My CHR-70+DSP Linkwitz Transform experiment (4" driver playing flat to 20hz!)

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Thanks for that test. I was surprised how well I matched the standard curves, except for a problem at 500Hz. Thought it might be the CHR70 but checked the levels with a sound meter and it's my ear that is the problem, 6b down :(

But I can hear the 16kHz which doesn't surprise me as I can still hear CRT monitors and cat scarer's at the ripe old age of 42.
 
i would like to share this nice vst i found to analyze the spectrum of the ongoing track, it is somewhat desynchronized but very complete and free under win and osx:

Real-time audio spectrum analysis plugin - SPAN - Voxengo

as usual just add the vst to your /program files/vst plugins folder then 'rescan' from the vst wrapper and enable and it should work (should... for it gives me a bug under win7 for i have to set the wrapper to 'bypass' and it will then work just fine to observe)

you have many regulations one important is 'edit' 'slope'
 
setting the slope to 0db i can see there is some energy going on in the very sub-sonic frequencies, more than i thought... is this something to care about? is it contributing to the amplifier and speaker duties?

here is shoot from Sweet Georgia, Chesky records. how should i read the graph, is it considerable energy there from 1 to 20hz? is it going to the driver?
 

Attachments

  • Capture.jpg
    Capture.jpg
    90.5 KB · Views: 382
setting the slope to 0db i can see there is some energy going on in the very sub-sonic frequencies, more than i thought... is this something to care about? is it contributing to the amplifier and speaker duties?

here is shoot from Sweet Georgia, Chesky records. how should i read the graph, is it considerable energy there from 1 to 20hz? is it going to the driver?

I think there's more going down there in general than what people realize.

This may be due to the fact that so few systems actually play that low, and the differences are subtle.

An interesting thing to note in that graph is that, despite all the effort people go forth in order to get high frequency response that goes all the way up to 20khz, there's actually notably more output at 20hz than there is at 15khz, let alone 20.
 
notice that the slope is set to 0, so you can notice how mastering almost naturally respects the loudness curve. the spl required for something below 20hz to be heard/felt would be huge, so no doubt it's not in the hearing range, my concer mostly being around drivers duty, does it count?

tresh are you experimenting further with a 30hz goal?
 
Chapter 5: The sound

--------------
Whoah! That was a long intermission! I hope you guys went to the lobby and bought plenty of treats!

Sorry about the big delay. I got distracted earlier and then we all got into some discussions and then I got onto other things and started procrastinating finishing this up. Would it be possible for an admin to perhaps edit the "intermission" post linking to this post? It would allow people to read straight through if they wanted to.
--------------

ANYWAY, back to the point! I want to talk, now, about listening impressions. This will be, of course, mostly subjective for the moment. I have not taken the effort to do measurements, and I may never on this particular set, as I'm not confident in these particular drivers (one has been dropped, the other overextended) nor my hacked-together measuring equipment to give meaninful results.

That said, where do I start?

Obviously they sound great. I've been living with them for some time now, so the initial "wow" factor has subsided, but these speakers, and this setup, still wows me on a regular basis.

In the higher frequencies, well, they sound probably like most CHR-70s do. They're not perfect, but they're so good at certain things, and everything else they do pretty darn well. There are a hundred posts out there about the CHR's performance in the midrange and treble, however, and I'm not doing anything special in that regard, here, so I'm going to focus on the lower frequencies.

First off, I have done a few listens of test tones, and as was discussed earlier in the thread, at a normal desktop listening distance of about 3 feet, they do NOT produce much audible sound at 20hz. 30hz is definitely a more reasonable expectation. However, the 20hz is about where they cut off, and there definitely is some noise in the 20-30 range that I feel is very useful.

