THD Analyzer Test of the Subwoofers

[basspig 62]

Very interesting test.. 20 Hertz signal, driven to Bassmaxx ZR18s, SPL of 110dB as measured on SPL meter. Behringer ECM8000 omni capsule mic measuring the sound, fed to Behringer MX1804X mixer, output of which drives Tektronix AA501 Distortion Analyzer and HP 3580A Audio Spectrum Analyzer. 

Not taking into account hum and noise, the distortion reading (unweighted) is 0.4%. On the spectrum analyzer, with the test signal stopped, there is 60Hz hum at -60dB, accounting for at least 0.1% of the total distortion reading. Of course there are many other factors, including room boundaries vibrating and re-radiating their own distortion, so this test is not completely accurate and probably contains much noise which adds to the THD % figure. I suspect as such, because distortion seems to go down as I increase level, up to a point.

The measurement mic has a distortion rating of 1% at full input at mid frequencies, so it's probably higher at very low frequencies and only kept down because the level at which the test is conducted is below the mic's max SPL. 110dB seemed to be the sweet spot between noise floor and room/mic distortion. It is safe to say that any distortion produced by the subwoofer is way below 1%.

If I could test in an anechoic chamber with a mic that has zero distortion, I could get a better picture of what the ZR18s are doing, but even with the distortion introduced by the environment and test setup, the figures look pretty darned good for a subwoofer.

Even with the sound system shut off, there is still a big spike at 60Hz on the spectrum coming into the distortion analyzer. No doubt that is contributing to the reading as it shows up as a third harmonic. Ground loops place a lower limit on THD tests because the noise is counted as distortion.

[TNRabbit 371]

That's AMAZINGLY low THD for bass. Most folks can't detect up to 10% or more distortion in low bass notes....

[RichP714 3,163]

As long as the signal falls within the range of the mic's linear region (110dB is probably near the upper limit of the mic) one would expect the result to be within the harmonic distortion spec of the mic itself (plus room boundary interactions if the measurement is not gated).

 

driver distortion is often quoted as 3-5%, but that's not harmonic distortion. It's a measure of the BL product nonlinearity (dynamic compression determined by voice coil position within the magnetic flux, voice coil temp, driver Xmax, etc.)

 

To measure that, you'd need a mic capable of measuring the driver into its nonlinear regions (above 120dB most likely) and you'd have to establish a reference (say, 100dB).

 

e.g. a 100dB/w/m driver exposed to 1W of power, measure the response, store in the SpecAn.; now increase power 20dB (1000W) and take another measurement; now subtract it from the reference.

 

This nonlinearity measurement would require a professional mic ($$$) and your drivers are probably loafing well within their linear region at 110dB......

[basspig 62]

Ah, indeed they are loafing.. probably about 1W of input power to achieve those levels.

David Lee, at Bassmaxx told me that when the lab in Pennsylvania did the Thiel-Small parameter testing, they got some incredible results, such that it prompted a call from the testing lab to David to discuss the unusual results. The BL is flat to about 3" of peak to peak excursion (Xmech is even higher at 3.6"), and the lab's test results indicated that with the rated 3600W of input power, the calculated SPL @20Hz would be 154dB. Since these were designed for use in a horn loaded system, I would presume those figures are for a horn loaded application. Power compression was said to be about 1dB at full power, which is also unusual, but then the drivers were built to rock football stadiums, not livingrooms, and as such are in a different league than what we're normally used to.

 

I figured out that my limiting factor was ambient noise. I've got about 75dB between the 110 and the 35dB ambient noise floor. That sets a theoretical low limit in the THD measurements. The microphone is the other unknown, since I know that at 1KHz it's 1% with something like 133dB, but that's probably the frequency where distortion is lowest. Just how much it rises as freq goes down, who knows. Also, those ECM8000 mics roll off several dB below 30Hz, so by 20Hz they may be less sensitive to the fundamental and more sensitive to the noise and harmonics. And finally, at -60dB is that 60 cycle hum. It's still there after I shut down the sound system, and appears to be a ground loop between mixer preamp and test equipment. I'll attempt this test again soon, after exploring alternative wiring methods to reduce the hum. The hum puts a hard lower limit of 0.1% THD+N on the readings. Add in room noise, room rattling/distortion, mic distortion, etc., and a lot of what I'm measuring isn't loudspeaker distortion at all.

[basspig 62]

I've been analyzing my test setup and have found some alternative wiring methods to get the 60Hz hum down another 20dB. In the process, I took distortion measurements on the AA501 analyzer and found that the mixer, preamp I'm using contributes .03% distortion itself. I am now taking the pre-EQ/Fader output to get rid of some noise and distortion. Another issue is the blocking capacitor downstream of the phantom 48VDC.. it's dropping the fundamental output because the response rolls off somewhat by 20Hz. 

 

 

(the 90Hz should be 80Hz --typo) 

 

 At least the setup is much cleaner now, and I measured the noise with the mic plugged in, and it is -84dB below preamp clipping, which represents 0dB at the top of the HP 3580A display. The noise floor is showing about -90dB at frequencies other than 60Hz.

As soon as I can get everyone out of the house, I'll conduct a followup test. Also try some different mics to see which mic has lower distortion.

 

 

 

 

 

 

[basspig 62]

Interesting contrast.. just measured my Dynaudio Acoustics speakers.. 9-1/4" long-throw woofers.

20Hz test signal, 3 watts power level, 90dB SPL attained on Cel 201/1 meter. 29.9% THD on Tektronics AA501.

[RichP714 3,163]

[peck555 366]

Wow

 

[basspig 62]

I switched out the ext proc loop and the Sonic Hologram, swapped the Behringer ECM8000 mic for the B-1 mic, and with 20Hz @ 110dB, the total harmonic distortion reading of the Bassmaxx woofers in operation shows 0.27% on the Tektronix AA501.

[RichP714 3,163]

What model SpecAn is that?  I searched for almost two years for an HP3580A; at the last moment I landed an HP 3585A (which needed repair, but I used to do that for a living, so I got it at a very reasonable price)

[basspig 62]

What model SpecAn is that?  I searched for almost two years for an HP3580A; at the last moment I landed an HP 3585A (which needed repair, but I used to do that for a living, so I got it at a very reasonable price)

 

This one's an HP 3580A. Got it for $212 on eBay and refurbished it myself (mostly cosmetic fixes).

How much did you spend for the 3585A? 

[RichP714 3,163]

 

 

 

What model SpecAn is that?  I searched for almost two years for an HP3580A; at the last moment I landed an HP 3585A (which needed repair, but I used to do that for a living, so I got it at a very reasonable price)

 

This one's an HP 3580A. Got it for $212 on eBay and refurbished it myself (mostly cosmetic fixes).

How much did you spend for the 3585A? 

 

3580A were going for more than I wanted to pay (about$700 IIRC), but I was about to bite the bullet anyway; then this 3585A came along (they were going for about $1500 in 'works' condition'), at $350; it had a problem in the IF section, but I was able to repair it.