Subwoofer Tests Explained

**SteveCallas** · **08-01-08, 07:03 PM**

Quote:

darrell wrote:

The 2xPR has more output only from 30 Hz down.

That does not appear to be the case. Look at the pink 115db sweep for instance - the PR sub is 1-2db higher from 40-80hz and that increases to 6db higher in the 20-40hz range.

Quote:

noah wrote:

Disagree; the PR's have no motor nonlinearities, only suspension which is mostly the highly linear air spring.

Are the PRs using the same soft parts as the driver?

**noah katz** · **08-01-08, 07:13 PM**

I believe so

**darrellh44** · **08-01-08, 07:36 PM**

Quote:

SteveCallas wrote:

That does not appear to be the case. Look at the pink 115db sweep for instance - the PR sub is 1-2db higher from 40-80hz and that increases to 6db higher in the 20-40hz range..

Even if you use the 118dB %THD curve from the sealed 5400, it has considerably less distortion than the 115dB %THD curve from the 2xPR 5400

**SteveCallas** · **08-03-08, 12:37 PM**

The sealed 118db sweep most closely matches the PR'd 115db sweep in terms of output.

But there is something causing a difference in the distortion profile - I'd still have to say the PRs. If we look at the SVS Ultra 20hz, 15hz, or 10hz tunes vs the sealed version, we see less distortion at higher output levels everywhere above tuning with the ported versions. In my mind this is because the air in a port is more linear than a PR.

**darrellh44** · **08-04-08, 02:07 PM**

Steve,

I agree about the SPL and %THD plots for the sealed vs 10/15/20 Hz tuned SVS PB-13 results, but they are much closer to one another (for upper frequencies) than the sealed vs PR'd versions of the LMS-5400.

Ilka,

For the next round of testing, would it be possible to perform the spectral contamination test for the sealed and ported (PR'd) versions of the SVS PB-13 and LMS-5400 subs?

Thanks,
Darrell

**DrWho** · **08-19-08, 11:03 PM**

Quote:

SteveCallas wrote:

The sealed 118db sweep most closely matches the PR'd 115db sweep in terms of output.

But there is something causing a difference in the distortion profile - I'd still have to say the PRs. If we look at the SVS Ultra 20hz, 15hz, or 10hz tunes vs the sealed version, we see less distortion at higher output levels everywhere above tuning with the ported versions. In my mind this is because the air in a port is more linear than a PR.

There is no way air in a port is more linear - the non-linearity of air in a port begins because you can't have absolute pressures going below a perfect vacuum. It's totally a physical limitation and the only way around it is to move to a PR.

The distortion I think you're referring to is very likely because the PR's have too narrow of a Q. You don't get to adjust the Q of a port, but you can choose the Q of a PR. When you go too narrow, the cone excursion of your active will go up around the tuning frequency...thus the distortion is higher there too. Take some mass off your PR's and let the tuning frequency go higher and that behavior goes away...the rise in response at the tuning frequency at higher driver levels starts to go away too.

**andru** · **08-19-08, 11:19 PM**

Hi all,
Just wondering... if you are testing your sub in a free field environment... that is out doors.
Would you be better to raise them away from the ground by 20' or 30"..? On a pole or a extending speaker stand or something....
Is it possible?... otherwise you do have a boundary effect to consider, and not really a totally free field. From a scientific point of view.

Andru

**Ricci** · **08-20-08, 04:16 PM**

20 or 30" is not going to have much of an effect at the frequencies being tested here. Also some of the subs weigh upwards of 100lbs, so that becomes a factor.

**Ricci** · **08-20-08, 04:22 PM**

Quote:

DrWho wrote:

There is no way air in a port is more linear - the non-linearity of air in a port begins because you can't have absolute pressures going below a perfect vacuum. It's totally a physical limitation and the only way around it is to move to a PR.

The distortion I think you're referring to is very likely because the PR's have too narrow of a Q. You don't get to adjust the Q of a port, but you can choose the Q of a PR. When you go too narrow, the cone excursion of your active will go up around the tuning frequency...thus the distortion is higher there too. Take some mass off your PR's and let the tuning frequency go higher and that behavior goes away...the rise in response at the tuning frequency at higher driver levels starts to go away too.

What about all the suspension non linearities, and distortions caused by the surround and spider/s at large displacements like in these tests? I'm pretty sure that there will be more noise/ non linearity from a PR or 2 that is close to it's limits than from a port that's close to it's. Also the PR's will have a hard mechanical limit.

**DrWho** · **08-20-08, 08:43 PM**

The SPL of a port near its limits is nowhere near the SPL of a PR near its limits...

The suspension non-linearities of a passive aren't going to be any worse than your active. In fact, increasing the surface area of the passives (ie, doubling up) is going to reduce whatever distortion you're worried about. The thing about passives is that you're free to use as many as you want...

Ports will show measurable power compression at only 10m/s air flow - and yet most designs are barely able to get under 50m/s at full output. As you make the port larger to reduce air flow, the cabinet gets bigger, and your natural resonances (of both the port and the cabinet) start moving into the passband of the driver. And even if some of these resonances are just above the passband, the harmonic distortions from the driver can trigger them too. The point being that it's not a free lunch to add more port...

