Thank you very much for clarifying GM. And thank you for everyone else for providing good information. It makes sense, and the sort of thing I was afraid of.
From what I have learned in this post I think I have to rephrase my question and requirements:
Which midrange driver suitable for home hifi is known for behaving well when pushed to its power limits but well within xmax? Lets say the minimum requirements are 100db/1m dB SPL in the range 350-2500Hz crossed actively with LR4 driven by 100W. Preferably 5" or less if doable.
From what I have learned in this post I think I have to rephrase my question and requirements:
Which midrange driver suitable for home hifi is known for behaving well when pushed to its power limits but well within xmax? Lets say the minimum requirements are 100db/1m dB SPL in the range 350-2500Hz crossed actively with LR4 driven by 100W. Preferably 5" or less if doable.
Well, partly misunderstood in that when you added the 'HIFI' tracking, this really narrowed up your point source choices since prosound historically use compression horns, [relatively] high Vas woofers to track all of the signal at low power.
For high SQ mids, the pioneers concluded that a wide range driver is ideally required to reproduce the [250-2500 Hz] analog telephone BW at high speech intelligibility, settling on a 8" three way [tri-ax], ~70-13 kHz 'single' driver [W.E. 755].
Today, a 6.5" co-ax seems the way to go or 8" if you don't like your options.
GM
Still looks like a job for the BMS 5S117.
Here is some guy who tested it as well as a few others where it ended up top of the bunch alongside a boutique model at twice the price I've never heard of: Robsan-DIY | Mid-Bass Tests
BMS 5S117 is not bad if someone likes higher 3rd than 2nd harmonic distortions in all of the pass-band with almost 1% at 85dB at the middle of the pass-band.
Hifi driver makers try to follow the physiologically better allocation where a higher order harmonic distortion is lower in level. One of the pro sound manufacturer that makes products with that trend with their mid-woofers is B&C. Another good example from pro drivers is the aforementioned 6ND430 (favored by many for home use).
Hifi driver manufacturers like Scan-Speak, SB Acoustics are all try to make drivers like that in mind. Check the linked tests at post #20. In the audioxpress bench tests the red line is the 2nd harmonic.
Might be just nit-picking, what do you guys think?
Hifi driver makers try to follow the physiologically better allocation where a higher order harmonic distortion is lower in level. One of the pro sound manufacturer that makes products with that trend with their mid-woofers is B&C. Another good example from pro drivers is the aforementioned 6ND430 (favored by many for home use).
Hifi driver manufacturers like Scan-Speak, SB Acoustics are all try to make drivers like that in mind. Check the linked tests at post #20. In the audioxpress bench tests the red line is the 2nd harmonic.
Might be just nit-picking, what do you guys think?
If you looked at the graph on Robsan-DIY you seem to have mixed up the colours for 4th and 2nd order since 2nd is clearly higher than 3rd.
You also seem to have failed at noticing the log character of the grading and THD never gets anywhere near the levels you stated.
He also tested a couple of SB Acoustic drivers whose THD is about 3x that of the BMS.
You also seem to have failed at noticing the log character of the grading and THD never gets anywhere near the levels you stated.
He also tested a couple of SB Acoustic drivers whose THD is about 3x that of the BMS.
Last edited:
Yes the 5S117 have very low distortions even at very high SPL (according to manufacturer datasheet) but the linked test shows higher D3 (red) than D2 (blue). Might not a problem and excellent result overall.
I don't know if this has been mentioned already, but what's the largest driver size you're willing to accept for this mid application?
If my current target of 100dBspl/1m at 350-2500Hz LR4 cannot be reliably met with a single 4" or 5", i can go as high as 6". Even 8" if there is a very good reason i should go for that, but i would much prefer a size that can cross over to a dome tweeter without a waveguide.
According to the specs, both SB12MNRX2 and 15M/4524G should theoretically be able to do it. But will they sound good doing it? Will they handle transient peaks some 10db above the target?
According to the specs, both SB12MNRX2 and 15M/4524G should theoretically be able to do it. But will they sound good doing it? Will they handle transient peaks some 10db above the target?
Last edited:
Yes and no per my previous response to you. Drivers are simple solenoids, so the motor controls what it can and cannot track signal wise. The suspension only needs to be stiff enough to keep the VC centered in the gap, so any stiffer for the needs of the app is a drag on the motor.
The B&C 5MDN38-8 is so stiff you only need look at its measured rising response all the way up to its upper mass corner to know it's relatively weak motor Vs tiny Vas [compliance] can't maintain tonal balance at low power, much less accurately tracking a fractional watt signal, so in short, sounds like a seriously over-damped/'sterile', etc., alignment compared to a decent quality 'full-range' driver's, or even a high Vas [mid] bass woofer's 'inner detail' response and of course any horn/WG system worthy of the name.
GM
100 dB @ 1 m is very easy. Those drivers can readily handle it. If you need more, the SB17NBAC35-8 would probably be able to do 110 dB cleanly (its measurements in volumes increasing up to 100 dB @ 1 M show no drastic change in higher order distortion).
Having done some measurements on this very phenomenon, I couldn't find any evidence to support it.
I set off running sweeps at a moderate in-room level, using a Kef HTS3001SE and a Beyer MC930 microphone (chosen for its low electrical noise and high output) positioned up close.
After each sweep, I reduced the speaker drive level by 6dB. The first sweep might've been around 50dB at my ears. The penultimate sweep was briefly audible as it passed through the sensitive kHz range, and the final sweep was inaudible entirely.
As the graph shows, even though I couldn't hear it, the frequency response was essentially unmodified.
Chris
Attachments
Chris I'm not understanding your graph ? The bass is rolling off higher and higher as the spl drops, and the top trace is flat whereas all the others have roll off in the treble ?
edit: unless the top 3 traces are on axis / off axis measurements and the rest are measured at the same position maybe ?
Rob.
edit: unless the top 3 traces are on axis / off axis measurements and the rest are measured at the same position maybe ?
Rob.
Last edited:
Hi Chris, i still don't get what you're showing either.
Are you saying the top two traces are bogus?
If so, i'd toss them to illustrate your (cool) analysis
Are you saying the top two traces are bogus?
If so, i'd toss them to illustrate your (cool) analysis
I'm saying the top two traces clipped the mic input for part of the frequency range. The unclipped sections are fine, though. With the directional mic close-up, there's some boost in the LF as expected.
If the top trace was around 50dBSPL nominal, with 7x 6dB steps down, the bottom trace was at 8dBSPL, which is probably why I couldn't hear it acoustically.
Next time, I'll calibrate the SPLs for 1m distance.
I just wanted to show that you could go from quiet to very quiet and (apart from the input clipping at the highest SPLs putting a ceiling on things) the frequency response stays consistent.
Chris
If the top trace was around 50dBSPL nominal, with 7x 6dB steps down, the bottom trace was at 8dBSPL, which is probably why I couldn't hear it acoustically.
Next time, I'll calibrate the SPLs for 1m distance.
I just wanted to show that you could go from quiet to very quiet and (apart from the input clipping at the highest SPLs putting a ceiling on things) the frequency response stays consistent.
Chris
8dB ! That's massively lower than any noise floor I can achieve out here in the Essex sticks. Are you in a studio ?
As an aside, I thought the top (red) trace looked the best measurement which does go to show..... 😀
Cheers,
Rob.
edit: clipping your mic at 50dB doesn't sound right to me at all ? that's 30dB below 'loud talking' according to the charts (from memory? )
Last edited:
