Re the 'N' part -that's just a typo; one which I should have picked up on (evidently can't see for looking) because I normally do an editorial check for them. I don't write the copy, but I do normally look over the data sheets to check grammar, spelling etc. In this case, I missed it (mea culpa) so if blame, incompetence or accusations of lying are to be raised and apportioned: please direct them at me.
Sd is normally taken by manufacturers as cone diameter, plus part of the surround diameter -usually 1/3 - 1/2 but sometimes something different depending on its profile and flex ratio as it forms part of the emitting area. The most extreme example of this can be seen in tweeters: that's actually what a ring-radiator is -a flexing surround or several concentric roll surrounds. That's also why they & regular domes with large roll surrounds have rising levels of HD2 < 2KHz, but I digress.
Fs is the resonant frequency of the driver. It doesn't mean the driver is flat to Fs: the mathematical -3dB mass-corner rolloff under anechoic IB conditions is taken as 2Fs/Qes so how low it goes 'au naturel' depends on Fs and Q. The actual detail varies in practice from the electrical filter derivation depending on the suspension / cone design though, as well as the measurement conditions, so like all nominals the theoretical mass corner frequency should be taken as a guideline rather than as a literal.
Sd is normally taken by manufacturers as cone diameter, plus part of the surround diameter -usually 1/3 - 1/2 but sometimes something different depending on its profile and flex ratio as it forms part of the emitting area. The most extreme example of this can be seen in tweeters: that's actually what a ring-radiator is -a flexing surround or several concentric roll surrounds. That's also why they & regular domes with large roll surrounds have rising levels of HD2 < 2KHz, but I digress.
Fs is the resonant frequency of the driver. It doesn't mean the driver is flat to Fs: the mathematical -3dB mass-corner rolloff under anechoic IB conditions is taken as 2Fs/Qes so how low it goes 'au naturel' depends on Fs and Q. The actual detail varies in practice from the electrical filter derivation depending on the suspension / cone design though, as well as the measurement conditions, so like all nominals the theoretical mass corner frequency should be taken as a guideline rather than as a literal.
Last edited:
To be fair, I very much doubt any driver manufacturer can claim 100% accuracy with their parameters tbh. It will vary to a greater or lesser degree between batches? That said I wouldn't say the same for published frequency response graphs, as I have read stories of inaccuracy.
If they do (claim 100%) you know they're lying. MA's usual per-batch QC tolerance is +/-3% to Fs assuming identical measurement conditions* although they sometimes make changes for production reasons (usually consistency) that will shift an average.
*And therein is a merry time for many manufacturers -Seas, Scan, SB, MA, TB, Dayton, Fane -you name it. Because they often measure under voltage drive conditions that a lot of affordable home measurement options can't provide (DATS for example -and I'm not knocking DATS because I own it, use it and like it). They're usually more representative (especially for 'modern' low-VAS, lower efficiency type drive units) of the actual operating conditions, but since T/S parameters vary with voltage drive, climatic conditions and a host of other factors that does mean you can get variations -and that's before you consider suspension break-in etc. Which isn't to make excuses: I could name a few that have some rather wild variations, including one extremely well known and well-regarded brand who use +/- 10% of Fs and sometimes have some wildly different values on everything else (in fairness, I've also had plenty of their products through here that measured very close on most values to stated)
*And therein is a merry time for many manufacturers -Seas, Scan, SB, MA, TB, Dayton, Fane -you name it. Because they often measure under voltage drive conditions that a lot of affordable home measurement options can't provide (DATS for example -and I'm not knocking DATS because I own it, use it and like it). They're usually more representative (especially for 'modern' low-VAS, lower efficiency type drive units) of the actual operating conditions, but since T/S parameters vary with voltage drive, climatic conditions and a host of other factors that does mean you can get variations -and that's before you consider suspension break-in etc. Which isn't to make excuses: I could name a few that have some rather wild variations, including one extremely well known and well-regarded brand who use +/- 10% of Fs and sometimes have some wildly different values on everything else (in fairness, I've also had plenty of their products through here that measured very close on most values to stated)
They can only give a statistical estimate based on a measurement of a sample set of a production run. MA for instance takes 50 random drivers from a batch, measures them, then do an average. Now, MA drivers are fairly tight QC wise, i have mesured many sets of 20+ drivers and typically the core set of T/S vary by plus-minus 8-9% (so 2 drivers at the opposite ends of the sameple could be 18% apart). Fostex samples were typically not quite double that, and others, which i won’t mention, are even further apart, i have had samples that are sp far apart it is hard to call them the same driver.
