This is not so simple as speaker SPL is not directly dependent from cone amplitude but from cone velocity.
I had symmetrical H-frame OB bass with 2 drivers where one driver is flipped and its THD is mostly 2nd harmonic except in region (lower than 40 Hz on image, heavily EQ-ed) where cone amplitude will grow close to Xmax.
How does one reduce 3rd and other odd orders? Is it mainly related to Sd(x), BL(x), CMS(x) or Le(x). Or multiple factors?
Charlie, I honestly think you use quad P835016 (10" alu XXLS) or 17 (12" alu XXLS) (if you can get them) in series parallel, or wait for the Purifi 10". The same fellows built both...
Charlie, I honestly think you use quad P835016 (10" alu XXLS) or 17 (12" alu XXLS) (if you can get them) in series parallel, or wait for the Purifi 10". The same fellows built both...
We must remember, that a loudspeaker is more than a driver. Klippel test and T/S parameters are the foundation yes, but...
Kaamelis's measurement is not nearfield, but obiously at some distance. To get high enough spl (and excursion) to show relevant distortion behaviour above noise floor of the environment, also introduces vibrations in the stand of the speaker and room (if done in-room)
I have done manymany indoor and outdoor measurements of drivers and speakers with microphone, so I have become very cautious about a single result, I mean making conclusions and deductions on it.
Also, the quest for the best or perfect bass driver for dipole, will be very difficult, one should decide what parameters to use, but not at the expence of all the others. High Sd and high number of drivers always helps, I suppose. I haven't tested compound systems at all, I am very happy with a single nude 12" pro driver highpassed around 150Hz LR2. Below that a sealed box woofer (in each speaker) gives the thump that I like and all the spl I can use in my living room.
This measurement outdoors at roughly 1,5m
And this same speaker indoors at roughly 2,5m (a bit different eq settings)
Kaamelis's measurement is not nearfield, but obiously at some distance. To get high enough spl (and excursion) to show relevant distortion behaviour above noise floor of the environment, also introduces vibrations in the stand of the speaker and room (if done in-room)
I have done manymany indoor and outdoor measurements of drivers and speakers with microphone, so I have become very cautious about a single result, I mean making conclusions and deductions on it.
Also, the quest for the best or perfect bass driver for dipole, will be very difficult, one should decide what parameters to use, but not at the expence of all the others. High Sd and high number of drivers always helps, I suppose. I haven't tested compound systems at all, I am very happy with a single nude 12" pro driver highpassed around 150Hz LR2. Below that a sealed box woofer (in each speaker) gives the thump that I like and all the spl I can use in my living room.
This measurement outdoors at roughly 1,5m
And this same speaker indoors at roughly 2,5m (a bit different eq settings)
Thank you Juha,
The requirements for dipole woofers are somewhat unique due to increased excursion requirements but also noises not directly related to cone/motor movement.
I am reminded of some comments from Josh Ricci; who has probably tested and provided results for more large high excursion drivers than anybody
https://data-bass.com/#/articles/5cc0bc36a75a260004255c88?_k=vbfc4v
Re-
Long term output sweeps
“Ideally what would be seen in this measurement is the least amount of response change throughout the entire bandwidth at the highest sweep levels. The DUT would produce the exact same response with every increase in output up to extremely loud SPL's and the repeat lower volume measurement would overlay the original exactly afterwards. In reality this never happens. Every subwoofer has limitations on its output and every one of them is driven to them during this testing. Voice coils heat up, impedance changes, air speeds (can) get high, linear excursion is exceeded and amplifiers run out of voltage or current…
This is the most demanding measurement type conducted during Data-Bass testing. It will mercilessly reveal any issues with: Overload, (port compression), air noise, driver distress, cabinet or grille rattles, buzzes, harsh amplifier clipping, etc... This testing is conducted outdoors, so there will be no nearby walls or objects to provide low frequency gain, no vibration from walls or objects and no other speakers are operating so there is no upper frequency content to mask the subwoofer.
These things would normally help to cover up objectionable noises from the subwoofer in a typical room, or operational setting. This testing is pretty much the worst case scenario for a subwoofer, putting it under the proverbial microscope with test signals. Any sort of audible distress or issues with the subwoofer will be readily apparent in this environment. Much more so than in a typical room with accompanying speakers and electronics while listening to music or movies.”
I don’t recall if Charlie has SPL/Hz targets, but he does allude to the fact that he would prefer not to have subwoofers; if possible. He mentions 30Hz as a target for music.
I would agree with 30Hz; and I would think that hitting the reference level of 105dB at the listening position is wonderful.
