Anybody measure distortion of diy subs?

[warnerwh]

I'm wondering how the common diy subs compare to commercial subs. Distortion I know with subs can and usually is double digits. Is this why almost nobody states their distortion figures in commercial subs and drivers for diyers?

With most electronics in the .01% arena the distortion is really inconsequential. With speakers however being several percent or more I'd think it would audible.

I realize that speaker measurements and how they sound are not much related. :W

[ThomasW]

We don't bother since there's nothing to be done to lower it other than throwaway the driver and buy a different one.

The audibility of distortion in low bass is the subject of much discussion. The big guns like Tom Nousaine typically measure output levels at 10%THD

If anyone uses TrueRTA, I have a copy of Excel worksheet Ed Mullin and Ilkka developed to calculate distortion from TrueRTA plots. Be forwarned it's pretty user hostile.

[JonMarsh]

The most important factor is basic motor linearity and inductivity moduation- many sub drivers I've measured are not that good in this regard. There's a lot of measured data here and there for the RS315 and RS265; notable is that they have lower distortion than many other similar drivers. The TC2+ series is pretty good for a moderate throw driver. Most of the longer throw drivers are best kept below 70 Hz because of the high VC inductance (Le).

[warnerwh]

That's good to know Jon regarding the 70hz cut off. It seemed that mechanically a long throw driver couldn't stay as accurate as a short throw driver even with a bigger motor and stiffer surround and spider. I don't understand most of a drivers design but the mechanical part I do to some extant.

The distortion of bass being controversial is a good point. Our ears are less sensitive under 80hz than in the midrange. I was just curious because I'm trying to learn. This is far more interesting that what SS amp is best to use with such and such speakers.

[Evil Twin]

Less sensitive may not be strictly correct- if you've been exposed to really clean bass in the lower reaches, the bass reproduction which pleases many ears may not be appreciated any longer... the better servo systems, such as a Pardigm Servo15 are the best examples in single cabinets, and multiple driver manifold array IB subs also show what can be done with more brute force methods.

[warnerwh]

I've heard alot of live music so have a good reference. To my ears the sound that is closest to live is the most accurate.

In the midrange a voice being distorted at 10% though would probably be more obvious than a stand up bass playing a 60hz note that is 10% distortion. Let's use harmonic distortion in both cases.

The above is just a guess as I've never measured it and can only imagine. It's funny that the most non linearities are caused by the room and speakers and therefore affect the sound we hear the most. Yet it's rarely discussed.

If I had a way to cause distortion in my system I'd like to do this test with varying amounts of distortion. From what I've read over the years even 1% distortion is very difficult to hear in music. I'd bet that most speakers are doing very well however to keep the distortion below 10% at any substantial volume of say peaks of 95db.

[Jonasz]

Jon: How high in frequency would you consider the TC2000 15" SVC to be useful to?

http://www.tcsounds.com/tc2000.htm

[JonMarsh]

That's a good question, and how high are you interested? Ls/Lp isn't particularly good, and that's in relation to a 3 ohm VC, but that's to be expected with a 28 mm Xmax. I would assume the cone is as stiff as most TC-sounds designs, so that won't be the limiting factor; I'd expect inductance rise and response roll off coupled with THD increaase to be the main factor. I'm sure you could use them up to 100 Hz, implying decent performance to 200 Hz (transistion band) possibly with a little EQ, like my HE-15.


But Le limits rise time and distortion performance. In this regard consider the Dayton RS390HF, with an Le of 1.0 mH, with Faraday rings to stabilize inductance over cone travel. Not as much total Xmax, of course, but pretty clean to above 300 Hz. It's in a "transistion area" between subwoofer and woofer, I'd say. A nice place to be. I'd expect it's usable upper range (you set the critiera) to be 2.5-3X greater than the TC2000. For the same money, I'd buy two RS390 over 2 TC2000. More Sd seems to win out in sonics- maybe better radiation resistance?

[Jonasz]

Jon: I was looking for clean performance up to 80-100 Hz in an H- or U-baffle. I was thinking the TC-2000 with it's 28mm Xmax would keep the size down when it would be sufficient with only one. Today I use a W-baffle with two XLS12 (crossed at 100 Hz) and one TC-2000 should move twice the amount of air with just a slight increase of bafflesize.

Is there any reason not to use the TC-2000 in this configuration? If not, would the SVC or DVC (to keep the impedance up) be the best alternative?

[Tommythecat]

More Sd seems to win out in sonics- maybe better radiation resistance?

Is there any information about this? I heard something about the Cult'oIB and this, but I've never heard of such a property before.

[Evil Twin]

Is there any information about this? I heard something about the Cult'oIB and this, but I've never heard of such a property before.

Air is compressible, and the more compression locally, the more non-linear. Radiation resistance is a function of Sd; horns are not necessarily the ultimate solution becuase of the distortion induced in the air at the mouth of a horn. Radiation resistance affects conversion efficiency, one reason you see 15's with similar motor's to 10's with a higher net sensitivity.

