Member
Joined 2003
Damped properly in a sufficiently sized and less tube shaped chamber, the reflection would be non-existent. SB has the engineering know-how to design a large cup with stuffing in it, you need to look no further than something like Morel MDM55 to see what that looks like. This reflection is purposely engineered, I am just a bit baffled to understand it.
The large flange is what irks me. To use it with a high crossed tweeter would require one that has a very small flange. I guess that's why they sent a few to be tested with it. Still though, if Morel can do a small flange, SB can too. It doesn't look as cool, but form should follow function
Last edited:
HiFiCompass,
Do you still have a W22EX001 kicking around? Would you be amenable to doing a one-off test on it, with the following condition:
- Woofer to have a low-pass filter with corner frequency around 1.4 kHz (either with passive components or DSP)
- Test the harmonic distortion behaviour with a sine sweep
I believe that the rise in H5 from 600 Hz to ~1.5 kHz would stay the same whether the driver is unfiltered OR has a low-pass filter that attenuates the breakup peak around 5 kHz, as long as the distortion is tested using a sine sweep (as opposed to a log-swept sine chirp or something similarly broad-spectrum). Would really really appreciate if you could confirm or disprove that theory.
Do you still have a W22EX001 kicking around? Would you be amenable to doing a one-off test on it, with the following condition:
- Woofer to have a low-pass filter with corner frequency around 1.4 kHz (either with passive components or DSP)
- Test the harmonic distortion behaviour with a sine sweep
I believe that the rise in H5 from 600 Hz to ~1.5 kHz would stay the same whether the driver is unfiltered OR has a low-pass filter that attenuates the breakup peak around 5 kHz, as long as the distortion is tested using a sine sweep (as opposed to a log-swept sine chirp or something similarly broad-spectrum). Would really really appreciate if you could confirm or disprove that theory.
Last edited:
Member
Joined 2003
You don't need to measure again to prove that theory, it has been done. Harmonics that excite the resonant still do so after you filter out the resonance, because they occur after the filter within the driver itself. The resonant peak still exists with a filter applied, the only way around it is to change driver characteristics such as cone material and damping, etc so the driver has less of a peaky resonant breakup. The W22 should ideally be crossed around 1kHz or lower to avoid the 3rd order harmonic peak, or 1/3 the frequency of where the breakup peak starts.
Cone resonances do have rise and decay times, true. But those are short enough to show up with most test signals. Five cycles at 5kHz is only 1ms... any serious gated measurement would have 5ms or more. For the sake of it you could measure close range with a 20ms gate. That would be more than enough to reach steady state and show up in the averaged level.
Yes, I have, but there is no need in repeating such a measurement again as I did it many times with different speakers. The DSP filtering doesn't work in this case. The only way is using parallel LCR connected in series with a speaker and tuned at breakup peak frequency.
Thanks for the response. The series notch could definitely flatten out the peak in terms of the frequency response, but are you saying that the notch can actually reduce the amplified subharmonic distortion (e.g. H5 from 600-1500 Hz in the W22EX) as measured with a sine sweep? How is that possible when the notch has no acoustic effect on the driver's self-generated distortion?
Yes, of course. The notch can down 3rd and 5th HD peaks corresponding to the breakup resonance as much as the peak is attenuated. This distortion is born in a speaker's motor and manifests itself as a distortion voltage source in the voice coil circuit. This voltage source creates the voice coil current, which drives the voice coil and is subjected to the resonance amplification by the cone at the breakup frequency. You need to break circuit for voice coil curent at breakup frequency by adding hi-impedance notch in series with the voice coil. It really works.
understanding measurment effects.
For sure people would like to see IMD measurments like S. Linkwitz did for few speaker in a study iirc (Pluto upgrade speakers choice?).
Slighty off-topic: What is the amount in db of H3 (and H5) you guys can accept with your loudspeakers? -40 db and -60 db ? Can you live with -30 db H3 in the mid ?
Do you find it more harmfull even if it is low but above H2 for the brain ?
For sure people would like to see IMD measurments like S. Linkwitz did for few speaker in a study iirc (Pluto upgrade speakers choice?).
Slighty off-topic: What is the amount in db of H3 (and H5) you guys can accept with your loudspeakers? -40 db and -60 db ? Can you live with -30 db H3 in the mid ?
Do you find it more harmfull even if it is low but above H2 for the brain ?
It isn’t about breakup, it’s about motor-induced distortion. Which could be reduced by altering the current flow through the voicecoil. Think of it as a passive feedforward solution.Hmm, I thought not all breakup modes will show up in the voice coil (back emf). Well, I guess impedance sweep shows what does and what doesn't show up there
This is influenced by if you are a H2 or H3 kind of guy.
From Nelson Pass, "Audio, Distortion, and Feedback" Article
PassDiy
"Many audiophiles believe that 2nd harmonic is to be preferred over 3rd harmonic. Certainly it is simpler in character, and it is well agreed that orders higher than third are more audible and less musical. However when given a choice between the sound of an amplifier whose characteristic is dominantly 2nd harmonic versus 3rd harmonic, a good percentage of listeners choose the 3rd.
I have built many examples of simple 2nd and 3rd harmonic “types” of amplifiers over the last 35 years. When I say “types” I mean that they used simple Class A circuits described as “single-ended” versus “push-pull” and so tended to have a 2nd harmonic versus 3rd harmonic in the character of their distortion, but were not made to deliberately distort.
Anecdotally, it appears that preferences break out roughly into a third of customers liking 2nd harmonic types, a third liking 3rd harmonic, and the remainder liking neither or both. Customers have also been known to change their mind over a period of time.
However the issue is partially obscured by the fact that the 3rd harmonic type amplifiers usually have lower total distortion. Third harmonic usually appears with a negative coefficient, resulting in what we think of as “compressive” - the example in figure 3. It's worth noting that odd orders on nonlinearity also can be seen altering the amplitude of the fundamental tone -something a distortion analyzer doesn't ordinarily display."
In my personal, not-backed-by-science view, not more than the following with loud-ish music at listening position:
- 30-40 dB H2
- 45-50 dB H3
- 60-65 dB H4
- 65-70 dB H5
Last edited:
Thank you Twinter.
I was talking about drivers measurements as performed by Hificompass here (graphs). Do you mean the harmonic distorsion measurment of drivers whatever the technic : Dayton V device, Clio, Arta-Box, etc is dependant of the outputted harmonics type of the amp in the measurment chain as well ?
I was talking about drivers measurements as performed by Hificompass here (graphs). Do you mean the harmonic distorsion measurment of drivers whatever the technic : Dayton V device, Clio, Arta-Box, etc is dependant of the outputted harmonics type of the amp in the measurment chain as well ?