**Ricci** · **08-20-08, 11:09 PM**

Yes the linearity problems with your driver will also be present but these are added to by the PR or multiple PR's also most PR systems don't have enough displacement to handle the drivers full output at the tuning freq. So they are under PR'd just like most systems are underported. Where are you getting this port compression at 10ms statistic? I don't think this is accurate for a 6" or much larger heavily flared port. True the PR will trap the noises that sometimes escape from a port inside the box. There is no free lunch with PR's either. Don't get me wrong I like PR's but I"ve never seen anything to prove that a PR will always be cleaner than a port. I think it depends on the application.

**SteveCallas** · **08-21-08, 01:28 AM**

Quote:

DrWho wrote:

The SPL of a port near its limits is nowhere near the SPL of a PR near its limits...

That's a pretty generic, all encompassing statement. Are you suggesting a single 15" PR can provide more clean output than a heavily flared 10" port, as an example?

**Ilkka** · **08-21-08, 03:58 AM**

Quote:

darrellh44 wrote:

It guess its not so much a question of restricting the higher frequency excursion, but adding distortion to it. A case in point can be seen in some of Ilkka's Round 5 results if you compare the % THD plots between the sealed LMS-5400 and the 2xPR LMS-5400. For every SPL level 105db and higher, there is significantly more THD between 35 Hz and 100 Hz for the passive radiator version (even though it has twice the volume).

2xPR LMS-5400:
http://personal.inet.fi/private/zipm...200l%20thd.png

Sealed LMS-5400:
http://personal.inet.fi/private/zipm...100l%20thd.png

-Darrell

Hi guys,

I too noticed the difference in upper bass THD right after plotting the results. When looking at the individual HD components more closely, I noticed the difference was mainly caused by 2nd order harmonic. Now, there could be a few reasons for that.

1) The drivers weren't identical. A small difference in BL linearity could cause that kind of difference in 2nd harmonic.

2) The enclosures weren't identical. The upper bass performance of a small sealed enclosure is different from larger sealed or ported/PR enclosure.

3) I'm almost 100% sure it isn't because of passive radiators. The radiators don't contribute at those frequencies, so all THD is produced by the active driver. The passive radiators are too "stiff" to move at those frequencies. There's no non-linearity problems because the radiators aren't simply moving at all (or at most very little).

When wanting a proper comparison (same driver, same enclosure) between ported and sealed, look at the SVS PB13-Ultra (as you already did).

**Ilkka** · **08-21-08, 04:04 AM**

Quote:

andru wrote:

Hi all,
Just wondering... if you are testing your sub in a free field environment... that is out doors.
Would you be better to raise them away from the ground by 20' or 30"..? On a pole or a extending speaker stand or something....
Is it possible?... otherwise you do have a boundary effect to consider, and not really a totally free field. From a scientific point of view.

Andru

You are being correct, the testing environment isn't free field. But actually it doesn't have to be because the ground acts like a perfect mirror for frequencies that low (below ~200 Hz). Therefore we simply get 6 dB more SPL compared to free space (usually compensated by measuring the subwoofers at 2 meter distance). The measuring technique is called ground-plane. I suggest looking into the following paper.

Gander, Mark: "Ground-Plane Acoustic Measurement of Loudspeaker Systems," Journal of the Audio Engineering Society, vol. 30, no. 10, October 1982.

**Ricci** · **08-21-08, 12:52 PM**

There were 2 different LMS drivers? All this time I thought that you tested the same one and just switched alignments with the one driver. I never noticed the slight increase in distortion above 50hz for the PR'd system before.

**Mike P.** · **08-21-08, 12:54 PM**

I think he means not identical as in production tolerances.

**darrellh44** · **08-21-08, 03:37 PM**

Illka,

Which 5400 driver from the Round 5 tests did you use for the recent 75L/200L sealed special test, the one from the sealed or PR'd enclosure?

Thanks,
Darrell

**Ilkka** · **08-21-08, 03:48 PM**

Quote:

Ricci wrote:

There were 2 different LMS drivers? All this time I thought that you tested the same one and just switched alignments with the one driver. I never noticed the slight increase in distortion above 50hz for the PR'd system before.

Yes, I have two LMS-5400 18" drivers. But even though the THD graph seems quite different, the difference is mainly caused by the 2nd order harmonic which isn't really that harmful or audible. Maybe some day I'll measure the other driver in that 100 liter sealed box so that we know for sure whether it's a difference caused by the driver or the alignment/box. In any case, I wouldn't worry about it that much.

**Ilkka** · **08-21-08, 03:52 PM**

Quote:

darrellh44 wrote:

Illka,

Which 5400 driver from the Round 5 tests did you use for the recent 75L/200L sealed special test, the one from the sealed or PR'd enclosure?

Thanks,
Darrell

I used the one from the sealed enclosure.

**Ricci** · **08-21-08, 05:16 PM**

Quote:

Ilkka wrote:

Yes, I have two LMS-5400 18" drivers. But even though the THD graph seems quite different, the difference is mainly caused by the 2nd order harmonic which isn't really that harmful or audible. Maybe some day I'll measure the other driver in that 100 liter sealed box so that we know for sure whether it's a difference caused by the driver or the alignment/box. In any case, I wouldn't worry about it that much.

Oh i wouldn't worry about that tiny bit of distortion myself either. It barely breaks 5% at a very high output level until 23hz. Still awesome performance, just not quite as squeeky clean as the sealed test.

Advertisement