Scan-Speek 4" for 49 cents?

[mzisserson]

Re: Scan-Speek 4" for 49 cents?

My science based designs should easily best your ear designs from what I have read. Even people who are good at working up a xo by ear tend to use a little science to help. Jbruner is associating what he hears to his measurements for example.

I agree with this.

But I don't think that darren should be written off, either. I have studied the science for 20 years now, honed in the correlations between what you see and what you hear. Worked extensively at (As Toole states) finsing what is meaningful measurably and what is measurably meaningful. The final step in my process is voicing, no matter how perfect I get them on paper, and no matter how well my paper results match my measured performance. This does not always yeild a technically perfect picture. The largest things, however is that it is always *close* to technically perfect. It has to be in several measurable domains or you will end up with audible issues regardless of the overall response.

[killa]

Re: Scan-Speek 4" for 49 cents?

I often wonder why the human ear, isnt the sound test tool of choice. It is in my world

My science based designs should easily best your ear designs from what I have read. Even people who are good at working up a xo by ear tend to use a little science to help. Jbruner is associating what he hears to his measurements for example.

[jbruner]

Re: NS6-ed

So now take an unmodified driver, build a contour filter to make the response match that of your modified driver and compare. That would be the "is all this mucking about worth it or should I just design a proper crossover" test. At the very least apply a 2nd order filter at 2.5kHz on both modified and unmodified driver to see how the distortion compares in a real world application.

Please Understand, Simply Satisfying You Won't Help Inform People Perceiving Every Distortion

[DarrenM]

Re: Scan-Speek 4" for 49 cents?

I often wonder why the human ear, isnt the sound test tool of choice. It is in my world

[Tin_Ears]

Re: Scan-Speek 4" for 49 cents?

I hope this isn't derailing...

What if the NSBs are crossed at 3K/24dB? Will the "unmodified" cone breakup still be an "audible" issue? What about 3K/12dB?

[weinstro]

Re: Scan-Speek 4" for 49 cents?

...Here is the response of sample 1 before and after Revelator style cut and glue job.


Blue is modded, brown is baseline. I added a small piece of duct seal to the slit dust cap to smooth the top octave on this one.

The net result is very impressive. The primary break-up is reduced by 11dB, distortions of all orders are lowered, and subjective sound quality is much improved.

In other words, there is no downside, unless you like the sound of paper cone break-up.

Wow, very cool. Looks like the slits interfere with the development of resonance on the surface of the paper cone. I wonder what slitting would do with a (cheap) metal cone driver with an inconveniently placed breakup node/frequency?

What voltage did you charge the (dust) cap to?

I probably need to stop at this point and clarify my intentions....
I wanted to find out if the well-known slit-cone method had any drawbacks. Could you damp the out-of band resonances without hurting the Clarity, Liveliness, Immediacy, Transparency?

Someone help me. I can't think of an acronym for this term.

Regards,

Rob

[dcibel]

Re: NS6-ed

I have completed testing on 2 sliced Aurasound NS6-255-8a samples and decided to bury the data here. These are unworthy of their own thread. Whoever it was that convinced me to buy these must be deaf.

3 NS6s baselined showing 3rd order distortion-



The slit technique definitely works for suppressing the break-up.

The slits destroyed the peak completely, and reduced the related 3rd and 5th order distortions by the same 8-10 db.

Sample-5




What it did not do is make the driver into something I would want to use. I can't stand to listen to them raw or modified.

There is one big problem that I may or may not address in this round of drivers. Distortion below 1k rises when testing the NS6 at high volumes. This indicates that the glue is too soft(which I suspect), the gap is too wide, or the floppy-coned NS6 is just not suited to this mod.

MY advice so far- sell the NS6s, keep the NSBs, and cut everything but don't glue yet!





Apparently not.

So now take an unmodified driver, build a contour filter to make the response match that of your modified driver and compare. That would be the "is all this mucking about worth it or should I just design a proper crossover" test. At the very least apply a 2nd order filter at 2.5kHz on both modified and unmodified driver to see how the distortion compares in a real world application.

[Sidi]

Re: NS6-ed

It's a driver evaluation. My opinion is that the driver sucks. I showed measurements that partially reveal why. It's obvious that someone would try to deal with it's flaws when designing a system. But the driver offers no low bass, poorly defined mid-bass and garbled mids. It's a lousy woofer that has to be crossed at 1500hz to avoid that 3rd order distortion. Doesn't sound very useable to me. Very Astute, Great Intelligence Never Assumes.

I didnt assume. I asked.

[Tin_Ears]

Re: Disregard

Thank you. I was trying to present the science in a format that everyone could understand.

I think he was referring to my post #96.

[jbruner]

Re: Disregard

This thread has become 8th grade level. Nice job, guys.

Thank you. I was trying to present the science in a format that everyone could understand.

[jbruner]

Re: NS6-ed

I guess I don't understand... ...They sound quite pleasant in a system.

It's a driver evaluation. My opinion is that the driver sucks. I showed measurements that partially reveal why. It's obvious that someone would try to deal with it's flaws when designing a system. But the driver offers no low bass, poorly defined mid-bass and garbled mids. It's a lousy woofer that has to be crossed at 1500hz to avoid that 3rd order distortion. Doesn't sound very useable to me.

Are you modifying them to use as a full range driver only? They aren't a full range.

Very Astute, Great Intelligence Never Assumes.

[johnnyrichards]

Re: Disregard

This thread has become 8th grade level. Nice job, guys.

[Tin_Ears]

Re: Disregard

* deleted * sorry :(

[scottq]

Re: Disregard

This thread is getting really exciting. I hope you can stick it out until all those slits are fully explored.

Wow. Well played.

[Sidi]

Re: NS6-ed

I guess I don't understand. Are you modifying them to use as a full range driver only? They aren't a full range. They sound quite pleasant in a system.