Hence why i went to so much effort to. ship matched pairs.
dave
Why i, unless unavailable, start designs using the factory data. It typically gives me a box that works first time, using data i created, i usually want a second crack after hearing the result. But i have measured many 1000s of drivers and use the data to match them into pairs (or occasional quads, and a couple octs).
dave
Dave you(I learn about MA from your FH and Onken ) , Scott (enclosure designer), Nandappe (with his excellent videos in Youtube) , Pearl Acoustics, ... to name some, you are very important for the commercial success of MA. You Dave you are a big client also. So you are in possession to negotiate directly with Fenlon and demand for "matched pairs" . But me, (maybe we) the simple diyer(s) I am not ! Furthermore a pair of drivers will be cost me much more expensive if I buy from you an "enabled" pair (the pair itself + your work on them + shipping from China to Canada + shipping from Canada to Greece) . In conclusion, trying to design a speaker seems completely futile. I just have to replicate Scott's "blueback" or Nandappe's SDDBH-T13.5R-AD design & just hope to achieve something "just acceptable". Not forget to mention the amplifier. To explain myself ! I have seen SPL plots for specific driver (not the CHP-90) that differs considerably, mainly in the lower bass-band, , for SE-tube amplifier, AB-class tube amplifier, AB-class transistor amplifier, AB-class transistor amplifier operating in current source mode, D-class amplifier.... All the above seems to me like no exit road .
I don't like the taste of disappointment at all
Elias
As a handful of minor points re some of the above, which may be useful:
-'Fenlon' does not own Mark Audio and hasn't for years
-Neither Dave nor myself is in any position to negotiate with anybody to demand anything. Dave has explained what he did above; for myself, if I want drive units from anybody, I buy them same as everybody else.
-No mass produced drive units, irrespective of manufacturer, will show 100% conformity to a generic spec. sheet across all units, with zero tolerance variation, nor any variation with different measurement method or conditions. The former isn't realistic, the latter not physically possible.
-All permanent magnet moving coil drive units will exhibit variations in their frequency response (especially at the low end) depending on the output impedance of the driving amplifier. That is because of the way it interacts with the drive unit's impedance. It's inherent to the type: there are no exceptions. How much they do so depends on the details of that amplifier and the specifics of the impedance curve, which results from the electromechanical design of the drive unit. Back in the day, most drive units were designed assuming use with a particular output impedance range, or amplifiers with variable output impedance were used to tailor the response. The majority of 'modern' drive units assume they will be used with a voltage source amplifier with extremely low output impedance (i.e. a very high damping factor). Some exceptions are the very low Q, high sensitivity types from Lowther etc., which like the old units, are designed assuming use with a high output Z amplifier which has the effect of artificially raising Qt and raising the LF output as the frequency response is modified by the amplifier following the impedance curve (since it has little in the way of electrical damping). Where relevant, the cabinet should be designed assuming these intended operating conditions.
Hope that helps a little.
-'Fenlon' does not own Mark Audio and hasn't for years
-Neither Dave nor myself is in any position to negotiate with anybody to demand anything. Dave has explained what he did above; for myself, if I want drive units from anybody, I buy them same as everybody else.
-No mass produced drive units, irrespective of manufacturer, will show 100% conformity to a generic spec. sheet across all units, with zero tolerance variation, nor any variation with different measurement method or conditions. The former isn't realistic, the latter not physically possible.
-All permanent magnet moving coil drive units will exhibit variations in their frequency response (especially at the low end) depending on the output impedance of the driving amplifier. That is because of the way it interacts with the drive unit's impedance. It's inherent to the type: there are no exceptions. How much they do so depends on the details of that amplifier and the specifics of the impedance curve, which results from the electromechanical design of the drive unit. Back in the day, most drive units were designed assuming use with a particular output impedance range, or amplifiers with variable output impedance were used to tailor the response. The majority of 'modern' drive units assume they will be used with a voltage source amplifier with extremely low output impedance (i.e. a very high damping factor). Some exceptions are the very low Q, high sensitivity types from Lowther etc., which like the old units, are designed assuming use with a high output Z amplifier which has the effect of artificially raising Qt and raising the LF output as the frequency response is modified by the amplifier following the impedance curve (since it has little in the way of electrical damping). Where relevant, the cabinet should be designed assuming these intended operating conditions.