I’ve measured “only 100dB” for my twin sealed unequalised woofers with an F6 of 40Hz, and for me that is enough.
Hitting 30 or 40Hz for dipole woofers at these levels is a tough task.
The requirements for dipole woofers are somewhat unique due to increased excursion requirements but also noises not directly related to cone/motor movement.
I am reminded of some comments from Josh Ricci; who has probably tested and provided results for more large high excursion drivers than anybody
https://data-bass.com/#/articles/5cc0bc36a75a260004255c88?_k=vbfc4v
Re-
Long term output sweeps
“Ideally what would be seen in this measurement is the least amount of response change throughout the entire bandwidth at the highest sweep levels. The DUT would produce the exact same response with every increase in output up to extremely loud SPL's and the repeat lower volume measurement would overlay the original exactly afterwards. In reality this never happens. Every subwoofer has limitations on its output and every one of them is driven to them during this testing. Voice coils heat up, impedance changes, air speeds (can) get high, linear excursion is exceeded and amplifiers run out of voltage or current…
This is the most demanding measurement type conducted during Data-Bass testing. It will mercilessly reveal any issues with: Overload, (port compression), air noise, driver distress, cabinet or grille rattles, buzzes, harsh amplifier clipping, etc... This testing is conducted outdoors, so there will be no nearby walls or objects to provide low frequency gain, no vibration from walls or objects and no other speakers are operating so there is no upper frequency content to mask the subwoofer.
These things would normally help to cover up objectionable noises from the subwoofer in a typical room, or operational setting. This testing is pretty much the worst case scenario for a subwoofer, putting it under the proverbial microscope with test signals. Any sort of audible distress or issues with the subwoofer will be readily apparent in this environment. Much more so than in a typical room with accompanying speakers and electronics while listening to music or movies.”
I don’t recall if Charlie has SPL/Hz targets, but he does allude to the fact that he would prefer not to have subwoofers; if possible. He mentions 30Hz as a target for music.
I would agree with 30Hz; and I would think that hitting the reference level of 105dB at the listening position is wonderful.
I’ve measured “only 100dB” for my twin sealed unequalised woofers with an F6 of 40Hz, and for me that is enough.
Hitting 30 or 40Hz for dipole woofers at these levels is a tough task.
Last edited:
Have you considered GR research servo-controlled open baffle woofers?
I decided on these after having tried AE 18" ob woofers. A set of three each side.
It is by far the best bass I have heard. punchy, fast, detailed, nuanced, and goes down to 20 hz.
And, it fits the 'expensive' tag as well.
The Priest
I decided on these after having tried AE 18" ob woofers. A set of three each side.
It is by far the best bass I have heard. punchy, fast, detailed, nuanced, and goes down to 20 hz.
And, it fits the 'expensive' tag as well.
The Priest
You are correct, it was made from listening position.
I now find also more nearfield measurements on level where mic amp is nearly clipping (absolute level somwhere close to 100 dB on dipole peak), 80 cm from drivers plane, mic exactly on middle distance from drivers. EQ is +6 dB/octave down from 100 Hz, LP filter is on 200 Hz.
To get random noise floor down it is 8 times repeated measurement.
I too am looking to build a compact, high quality OB. Regrettably, my budget is smaller, so I'm considering a SLOB with 4 Anarchy 708s per side. That said, I'll definitely be following this thread for "inspiration".
Nearfield measurements are not that helpful for dynamic range and/or distortion measurements for woofers. Mainly because it overloads the mic. Suppose you’re expecting a woofer to hit 90dB @ 1m, which is not unreasonable. At 1cm that’s 130dB well beyond the capabilities of most electret condenser mics that hobbyists use.
Recently I did some nearfield measurements at 1cm, then double the distance gradually eg. 2cm, 4cm, 8cm, 16cm until 1D, 2D, 3D; D representing the diameter of the woofer, as well as 1m and then 2m ground plane measurements. all with the same woofer, in 1 cu ft cabinet heavily filled with fibreglass wool.
I found that measuring in the nearfield was excellent for determining the anechoic frequency response up to about 1Khz (exact figure depends on size of cone diameter). BUT, I did easily overdrive mic and forgot that I was sending a full 2.83V!
So a 85dB/2.83V@1m woofer; the mic is trying to
anywhere above 1-2W eg. 2.83V or 4V and the mic is supposed to read 125dB at 1cm. Well, compression sets in and the mic didn’t like that; showing up high levels of H2 distortion.
