Hi,
Eg. : http://medleysmusings.com/kef-q100-...fer-displacement-on-tweeter-response-example/
I have 2 questions here :
1) I know that high pass will help here, but would like know the threshold value for woofer movement (in mm) below which impact on HF response would be inaudible. Based on that and the intended SPL, one can arrive at HP freq value for other size coaxials..
2) In the Q100 example, I wonder why the impact is pronounced only between 6kHz-9kHz ( and then again above 15kHz) ?
Thanks !
Eg. : http://medleysmusings.com/kef-q100-...fer-displacement-on-tweeter-response-example/
I have 2 questions here :
1) I know that high pass will help here, but would like know the threshold value for woofer movement (in mm) below which impact on HF response would be inaudible. Based on that and the intended SPL, one can arrive at HP freq value for other size coaxials..
2) In the Q100 example, I wonder why the impact is pronounced only between 6kHz-9kHz ( and then again above 15kHz) ?
Thanks !
thanks - extensive testing - static offset of cone -- I'd be more interested in its modulation distortion with a low frequency moving the woofer section and a high frequency tone such as a high flute note "gargle effect" - its a wild looking little speaker - - when the woofer was biased in the rear position, cavity depth appeared greater -- its a small coaxial speaker so perhaps it doesn't take much to alter the tweeter response.
Last edited:
Yes, Erin has extensively tested Q100 and few other KEF drivers (and many other non coax as well) . Thanks to him! ( He is a member on this forum too - bikinpunk)
eg. TAD Reference speakers and KEF R, Q and References speakers have coaxial mid-tweeters. I don't hear complaints about modulation artefacts. A 2-way coaxial might be a different story, but they are not bad either, limited bass capability is the worst issue. There are obviously many far more important details than modulation.
KEF R700 loudspeaker | Stereophile.com
KEF Blade Two loudspeaker | Stereophile.com
KEF LS50 Anniversary Model loudspeaker | Stereophile.com
TAD Evolution One loudspeaker | Stereophile.com
TAD Compact Reference CR1 loudspeaker | Stereophile.com
KEF R700 loudspeaker | Stereophile.com
KEF Blade Two loudspeaker | Stereophile.com
KEF LS50 Anniversary Model loudspeaker | Stereophile.com
TAD Evolution One loudspeaker | Stereophile.com
TAD Compact Reference CR1 loudspeaker | Stereophile.com
Hi Juhazi,
The 3-ways would avoid this by limiting the excursion.
Though LS50 should have this....
(and my first question was related to threshold of cone movement)
As for the stereophile measurements :
Forgive my limited knowledge, but which measurement would involve both LF and HF tone playing together , as a simple freq sweep wouldn't reveal this, right ?
The 3-ways would avoid this by limiting the excursion.
Though LS50 should have this....
(and my first question was related to threshold of cone movement)
As for the stereophile measurements :
Forgive my limited knowledge, but which measurement would involve both LF and HF tone playing together , as a simple freq sweep wouldn't reveal this, right ?
Last edited:
I was not referring to measurements, but auditioning - those speakers are praised for natural and precise, uncoloured sound.
A dual tone distortion test would perhaps reveal modulation?
A dual tone distortion test would perhaps reveal modulation?
You would think it would pretty easily...setting a tone at LF driver FS and looking at the resultant intermodulation products. The catch would be making sure you had enough frequency resolution to clearly separate the intermod products from the shoulders of the hf test tone.
Scott
With concentric drivers, the main form of modulation distortion is amplitude modulation from the change in frequency response as the cone is displaced forwards and backwards. This effect can be seen by measuring the tweeter response while physically displacing the cone. Two things affect this: the design of the fixed waveguide between the tweeter and the apex of the cone, and the shape of the surround.
The effect of Doppler distortion due to the velocity of the cone is a non-event, the ear is very insensitive to Doppler distortion. That this is so can be easily understood from listening to a full range electrostatic speaker: the tweeter signal is riding on top of the diaphragm that is also producing the bass. No-one has ever thought to question the Doppler distortion of these speakers!
The way to minimize amplitude modulation distortion is to simple minimize the cone displacement.
In my TAD speakers I do this by going to a three way system. This is always the best way to use a concentric driver.
Andrew
The effect of Doppler distortion due to the velocity of the cone is a non-event, the ear is very insensitive to Doppler distortion. That this is so can be easily understood from listening to a full range electrostatic speaker: the tweeter signal is riding on top of the diaphragm that is also producing the bass. No-one has ever thought to question the Doppler distortion of these speakers!
The way to minimize amplitude modulation distortion is to simple minimize the cone displacement.
In my TAD speakers I do this by going to a three way system. This is always the best way to use a concentric driver.
Andrew
Appreciate your response Andrew!
Just coming to specifics ( as I have mentioned in my first post ) what would be the approx threshold value of cone excursion, below which the AM distortion will become inaudible ?
And is this AM distortion always for certain range as seen in the link ?
Thanks a lot!!!
Just coming to specifics ( as I have mentioned in my first post ) what would be the approx threshold value of cone excursion, below which the AM distortion will become inaudible ?
And is this AM distortion always for certain range as seen in the link ?
Thanks a lot!!!
