Sub testing results confusion...

[Vader]

Hi guys,

As I learn more about acoustics, my complete ignorance is beginning to wane somewhat (now I understand the axiom "Ignorance is Bliss"). There are two main areas of confusion (currently), and I am really hoping that some of the "subhumans" here can point out where I am all wet:

First, is the THD of my twin 20-39 PC+ cylinders. I have been unable to find THD data on the 12.3 drivers for these, so I am improvising using either the THD obtained by AVTalk for a single 20-39 PC+ using the 12.1 drivers (apples and oranges?), or the PB12 Plus/2 reviews posted by Ilkka here on the shack. The AVTalk review is only of a single cylinder, so I am assuming that, since co-located duals give a 6dB headroom boost, a good approximation would be to use the curve 5dB below that of a single (i.e. the THD given by my duals at 95dB should be close to that from the single at 100dB). But, then I am still comparing the 12.1 to my 12.3, so I am probably way off here. OTOH, I am thinking that the PB12 Plus/2 is a better approx, since it is two 12.3 drivers (and the PB12 Plus is basically the box brother of one of my cylinders). That seems to match of resonably well, except below 25 Hz, where the cylinder's distortion seems to be considerably higher. To my reasoning, dual cylinders should outgun a single PB12 Plus/2 (same tune) slightly in both SPL and THD; due to increased port space and internal volume, even close to the tuning frequency (no EQ required due to greater volume?). Here's my confusion: When comparing the 12.3 to the 12.1 chart, it looks like the 12.3 actually has higher THD (most notably the 110 and 115 dB curves) than the 12.1. Everything I have read says that the 12.3 drivers are nipping at the heels of the (current) Ultra, but the tests of the Ultra cylinder at AVTalk show a significant discrepancy relative to the 12.3 drivers (from Ilkka's reviews). I thought I read somewhere that the 20-39 had < 10% THD down to the tuning frequency, but I couldn't swear to it. I am also interpreting from AVTalk that re-tuning to a lower TF (probaby 16Hz) might help. Where am I missing the boat here?

Second, the review of the single 20-39 cylinder at AVTalk shows that the FR starts a rapid drop-off at around 25Hz - about the same place where the THD shoots up (upwards of 28-30% at 100dB, if I'm reading this right...?); I would have thought it would not drop off that fast until the tuning frequency. The curves at SVSound show just this (even though they appear to be the 20-39 CS, using the 12.1 drivers). And Tom stated (I don't recall the forum or the post) that "even if you have a strong 20Hz response, if you drop off strongly a full octave above, that will seriously compromise your 20Hz capability" (I'm paraphrasing here). Yet the PB12 Plus/2 drops off at the tuning frequency of 20Hz. Is the 20-39 really closer to a 25 Hz tune than a 20Hz (native) one?

Prior to getting the SVS cylinders, I *was* blissfully ignorant about this stuff (thanks a lot, guys….. ), but since I am somewhat compulsive/obsessive to begin with (so I am told…), now am trying to learn all I can about it (and improve my soundfield where I can). Don't get me wrong: I have absolutely no complaints about the Plus cylinders (except having to put padding under my jaw when I use them – hard floor, you know ). But I don't want to turn left on the learning curve where I should have gone straight, if you get my meaning.

[Sonnie]

If this doesn't get some attention here, we might move it to the SVSound forum.

[Vader]

Sonnie,

Yea, I probably should've put it there to begin with. If you could move it there, I would be much appreciative...

[Sonnie]

Done!

[SteveCallas]

I'd use Ed's PBPlus review at Secrets to gauge your 20-39 PCPlus - they should be very similar performers.

As for tuning, you can't really determine it based on a sub's FR, as the FR will vary based on enclosure size and any built in EQ/filters/limiters. The THD measurements should be a pretty good reflection though, as the varibales I mentioned won't really affect where the dip in THD takes place, which will be at or very near the tuning point. Based on that, it does appear the native tuning is closer to 25hz than 20hz.

[Vader]

Steve,

If I understand you correctly, the enclosure size/shape should have little to do with the THD, and the distortion coming from dual cylinders should be essentailly the same as from the Plus/2. If that is true, using Ilkka's review of the Plus/2 (12.3 drivers, 20 Hz tune) does indeed show a TP of 20 Hz. The review of the 20-39 PC+ on AVTalk used the 12.1 drivers. Could that be the cause of the discrepancy?