Without an anechoic chamber ) or Klippel Near Field Scanner) I’ve found that the best way measuring large woofers is at a distance like 1m. If you have a small room eg. 10m^2; or a room with lots of background noise, then you may need to move the mic closer eg. 50cm (+6dB) or 31.6cm (+10dB). But if room is and large then (eg. 30dB(C) then measuring at 2m or more on the ground plane is better.
Stress testing super subwoofers; yeah; I’d take it into a basketball court and measure at 5m at 2AM. But that is kind of nutty professor type work.
Last edited:
Yes.
Real nearfield measuremnt of OB speakes in not very corrct way, specially when OB has 2 drivers and if one is flipped.
Ah, sorry for any confusion. I was directing my remark to the OP, not to you. But I am curious now since you believe the HiFiCompass nearfield measurements to either have been done incorrectly, or in any case done in such a way that the results are not useful?
In that case, yes I certainly have. This has driven my choice of driver for both current and future systems.
Was there something in particular you would like to point out?
Also, at this time I am leaning towards using multiple smaller drivers, e.g. four 10" per side. I currently have eight ScanSpeak 26W/4534 woofer that I will start working with as soon as it is again warm enough to do outdoor measurements here (in Michigan). Depending on how that all works out I wilil decide on the direction to take for future projects of this type.
@454Casull
Yevgeniy’s setup; he uses a lab grade mic capsule and associated pre-amplifier electronics and can measure up to 160dB.
160dB @ 20mm = 126dB @ 1m
@HiFiCompass measurements are VERY useful. Read with confidence and donate if you find it useful.
Yevgeniy’s setup; he uses a lab grade mic capsule and associated pre-amplifier electronics and can measure up to 160dB.
160dB @ 20mm = 126dB @ 1m
@HiFiCompass measurements are VERY useful. Read with confidence and donate if you find it useful.
@CharlieLaub Can you explain what it was about pro drivers that you didn't like for this application?
@ianbo Pro drivers are optimized for high efficiency. Part of this is making the cone relatively light weight to the detriment of cone resonances and damping. I was able to perceive their presence around 200-250 Hz for many 18" or larger drivers and I need to cross over a bit higher in frequency in my project. So I came up with the idea of using multiple smaller drivers in a square baffle that is about the same width, e.g. 20". This has worked well in many ways and avoids these resonances because they are occuring much higher in frequency for e.g. a home audio 8" driver.
No, HiFiCompass nearfield measurements of drivers alone are correct. But if drivers are installed in OB frame and target is to measure asymmetries cancellation caused by one driver flipp, then very driver cone close nearfield measurements cannot gve reasonable result.
I made better measurements to take clear asymmetries cancellation caused by one driver flip.
Both are made at same voltage level on drivers, room correction EQ and filter is turned off, +6db/octave compensation down from 100 Hz remains on.
Fist is single driver working, neafield 1 cm from cone center.
Second is 60 cm from drives installation plane, on same distance from both drivers center, one driver is flipped.
Mic amp gain is changed to keep maximum signal close to full scale 0 dB, both are 8 times repeated measurements.
Nearly 10 db of 2nd harmonic improvement can be seen on some frequencies. But general THD is same, where 2nd harmonic is decreased, there 3nd harmonic is increased.
Both are made at same voltage level on drivers, room correction EQ and filter is turned off, +6db/octave compensation down from 100 Hz remains on.
Fist is single driver working, neafield 1 cm from cone center.
Second is 60 cm from drives installation plane, on same distance from both drivers center, one driver is flipped.
Mic amp gain is changed to keep maximum signal close to full scale 0 dB, both are 8 times repeated measurements.
Nearly 10 db of 2nd harmonic improvement can be seen on some frequencies. But general THD is same, where 2nd harmonic is decreased, there 3nd harmonic is increased.
Charlie, how will you arrange / configure the 4 drivers per side? I have 8 x AE Dipole 12 and I’m currently trying to decide how to implement them - I’m seriously considering cloning the Model B woofer arrangement albeit with 4 not 6 woofers.
My current system uses a 20"x20" baffle that sits on the floor. This holds four 8" drivers, just barely. The paradigm for the system is "no or minimal baffle". There is no baffle at all for the other drivers - they are suspended by wires inside a metal cage. This is why I cannot operate the woofer to all that low of a frequency, because the baffle losses would be too large. Anyway I prefer sealed subs for the lowest octave or so, so I cross over to them around 70Hz.
You can certainly clone the Model B, using a solid baffle and four 12s per side instead of 6. Nothing wrong with that, just not my particular design style for dipole systems.
You can certainly clone the Model B, using a solid baffle and four 12s per side instead of 6. Nothing wrong with that, just not my particular design style for dipole systems.