[JonMarsh]

Jon: I was looking for clean performance up to 80-100 Hz in an H- or U-baffle. I was thinking the TC-2000 with it's 28mm Xmax would keep the size down when it would be sufficient with only one. Today I use a W-baffle with two XLS12 (crossed at 100 Hz) and one TC-2000 should move twice the amount of air with just a slight increase of bafflesize.

Is there any reason not to use the TC-2000 in this configuration? If not, would the SVC or DVC (to keep the impedance up) be the best alternative?

For what you're looking for, in an H-Baffle, the TC2000 should work well- 100-110 Hz being a reasonable upper limit for that baffle type, IMO.

The DVC version seems to offer a better ratio of Res to Le, and paralleling two in an H baffle will result in a nice load for a moderate size current oriented solid state amplifier- probably what you're looking for? With the SVC version, you'd want to series them.

Another possibilty using them in a diople is to measure parameters of one VC only connected for the DVC model; this will cut inductance, and raise the Q, and like the TC2+, this may be exactly the mode you're lookingn for in a Dipole, as you don't need the full VC power handling to hit over excusion limits, and the driver response may be more optimum.


TC2+ with both voice coils



TC2+ with one voice coil used


As expected, the sensitivity increases due to lower impedance. The LF corner behavior is also quite a bit different (higher Q, I'd say), and note how much less the response above 100 Hz rolls off- probably a function of VC inductance.

[Dennis H]

Lest warnerwh get the wrong idea, plenty of distortion testing goes into choosing a sub driver. As Thomas said, once it's built it's too late to do much about it. Here's one link:

http://www.mfk-projects.com/woofers.htm

Too many sources for tests around the net and I'm too lazy to track them all down but they are out there. Drivers with good measured performance and current DIY favorites include the Dayton RS series, TC Sounds (and all their renamed resellers), Peerless XLS and Aurasound. With all the drivers, it's best to keep excursion below 1/2 the rated Xmax. Distortion goes up pretty fast after that.

[---k---]

I guess I would be more interested in distortion created by things other than the driver.

[Tommythecat]

Air is compressible, and the more compression locally, the more non-linear. Radiation resistance is a function of Sd; horns are not necessarily the ultimate solution becuase of the distortion induced in the air at the mouth of a horn. Radiation resistance affects conversion efficiency, one reason you see 15's with similar motor's to 10's with a higher net sensitivity.

Very interesting, I haven't looked closely enough at the full properties of radiation impedance. While having a small knowledge of math/physics/acoustics, im attempting to understand more so i can write programs to model drivers, boxes and horns - from power dissapated to excursion to radiation patterns. Hopefully I'll have a grasp on it by the end of this next summer.

[warnerwh]

Dennis: Thank you for the link. I'm just curious. I can tell that the sub that I just built has very little distortion. My knowledge regarding the entire subject of subwoofers is very limited. The more I learn the more I find I don't know much. Sub's are much more complex than I'd guessed. This subject is much more interesting than the other audio stuff I'd been into for quite a few years.

[AJINFLA]

Jon: I was looking for clean performance up to 80-100 Hz in an H- or U-baffle. I was thinking the TC-2000 with it's 28mm Xmax would keep the size down when it would be sufficient with only one.

Jonasz, I don't know if Eclipse car audio drivers are more available in Sweden than the TC brand ones, but here is a TC based unit that would top my list for making a compact dipole baffle single 12 http://www.diymobileaudio.com/forum/...54&postcount=1
Hope you got one of my PM's from here or diyaudio .

cheers,

AJ

(originally attached in wrong thread :gah: )

[Dennis H]

AJ, thanks for reminding me about the diymobileaudio site. Look for threads started by npdang. He tests a bunch of drivers. Also the first sticky lists a bunch of other test sites that I was too lazy to look up. Looks like he caught all the main ones except of course for the tests our own JonMarsh runs on this forum.

http://www.diymobileaudio.com/forum/forumdisplay.php?f=7

[cdwitmer]

Related to this, of course, is the non-linearity of our ears. I'm thinking out loud here and would be grateful for any light that people might be able to shed on the subject . . .

I should think that at high volumes, something akin to compression is occurring in our ears. So it may be that at about the time the subwoofer is distorting, our ears are also going into distortion. Of course, distortion in our ears is something that our minds ought to be able to compensate for.

Of course, "distortion" encompasses a huge variety of phenomena, some more obnoxious than others . . .

[Jonasz]

Jon: So probably the best bet for me is to buy the DVC and use only one of the VC's? I think that makes sense... Thanks for your advice! :P ;x(

[Jonasz]

AJ: You have PM

Btw, I'm actually moving to the US in january so I will be able to buy them at your dirt cheap prices... :T That's the reason I'm building something new too, it's cheaper to build something fun than shipping my current system over the pond! :E