I probably have it in one of my JAES journals somewhere I've got 20 years of back issues...I'll have to look it up. I disagree with Andrew's assertion that electrostatics are a reason to believe that it's not very audible, even though hf and lf are on the same diaphragm element. The excursion (and therefore velocity) is very small on an electrostatic so I don't think it's a valid assumption that audibility is low.
-Scott
Last edited:
Yep, and is there an electrostatic that can reproduce high levels of low frequencies?
It is harmonics that teh ear is not particularly sensitive to. It is widely acceptance that the ear IS sensitive to intermodulation distortion.
Doppler distortion is just another form of intermodulation distortion, thus the ear is sensitive to it. That this is so is easily seen when you consisder it is a faorm of frequency modulation: a lower frequency tone causes the amplitude of of the higher frequency to be slightly reduced, and the energy transfered to sidebands, excatly as happens with FM radio transmission.
Dopler is not a issue with full range electrostattics becasue their cone area is huge, and so their back-and-forth travel is very small.
Coaxial and triaxial drivers with pretensions to high quality have separate voice coils for each cone - they are no diffrent to any other 2-way or 3-ways system.
Except for a big 1940's Jensen, such bi-axial or tri-axial drivers have the woofer at the back, and the midrange and or/tweeter at the front. Therefore any interference from the woofer cone is indirect, arising from diffraction of the midrange & tweeter output reflecting off the woofer cone. Therefore doppler distotion cannot be as great as with a single (dual cone or ridged) widerange driver, The reflection would make the sidebands further apart, but the amplitude would not be altered by the reflection. In fact, reflection will not be 100% effective, so the situation is even better.
I would expect doppler distortion to be a lot less than for a single widerange driver. Not very different to a conventional 2-way or 3-way system.
It would be interesting to set up a test with a probe microphone 15-20 mm in front of the tweeter, and another within the woofer cone space. and comparing the tweeter generated levels. If there is only a few dB difference you can ignore moving waveguide effects.
Last edited:
"The excursion (and therefore velocity) is very small on an electrostatic"
You mean I can't compare a Quad 57 with it's over 700 square inches of area with a 6-1/2" woofer with its 22 square inches of area?
PWK published measurements on an EV 15TRX with the same componenets with the tweeter mounted on the baffle.
The BBC used a perforated mesh in front of a 15" woofer that had the tweeters mounted in the woofer opening of the cabinet.
CV has a patent on a co-ax with perforated mesh in front of the woofer cone ( Patent #: US004619342, Patent #: US004283606 ).
You mean I can't compare a Quad 57 with it's over 700 square inches of area with a 6-1/2" woofer with its 22 square inches of area?
PWK published measurements on an EV 15TRX with the same componenets with the tweeter mounted on the baffle.
The BBC used a perforated mesh in front of a 15" woofer that had the tweeters mounted in the woofer opening of the cabinet.
CV has a patent on a co-ax with perforated mesh in front of the woofer cone ( Patent #: US004619342, Patent #: US004283606 ).
I know right!
For the older co-axial units (UREI, etc) and modern units where the waveguide is well in front of the woofer cone I don't see an issue occurring with doppler distortion between the tweeter and woofer. Diffraction is another issue and somewhere between lots of it, and approaches taken by Geddes to eliminate it is probably a nice middle ground.
For units where the woofer cone can be seen by the tweeter (Seas, KEF, some pro units like FaitalPRO, B&C...etc) I've always wondered why no one did measurements on the impact of cone movement to the frequency response of the tweeter. The Medley musings measurements show that well for the Q100 in a static condition...the woofer cone isn't moving. You're seeing the diffraction/reflection effects of the woofer disappearing from acoustic view and then appearing. You're not seeing the modulation of a tone generated by the movement of a second source.
I don't own any of the B&C or Faital (or 18Sound, or Seas) units I'm curious about, and I won't buy them just to find out. If someone wants to send me one....I'll measure it and pay return shipping. If you're in Central Texas and want to spend an early Sunday morning out in a parking lot measuring speakers....PM me. ;-)
Scott
Wouldn't a fullrange driver of equal size suffer from comparable doppler distortion ?
Also, in every multi-way loudspeaker (especially 2-way ls) the tweeter 'sees' a substantial part of its baffle moving.
I've never read anything indicating people worrying about that enough to measure the effect.
Also, in every multi-way loudspeaker (especially 2-way ls) the tweeter 'sees' a substantial part of its baffle moving.
I've never read anything indicating people worrying about that enough to measure the effect.
This is a diffraction effect based on the diameter of the diffraction "ring" around the tweeter. The fact that it is a "ring" exaggerates the effect due to all the edge sources of the diffraction "concentrating" on the 0 degree axis. It won't be the same as you move off-axis.
Though not shown, it is an effect that is not only relative to the degree of offset (either inward or outward), but ALSO the playback level. (..twice-over when you consider that amount of diffraction is also dependent on the woofer's position.)
This same problem shows up numerous times with tweeters that either either "proud" or are to far inset into the baffle as measured in Soundstage's NRC measurements under the category of "Deviation from Linearity".
Ex.
SoundStage! Hi-Fi | SoundStageHiFi.com - Magico S5 Loudspeakers
SoundStageNetwork.com | SoundStage.com | NRC Measurements: Magico S5 Loudspeakers
Attachments
Last edited:
- Status
- Not open for further replies.