[SteveCallas]

If I understand you correctly, the enclosure size/shape should have little to do with the THD

No. In a ported sub, there will almost always be a temporary decrease in THD near the tuning point, as the driver shifts from having to use a lot of excursion above tuning, to not much at all at tuning, and then it rises exponentially at some point below tuning. Other variables can and will change the specific levels of THD, including enclosure volume with a static tune and amount of power, but excursion plays a large role, and THD measurements of a ported sub typically resemble excursion use closely. The more linear the motor technology being used, the less this holds true, but it holds pretty true for the Plus drivers.

and the distortion coming from dual cylinders should be essentailly the same as from the Plus/2.

No. What I'm saying is that Ed's review of the PBPlus (NOT Plus/2) should closely resemble the performance of one 20-39 PCPlus. Here is his review.

Could that be the cause of the discrepancy?

No, tune is determined by enclosure size and porting - the drivers used have nothing to do with it. Looking at Ilkka's measurements of the 25-31 PCPlus and the PB10, using THD as a guide, it would seem they are both tuned a bit higher than what one would expect.

The only way to know for sure is to do a close mic FR measurements of the driver. Do you have an spl meter?

[Vader]

See.... told ya I was way off kilter here....

I do have an older analog RS meter, which I have been using with REW (though, not near field). The response I have (incl. room response) is not too bad, but I need to get a BFD. There is a significant peak right before 20Hz, followed by the falloff I would expect. I am somwhat suspicious of my results, though (I think the problem lies with the idiot doing the measurements), as my response seems to come back below 16 Hz and is strong down to 10Hz ?!?!?

http://www.hometheatershack.com/foru...-response.html

[SteveCallas]

Ok great, you know how to use REW. Take your spl meter and lay it on the baseplate with the mic towards the center of the driver if possible - if it's too thick to fit between the bottom cap and baseplate, then lay it on the ground right next to the baseplate. If you don't have carpet floors, use something like a mousepad to keep the meter from rattling.

Now take a FR sweep.....it doesn't even have to be a full sweep, 10-50hz will be plenty fine. This should reveal the exact tuning point, as the output from the driver will overpower that from the ports at such close proximity. You'll see pretty flat response, then a dip, then one last hoorah. That valley of that dip is the tuning of the sub. Here is an example of a close mic FR of my sub:



Tuning ended up at ~13.5-13.6hz. Make sure you keep all three ports open. You can then try it with one closed and two closed and see where the real tuning ends up in those configurations.

[Vader]

Kewl! I will try this over the weekend and post the results back here. Thanx, Steve!

[SteveCallas]

The weekend? But it's only Monday.....and now you've got me interested in finding out what the results are going to be

[Vader]

OK, last night's results:

First, I re-did the room response from the sweet spot. Without EQ, it doesn't look too bad (and the room gain was increased slightly as all doors leading out were closed...)

raw response.jpg

Then, I measured nearfield (this is where I think I messed something up). As I have dual cylinders (placed about 1" apart), I placed the SPL meter between them, on the axis of symmetry (I couldn't have gotton the mic any closer w/out cramming it between the base pate and the woofer):

near field.jpg

It occured to me (after shutting everything down) that the subs were still corner loaded when I did the near field, so room gain did still come into play. Once I get a BFD, I think it will improve a great deal...

[SteveCallas]

No, you can't place the mic between them. Run one sub at a time with the mic on the baseplate. The BFD isn't needed for this test.

[Vader]

OK, I performed measurements of both cylinders seperately (with the meter between the base plate and the tube), and here is what I came up with (both subs were, for all intents, the same):

LeftCylinderNF.jpg

Looks as if the TF is around 22.8 Hz, correct?

[SteveCallas]

Yep, thanks.

[Manic Miner]

If I'm not mistaken I seem to remember that Ilkka measured the tuning frequency of the PB12-Ultra to be 20.5hz, and if I try to go even further back in my memory I can remember reading that the box versions are tuned deeper than the cylinders

[SteveCallas]

That would seem to jive wouldn't it? Vader is measuring a 20-39 cylinder.

[lienly]

I ain't a techy guy, so can anyone explain how to analyze such graphs? tks