Hope that helps a little.
If you want matched pairs, MAOP is the easiest way to go.
You can of course buy a few other pairs and try to make matched pairs. However, I think it will be more costly.
Maop are expensive, but that extra cost is in the long run the most economical if you have such demands on your sound.
You can of course buy a few other pairs and try to make matched pairs. However, I think it will be more costly.
Maop are expensive, but that extra cost is in the long run the most economical if you have such demands on your sound.
Last edited by a moderator:
In the middle of this discussion, sorry, I try to post some data!
This are 2 measurement I just took of the CHP-90 and the RS100-4.
Both speakers are placed on a desktop and I took the measurement as in my listening position, so approx 1 meter, with boxes distance same approx 1m.
Smoothing 1/48
I'm total newbie in measurement so not sure I've made everything 100% correct, but here my results...
Do you need more data? Is that little higher line of the green RS100 in the 200-300 region, what I like of the bass of those speakers?
What else your expert eye sees on this graph?
The spike of the CHP-90 in the 6k-7k region can be the "shouting/harsh"???
Also the 1k and 2.5k-3k might be disturbing me???
I've APO/PEACE so if you see and can advice some quick setting to try based on the FR of the CHP90... would be great!
Wondering why in the high (and ??bass??) region is so limited? the ??? MIC ??? Behringer ECM8000, calibrated with a generic file for it...
Streaming from PC through an AYIYMA T9 for the CHP-90 and an AIYIMA A07 (so great amp yes!) for the RS-100.
Soon I'll post also the W5-2143 and MAOP...
This are 2 measurement I just took of the CHP-90 and the RS100-4.
Both speakers are placed on a desktop and I took the measurement as in my listening position, so approx 1 meter, with boxes distance same approx 1m.
Smoothing 1/48
I'm total newbie in measurement so not sure I've made everything 100% correct, but here my results...
Do you need more data? Is that little higher line of the green RS100 in the 200-300 region, what I like of the bass of those speakers?
What else your expert eye sees on this graph?
The spike of the CHP-90 in the 6k-7k region can be the "shouting/harsh"???
Also the 1k and 2.5k-3k might be disturbing me???
I've APO/PEACE so if you see and can advice some quick setting to try based on the FR of the CHP90... would be great!
Wondering why in the high (and ??bass??) region is so limited? the ??? MIC ??? Behringer ECM8000, calibrated with a generic file for it...
Streaming from PC through an AYIYMA T9 for the CHP-90 and an AIYIMA A07 (so great amp yes!) for the RS-100.
Soon I'll post also the W5-2143 and MAOP...
Last edited:
Without knowing exactly how you took the measurements & what they are, bit difficult to say anything conclusive. To a point (to a point) different things will be heard differently by different people -where some like or require a peak at a given point for e.g., others will prefer a flat or null response there.
If I read correctly, the above is for a stereo pair at the listening position? That's good from the POV that it's representative of what you're listening to, but for assessing the behaviour of a given design, it really needs to be single-channel. For relatively compact widebands, 1m is fine although a little closer is OK -I tend to follow Jeff Bagby's advice on this: you're effectively in the farfield when 'you' (as in the microphone) are positioned at ~3 - 5x the radiating diameter away from the driver. So for a CHP-90, with an Sd of 85dm^2, its diameter is 104mm, so your mic. is nominally farfield from about 31.2cm away. That's a bit close but 52cm (20in) should be fine. Caveat to that the distance should also be at least twice the baffle width, so if you've a wide-baffle design, by necessity the mic. will need to be further out to accurately show step-loss. Using those two is a good guide though as you're avoiding some of the issues with reflections that can come in with 1m measures & with single-driver designs you don't need to worry about integration of different units of course.
As general points which you already know, but I'll put here 'just because': make sure the microphone is calibrated properly -the generics are OK for the ECM but make sure its loaded up (individual is better, but the generic is decent), and take some time over your mic. alignment: it can make quite a difference. Select your gating window based on 1st reflection, and I'd probably use a maximum vertical scale of 100dB; a bit smaller is good. Zaph used to use 90dB; not many moaned about his measurements, so it's a good 'default' to use IMO.