Second, I want to talk about distortion a little bit. At 30hz on a test tone there are noticeable amounts of harmonic distortion, which seems to be primarily of the 2nd order nature. I have not measured this, but it's quite noticeable. at mid to high volumes. At lower volumes it still exists, though less prevailant. This doesn't startle me much, as this is considerably lower than the speakers are generally designed to play, and I would be totally astounded if this weren't the case. Closer to 20hz it becomes even more obvious, as the distortion is significantly louder (not to mention easier to hear!) than the fundamental.

That said, it's not that bad and considering the size of the speakers I'd go so far as to say that it's pretty damned good, as I've heard worse harmonic distortion out of much bigger speakers (*cough* B&W 602 *cough*).

Also, I'm not sure you noticed, but 30hz is really ******* low!

And finally I want to touch on, ehm.. intermodulation distortion? Is that what one would call it? Essentially, low frequency distortions causing distortions in the higher frequency sounds. I'm not sure about the physics behind it, whether it has to do with the physical movement of the cone causing actual turbulence in the air in front of the cone that distorts higher frequencies, if it has to do more with the magnetic field saturating and so some of the HF energy is lost, or if the cone itself is flexing and absorbing energy which affects it (I'd like to read more about this!)

Nonetheless, in certain very specific situations, an extremely low, loud, sustained note will cause the higher frenquency sounds to kinda "pulse" in time with the low frequency note. The sound is similar to talking through a fan.

This doesn't happen often. The best example I can find of this is at the end of Uranus (haha) from The Planets suite by Holst. The song ends with an extremely ominous, extremely LOUD, extremely LOW tone, with a horn blare that comes in over top. This suite is known for using some unusual instruments with extremely low ranges, like the contra-bassoon and the bass-clarinet, the latter of which has a lowest note that's BELOW 20hz!

This is an extreme case, of course. It's not noticeable at low volumes, and definitely gets exponentially worse as volume is increased, which leads me to believe that it's not a fundamental problem of a full-range driver itself, but more of an interaction the high levels of harmonic distortion caused by the loud lower frequency note causing the motor structure to respond differently to the high frequency sounds.

My best guess is that the driver is extending beyond the motor's linear range, and thus, at the peaks of motion, the affect of the magnetic field is lessened, which doesn't just distort the LF sound, but also reduces amplitude on HF sounds while the cone is at that peak. Since this is happening at 20-25hz, it's an audible "pulsing" sound, as if you were playing the speaker through a fan.

I know this is all well known documented stuff, but it's fun to work it through in my head!

Final chapter, "The Music" will cover my experiences listening to various music with these speakers, how they sound in practice, and how it's opened my eyes a bit regarding sound in general
 
Last edited:
And finally I want to touch on, ehm.. intermodulation distortion? Is that what one would call it? Essentially, low frequency distortions causing distortions in the higher frequency sounds. I'm not sure about the physics behind it, whether it has to do with the physical movement of the cone causing actual turbulence in the air in front of the cone that distorts higher frequencies, if it has to do more with the magnetic field saturating and so some of the HF energy is lost, or if the cone itself is flexing and absorbing energy which affects it (I'd like to read more about this!)

Nonetheless, in certain very specific situations, an extremely low, loud, sustained note will cause the higher frenquency sounds to kinda "pulse" in time with the low frequency note. The sound is similar to talking through a fan.

Doppler Distortion in loudspeakers - Have a read of that, I believe that's what you're hearing.
 
Doppler Distortion in loudspeakers - Have a read of that, I believe that's what you're hearing.

I'm familiar with that, but I'm not sure that's the effect that I'm hearing here. What I'm hearing is not a subtle coloration of the sound, but a very prominent "pulsing" of the sound. In actuality, it's probably a combination of the effect of doppler distortion and non-linearities of the speaker at high excursion levels. I feel the majority of what I'm hearing is due to the latter.
 
--------------
Whoah! That was a long intermission! I hope you guys went to the lobby and bought plenty of treats!


