten Tinggard (ex-Peerless) and Peter Lyngdorf (DALI, Steinway Lyngdorf, ex-NAD, Snell). Between them probably have over a hundred man-years of experience in the speaker/amplifier industry.
So despite knowing a lot, they continued to learn and continue to grow. And look at original data/research.
Who knew that flappy bit of foam or rubber connecting the cone to the frame was responsible for at least 20dB of distortion??
Perhaps that's why 20 years ago people like John Krutke were ranting about "woofers with a lower Xmax sound better" weren't completely off base...
Last edited:
Bruno Putzeys, I mean
Apologies to Bruno. I really should stop dictating to my computer and go back to old fashioned typing... plus this 30 mins to edit limit is too short...