If I read correctly, the above is for a stereo pair at the listening position? That's good from the POV that it's representative of what you're listening to, but for assessing the behaviour of a given design, it really needs to be single-channel. For relatively compact widebands, 1m is fine although a little closer is OK -I tend to follow Jeff Bagby's advice on this: you're effectively in the farfield when 'you' (as in the microphone) are positioned at ~3 - 5x the radiating diameter away from the driver. So for a CHP-90, with an Sd of 85dm^2, its diameter is 104mm, so your mic. is nominally farfield from about 31.2cm away. That's a bit close but 52cm (20in) should be fine. Caveat to that the distance should also be at least twice the baffle width, so if you've a wide-baffle design, by necessity the mic. will need to be further out to accurately show step-loss. Using those two is a good guide though as you're avoiding some of the issues with reflections that can come in with 1m measures & with single-driver designs you don't need to worry about integration of different units of course.
As general points which you already know, but I'll put here 'just because': make sure the microphone is calibrated properly -the generics are OK for the ECM but make sure its loaded up (individual is better, but the generic is decent), and take some time over your mic. alignment: it can make quite a difference. Select your gating window based on 1st reflection, and I'd probably use a maximum vertical scale of 100dB; a bit smaller is good. Zaph used to use 90dB; not many moaned about his measurements, so it's a good 'default' to use IMO.
Hi Scott, thx for answer! Yes, stereo at the listening position!
Finally I don't really need to measure the driver "itself" I just want it to sound good with some good correction with APO/MINIDSP/Physical XO/ I'll decide later.
Here the CHP only for better understanding... for what it can help!
"Select your gating window based on 1st reflection" really don't know what is/how to do it... I'll search and also the vertical scale of 100db not fully getting it.
The main reason of my measurement are to understand, if possible! through the freq response what and why I like so much the sound of the RS100 and not the CHP90/W5-2143! Let's continue to play!!!
Finally I don't really need to measure the driver "itself" I just want it to sound good with some good correction with APO/MINIDSP/Physical XO/ I'll decide later.
Here the CHP only for better understanding... for what it can help!
"Select your gating window based on 1st reflection" really don't know what is/how to do it... I'll search
and also the vertical scale of 100db not fully getting it.The main reason of my measurement are to understand, if possible! through the freq response what and why I like so much the sound of the RS100 and not the CHP90/W5-2143! Let's continue to play!!!
For a different microphone, but should help a bit re REW setup: https://www.minidsp.com/applications/acoustic-measurements/loudspeaker-measurements
Vertical scale i.e. SPL. Avoid having a range of more than 100dB (e.g. 0dB - 100dB, 30dB - 130dB & so on). 90dB range is better. For this purpose there's not a whole lot of value in going coarser, it just ends up looking a complete mess & it's often difficult to home in on trending because it's been so heavily inflated.
Neither CHP-90 nor the RS100 are BW limited to the ~ 180Hz - 6.5KHz range so unless you have some kind of crossover in there, there's something amiss with the measurement setup.
Vertical scale i.e. SPL. Avoid having a range of more than 100dB (e.g. 0dB - 100dB, 30dB - 130dB & so on). 90dB range is better. For this purpose there's not a whole lot of value in going coarser, it just ends up looking a complete mess & it's often difficult to home in on trending because it's been so heavily inflated.
Neither CHP-90 nor the RS100 are BW limited to the ~ 180Hz - 6.5KHz range so unless you have some kind of crossover in there, there's something amiss with the measurement setup.
OK yes thx! There was something weird in fact on the scale limited from 180hz to 6.5khz!
I found it was some really stupid "noise reduction" activated in the audio card settings of the mic! Now solved and tested, i have the whole frequencies so I'll have to post them again... ://
I found it was some really stupid "noise reduction" activated in the audio card settings of the mic! Now solved and tested, i have the whole frequencies so I'll have to post them again... ://
Yes, true!!! I understand really well! I was already thinking that maybe the mic was in same way defective... but then made full check in REW params and then went to the Audio Card Interface and thought that the "mic noise reduction" button was not really good! and voila, solved!
Post new measurement soon! Later after the match or tomorrow...
Post new measurement soon! Later after the match or tomorrow...
Not true. A friend bought a pair of Maop 11 from Soundimports and although stipulated as matched, they weren't according to him. I said to sent them back but he said its not worth it.