First off, I have done a few listens of test tones, and as was discussed earlier in the thread, at a normal desktop listening distance of about 3 feet, they do NOT produce much audible sound at 20hz. 30hz is definitely a more reasonable expectation. However, the 20hz is about where they cut off, and there definitely is some noise in the 20-30 range that I feel is very useful.

Second, I want to talk about distortion a little bit. At 30hz on a test tone there are noticeable amounts of harmonic distortion, which seems to be primarily of the 2nd order nature. I have not measured this, but it's quite noticeable. at mid to high volumes. At lower volumes it still exists, though less prevailant. This doesn't startle me much, as this is considerably lower than the speakers are generally designed to play, and I would be totally astounded if this weren't the case. Closer to 20hz it becomes even more obvious, as the distortion is significantly louder (not to mention easier to hear!) than the fundamental.

Also, I'm not sure you noticed, but 30hz is really ******* low!

Nonetheless, in certain very specific situations, an extremely low, loud, sustained note will cause the higher frenquency sounds to kinda "pulse" in time with the low frequency note. The sound is similar to talking through a fan.

This doesn't happen often. The best example I can find of this is at the end of Uranus (haha) from The Planets suite by Holst. The song ends with an extremely ominous, extremely LOUD, extremely LOW tone, with a horn blare that comes in over top. This suite is known for using some unusual instruments with extremely low ranges, like the contra-bassoon and the bass-clarinet, the latter of which has a lowest note that's BELOW 20hz!

My best guess is that the driver is extending beyond the motor's linear range, and thus, at the peaks of motion, the affect of the magnetic field is lessened, which doesn't just distort the LF sound, but also reduces amplitude on HF sounds while the cone is at that peak. Since this is happening at 20-25hz, it's an audible "pulsing" sound, as if you were playing the speaker through a fan.

I know this is all well known documented stuff, but it's fun to work it through in my head!

Hi Tresch,
Very interesting experimentation, possibly a little scary from my side. All the same I'm glad fellas are experimenting and enjoying the process.

Can you give me some idea of the length of the cone's oscillations (in mm) that you generated on the LF signal @ 30Hz? Appreciate its difficult to measure but some observation of the cone's edge would give some indication. And let me know the voltage you applied if possible.

Re "Doppler distortion", its unlikely to be significant given the CHR's cone's size and dispersion pattern. You may be correct in thinking its more likely the driver's reaching its linear mechanical limit (hence my tech enquiry). The characteristic you describe would show up the most on artificial single length generated signals. You'll likely find more musical listening experiences quite interesting from the low range perspective. Much depends on how well the box and driver are optimised; and the level of room gain will also be a factor. The CHR's are capable of generating a promising level of LF given their small size considering they are full rangers rather than woofs. Please fellas, remember the CHR's are smaller sized drivers, so there's a risk of damage if they're driven hard on larger LF loads for long time periods.

Cheers

Mark.
 
Last edited:
Hi Tresch,
Very interesting experimentation, possibly a little scary from my side. All the same I'm glad fellas are experimenting and enjoying the process.

Can you give me some idea of the length of the cone's oscillations (in mm) that you generated on the LF signal @ 30Hz? Appreciate its difficult to measure but some observation of the cone's edge would give some indication. And let me know the voltage you applied if possible.

Re "Doppler distortion", its unlikely to be significant given the CHR's cone's size and dispersion pattern. You may be correct in thinking its more likely the driver's reaching its linear mechanical limit (hence my tech enquiry). The characteristic you describe would show up the most on artificial single length generated signals. You'll likely find more musical listening experiences quite interesting from the low range perspective. Much depends on how well the box and driver are optimised; and the level of room gain will also be a factor. The CHR's are capable of generating a promising level of LF given their small size considering they are full rangers rather than woofs. Please fellas, remember the CHR's are smaller sized drivers, so there's a risk of damage if they're driven hard on larger LF loads for long time periods.

Cheers

Mark.

