I've just unpacked a pair of S/H Jaycar woofers.
Tested them with the DATS and OK but reading 30Hz rather than the catalogue 23Hz but +/- 10% so no surprise there but I hooked them up and generated a set of tones at 32/ 16 and 8 Hz and I can here them at both of the lower tones, and that shouldn't happen at 8Hz so what I'm hearing has to be overtones/ multiples of the base note.
Is there a fix for this?
Both woofers are the Jaycar carbon fibre woofer CW2145 from a decade ago and generally rated as reasonably good, I have another pair that do something similar but not as noticeable
Tested them with the DATS and OK but reading 30Hz rather than the catalogue 23Hz but +/- 10% so no surprise there but I hooked them up and generated a set of tones at 32/ 16 and 8 Hz and I can here them at both of the lower tones, and that shouldn't happen at 8Hz so what I'm hearing has to be overtones/ multiples of the base note.
Is there a fix for this?
Both woofers are the Jaycar carbon fibre woofer CW2145 from a decade ago and generally rated as reasonably good, I have another pair that do something similar but not as noticeable
Woofers can produce quite a bit of second and third harmonics if given a sufficient signal. A woofer may seem to pass a 20 hz test tone but you also want to be sure most of the output is really 20 hz.
Running a bass test using REW is a better option.
Running a bass test using REW is a better option.
Second hand/used woofers so I am assuming well broken in
Better explain REW.
I was generating the tones using Audacity
Better explain REW.
I was generating the tones using Audacity
This lovely tool is really good for speaker testing. https://www.roomeqwizard.com/
You need a microphone that works for the given frequency range.
Basically it can run a frequency response sweep test and give you distortion products along with frequency response.
Also really useful for testing crossover behavior when building a speaker.
You need a microphone that works for the given frequency range.
Basically it can run a frequency response sweep test and give you distortion products along with frequency response.
Also really useful for testing crossover behavior when building a speaker.
Almost any woofer from any manufacturer would distort (generate harmonics) significantly at an 8Hz frequency.
Dispute that. I have old school CV subs that do not and other that show much lower volumes of distortion
OK More playing around. It's probably going to be related to the limited excursion of these drivers. When dropping the levels the obtrusive harmonics diminish. I shouldn't really compare subwoofers with 14mm of one way excursion to woofers with only 5mm.
While the voice coils are large enough to handle 100 Watts modelling says 20Watts at 30Hz in a sealed Q-0.707 box; X-Max limited, naturally enough doubling up will help
While the voice coils are large enough to handle 100 Watts modelling says 20Watts at 30Hz in a sealed Q-0.707 box; X-Max limited, naturally enough doubling up will help
Not unexpected. DATS does not collapse the T?S curves in the same place as the typical factory software. Not to mention the T/s are guessed from the impedance curve instead of direct measurement. Good for matching drivers, i wouldnot design a box using them unless they were all i had (a Tannoy Red i’d put in a biigger box if i could do it again).
Not to mention i expect teh woofers have been sititng and coulduse some run-in.
dave
This is actually about subwoofers, but applies to hifi woofers. Lots of general info and good quality outdoor measurements too. Indoor measurements of bass tell more about the room than the woofer/box itself! Nearfield measurements easily have spl over 120dB and start to distort the mic and preamp!
Distortion is not linear, it rises very fast when speaker gets close to it's spl limit. Excessive distortion rises spl response at harmonics and lowers spl of fundamental!
https://data-bass.com/#/articles/5cc0bc36a75a260004255c88?_k=ix2ml1
Distortion is not linear, it rises very fast when speaker gets close to it's spl limit. Excessive distortion rises spl response at harmonics and lowers spl of fundamental!
https://data-bass.com/#/articles/5cc0bc36a75a260004255c88?_k=ix2ml1
Last edited:
Somewhat tangential/related, but what really is the source of these harmonics, especially the odd harmonics in woofers? I can understand if the voice coil movement is "inaccurate" due to eg hysteresis, but I would expect that to show up as random distortion, and not specifically at 2x or 3x the input signal. The surround moving makes sense too but ought to be much lower and (I would assume) at a lower frequency multiple (eg 0.5x). Or is the surround significantly responsible for odd harmonics?
A woofer is not a linear device. If you feed it a sine wave, you get a somewhat distorted wave. That can be decomposed into the fundamental and harmonics, using Fourier analysis.
The principal sources of non linearity are summarized in this poster: https://www.klippel.de/fileadmin/_migrated/content_uploads/Klippel_Nonlinearity_Poster.pdf
The principal sources of non linearity are summarized in this poster: https://www.klippel.de/fileadmin/_migrated/content_uploads/Klippel_Nonlinearity_Poster.pdf
At Xmax, THD, usually dominated by second harmonic, generally has reached 10%.
At 10% distortion, the harmonic is 20dB below the fundamental.
Looking at the equal loudness curves, we see a 90dB level at at 16Hz is just audible (for young adults with average hearing), while at 32 Hz ~60dB SPL can be heard.
If the driver was just reaching Xmax at 90dB, the second harmonic would be ~70dB.
At 1kHz, a 10dB change sounds twice (or half) as loud.
At 32 Hz, a 10dB change sounds ~four times (or 1/4) as loud, so as the harmonic level reaches ~65dB it may be perceived as loud as the fundamental at 90dB.
A sealed 12" woofer at 5mm excursion can produce ~84dB at 16Hz (likely to be inaudible other than the harmonic distortion), one with 15mm almost 10dB more, ~four times more perceived level.
When you compare subwoofers at low frequency, their Xmax difference really becomes obvious, especially when you can't hear the fundamental.
Art
Odd harmonics are created by symmetrical waveshape distortion, which is exactly what you would expect when the voicecoil excursion approaches Xmax.
If they were laying around, being busy gathering dust, the surrounds can harden again, especially subwoofers with thick surrounds. The rubber softener gasses out from the upper layer and it hardens. The surround becomes soft again if you break them in again, the rubber softener of the middle layer 'moves'/spreads to the outer layers again.
Something similar can happen to the sticky surround coating at PA drivers. There it is not a gas slowly escaping, it's the surface tension or tension of the molecules of the tough goo.
What power level were you driving them at and what excursion?
Does the the CW2145 have a vented pole piece? if so - was it free.
I have a pair of CW2110 but I'd hardly call them decent. In the sense, I'd expect high levels of bass distortion if measured, especially if driving them at 8Hz. It wouldn't take much power to take them to (Im guessing) a lowly 4mm likely Xmax and get all sorts of excursion related non-linear distortion higher up.
Does the the CW2145 have a vented pole piece? if so - was it free.
I have a pair of CW2110 but I'd hardly call them decent. In the sense, I'd expect high levels of bass distortion if measured, especially if driving them at 8Hz. It wouldn't take much power to take them to (Im guessing) a lowly 4mm likely Xmax and get all sorts of excursion related non-linear distortion higher up.
Maybe 10 watts and the vent was free.
The is a huge difference between the single wound CW2110 8 inch [ I have a quad used OB as midbass] and the Carbon Fibre CW2145 but all Jaycar drivers are X-Max limited even if the voice coils can handle hundreds of Watts, 4mm would be safe but 6mm is the limit I think
The is a huge difference between the single wound CW2110 8 inch [ I have a quad used OB as midbass] and the Carbon Fibre CW2145 but all Jaycar drivers are X-Max limited even if the voice coils can handle hundreds of Watts, 4mm would be safe but 6mm is the limit I think