Oh indeed! These particular drivers are ones that I ordered some time ago and have used as "shop mules," so they are now "test subjects." I'm fully aware of the risks. In fact, one of them I dropped in the shop and it's gained a bit of a buzz at a specific frequency in the midrange, so I'm pushing these things pretty hard to see where the limitations are.

As far as excursion goes, a lot. In regards to the issue pointed out above, with high freqency sounds pulsing when played along with intense low frequencies, I've noticed it most IN music, but in specific rare kinda cases. The most extreme being the final hit at the end of Host's "Uranus, the Magician" (from The Planets). An extremely low blasting bass note, with a horn blare overtop. The bass note is somewhere around 30hz, and at the volumes I'm noticing the distortion, the cone is moving an easy 5 or 6mm peak-to-peak for a sustained note. If I turn it down just a little, the artifact completely dissapears, which leads me to believe that it's an issue with reaching nonlinearities in the driver due to high-travel in either the motor or the suspension.
 
Chapter 6: Revelations

This is a tough one to write about. I've been procrastinating this chapter a bit because honestly, I don't really know exactly what to say, other than "they sound great."

Frankly, I'm not the most experienced listener. Over the years I've heard some great speakers, but I've certainly not done a huge amount of objective listening of many different speakers, DIY and otherwise. Furthermore, at least when it comes down to judging/describing things, I'm a bit of an objectivist, so I'm not a fan of the general audiophile "lingo" that people use to describe loudspeaker sound, as I find it all to be so subjective, and the words used to be so ambiguous, that very little meaningful communication happens in these conversations, with most people just taking from them whatever meaning they choose to hear.

That said, there is of course value in the attempt, and I'll do as best I can. Since this article is primary focused on the low frequency response of a driver that is otherwise fairly familiar to most interested parties here, it makes my job easier.

That said, well, these speakers sound great. I will not say they are the best speakers out there, but they are absolutely amazing in this configuration within this set of constraints. So good, in fact, that it has significantly changed the way I look at loudspeaker design.

What these have taught me is that low-frequency response, specifically high qualitylow frequency response is extremely underrated in most situations. There's a good reason for this, which is that it's hard to reproduce well, so it's expensive, so two things happen. 1) people undervalue it because it makes them more comfortable with the fact that they can't have it, and 2) Many who DO value it execute it poorly so it sounds terrible, which then feeds back into #1, giving them more reason to undervalue it.

The fact is that audio is a visceral experience, which explains the huge amount of subjectivity in this craft. Furthermore, the lower frequency sounds become increasingly more important, as they stimulate the brain differently than other sounds. They transcend the "hearing" into the "feeling," and thus, even though the extreme low end of the audio spectrum gets much less activity than the rest, the mind's reaction to it is proportionately greater.

If you've ever been in the same room with a kick drum, you not only know how it should sound, but how it should feel, and as long as that isn't represented correctly, the recording will always just seem like a recording.

The astounding thing is that this revelation has come to me by way of a couple 3" drivers.

I've heard a lot of ported speakers and subwoofers and various combinations of the above, but they always seem to be a bit off. I wasn't aware of this until this experiment, of course, they sounded fine. They sounded like speakers. Now I know what people mean by "delay" and "decay." I can hear the big harmonics in cheap car subwoofers, and the peaks in response from improperly sized/stuffed boxes. Finally, I'm always left hanging by that fancy B&W 602, which plays nice and flat down to about 40hz, like a good ported box should, and then just runs out of breath.

The kick drum in "Sailing the Seas of Cheese" by Primus comes alive. It sounds like I'm listening to an actual DRUM instead of just some piece of production. The Contrabassoon and bass clarinet in "The Planets" play solid, clean, and low, and not just in the background of the song, but right along with the other instruments, filling the room with ominous vibrations.

On the kick drum note, the delay of a port, even a properly sized one, as well as it's tendency to store energy (even thwacked the end of a PVC pipe and listened for how long it takes for the sound to stop?) are much more obvious to me, now. I always thought kick drums were one of those things that were just lost in modern production techniques, reduced to techno-esque "bumping." Now I believe that is largely due to the port separating the sub from the midbass and killing cohesion of a single "hit" on a drum.

Occasionally I fall asleep to some ambient electronic music, and those electronic artists get tricky. I may be laying in bed and thinking that people are walking around two floors up... sounds that I can barely even detect AS SOUND, and instead I "feel" them, and yet, I'm not talking big boomy bass, but subtle near-subsonic... rumble is not the right word, but perhaps "pressures." Can you imagine a "whisper" at 24hz?

People say there is very little in music below 40hz. By the numbers I could agree, but I could also say that your eye picks up very little energy in the "red" wavelength compared to the rest, but that doesn't make it any less important. I am convinced that there IS plenty of information below 40, and it is not the huge, booming, hollywood bass that we associate with that. It is the subtle sounds of a world constantly colliding with itself. The thumps and rumbles in the background of every day life. It is merely that the combination of subtlety and difficult of reproduction make it easier to ignore than respect.

Now, a big aspect of this whole test is that I'm listening to these speakers as computer speakers, and they are about arms-length from my ears. In some situations I'm laying in bed while they play, so they're more like 8-10 feet away, but complimented by a very quiet room.

I believe this further enhances my point on the importance of these lower frequencies, however. I find myself much happier with lower volumes with proper equalization, even without any added "loudness" EQ happening at lower volumes in order to compensate for the whole non-linearity of human hearing.

To sum up, I love these speakers, and I love them even more EQed! They've changed how I think about speaker and furthermore, system design. I already have parts here for a chip-amp kit, and as soon as that is finished I'll be doing some testing with dedicated linkwitz transform circuits.

My next project will be two Alpair10s in sealed enclosures with a matching LM3886 amp with a custom designed linkwitz-transform circuit. I hope to implement a bypass potentiometer so you can "turn down" the bass boost if you want to play higher volumes to a bigger room, and flatness is not as important. I'm extremely excited about this project, as the A10s have both nearly double the surface area AND double the excursion capability of the CHR-70s. They should be able to produce TRULY flat, clean, 20hz response at computer monitor volumes and distance.

I'll keep you up to date on the that project as it comes together!

Thanks for listening!
 
Hi Tresch,

"An extremely low blasting bass note, with a horn blare overtop. The bass note is somewhere around 30hz, and at the volumes I'm noticing the distortion, the cone is moving an easy 5 or 6mm peak-to-peak for a sustained note. If I turn it down just a little, the artifact completely dissapears, which leads me to believe that it's an issue with reaching nonlinearities in the driver due to high-travel in either the motor or the suspension".

Yes, you've reached the limits of the driver's applied excursion. As per your experiments, it's normally easy to recognise the limit and adjust the load on the driver.

"My next project will be two Alpair10s in sealed enclosures with a matching LM3886 amp with a custom designed linkwitz-transform circuit. I hope to implement a bypass potentiometer so you can "turn down" the bass boost if you want to play higher volumes to a bigger room, and flatness is not as important. I'm extremely excited about this project, as the A10s have both nearly double the surface area AND double the excursion capability of the CHR-70s. They should be able to produce TRULY flat, clean, 20hz response at computer monitor volumes and distance".

This is interesting feedback. I'm in the pre-production stages of the Gen 2 Alpair 10. The plan for June is to test another batch of coil designs, 5 new types in all. Also 2 designs of rear suspension. The Multiform cones have been made and tested.

I'd be most grateful for feedback from forum members on the new Alp10's useful excursion, power handling and efficiency. Hoping members will post on this thread:

http://www.diyaudio.com/forums/markaudio/158417-alpair-10-generation-2-sneak-peek-5.html

Cheers

Mark.
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.