I'm finding it hard to choose between ignoring this thread entirely and helping.
It would be good if you got a copy of Acoustics-Beranek and Acoustical Engineering-Olson from a university library and tried to understand their explanations of moving magnet.
A low Qms is indicative of low Bl. A high Qms shows high Bl^2 / Rdc. Low Qms & high Bl are not strictly exclusive but they are in practice
You really need to understand why this is so to design what you want .. or more important ... define what you want.
You also need to understand the reasons for the 'poor' performance of the 2 examples you quoted. HINT: its nothing to do with their Voltage drive.
But I'm quite prepared to be surprised so don't let me stop you experimenting. 🙂
I've been experimenting successfully with drivers having Qms of about 2, and in current drive that is then effectively the same as Qts. I've found this to be fine for damping in practice, even preferable to Qts of about 0.5 for voltage drive. Timing seems earlier, and no nasties.
I only have subjective views so far, not measured key listening params, but timing and distortion appear preferable with current drive in my tests, well worth pursuing I think.
I only have subjective views so far, not measured key listening params, but timing and distortion appear preferable with current drive in my tests, well worth pursuing I think.
If this is my fault, then I am sorry. I don't mean to offend anyone. I just write stuff based on my own experience and what ever I find interesting in the ongoing development mainly in professional sound-reproduction.
I don't understand what "two examples" you mean?
The experimentation is 90% of the fun. 🙂
Cheers,
Johannes
If I remember correctly there are diminishing BW results if you increase the flux density to high.
Not even that. Low Qes is caused by high Bl product. Qms resides in the mechanical realm and is not influenced by electromagnetic parameters.
Last edited:
No.
I said "A low Qms is indicative of low Bl. A high Qms shows high Bl^2 / Rdc. Low Qms & high Bl are not strictly exclusive but they are in practice"
It's not offensive.
But crazy ideas need to be looked at properly. Sometimes, crazy is so crazy that it will take a LOT of work to make anything sensible. 🙂
But talking crazy sometimes gives good ideas. I'm mulling over a practical speaker designed for current drive.
It would have low Qms as well as low Qes. I'm not sure it take full advantage of current drive but it would be an easy way to tune a Current Drive system for flat LF response.
luckythedog is on the right trail 😀
Have you read Nelson Pass's article on his website?
The servo motor sub and the 50dB/W experiment.
If you speak in quality terms, I agree. High efficiency speakers require both high Bl and high Qms. But in engineering terms it doesn't make sense, see my earlier post.
I know a high Bl is more then indicative of a low Qes. A low Qms is a low mechanical resonance propensity of the moving mass and suspension. I can't see how Bl has anything to to with that.
Modern drivers like B&C 12NW100 has both a very high Bl (and of course a low Qes) and a low Qms. A couple of years ago most high Bl low Qes drivers had a high Qms, but this leads to a very pronounced impedance peak at Fs, and much more back-EMF from the voicecoil. This is much harder for a normal voltage amp to drive, as the driver is much more reactive. Qms should always be as low as possible to linearise the impedance and counteract back-emf. It is always better to have a well controlled mechanical system that is not dependent on the the damping of the amp, then a highly reactive driver always taxing the amp very hard for control. More so in a bassreflex box with its very pronounced rapid changes of operating conditions for the driver and a very high Q port resonance that will store a lot of energy and smear it out in the time domain.
I can't find any other area of mechanical devices that are inherently mechanically resonant and controlled and damped only through the use of electric devices. If a high Qms is good why don't we see this in industrial welding robots, the rudder and wing flaps on fighter-jets, car suspension etc???
The increase in Rms is easily compensated for by increasing the Bl if spl/w/m is important.
It is always better to design inherently linear devices that does not rely on outer help to behave as intended.
Cheers,
Johannes
I may be wrong but I don't think any of what you just said will affect the back emf. That is strictly a function of voicecoil velocity or acceleration from what I remember.
High Qms implies high Rms (mechanical resistance), which has often been reported as a good thing for low level subjective dynamic.
Less non linear mechanical phenomenons maybe?
Less non linear mechanical phenomenons maybe?
Replacing the magnet with a non-magnetic structure should'nt change the Qms.
The "control" of the amp on the driver is only the behavior of the closed circuit. It can be predicted and observed by looking the impedance of their whole circuit.
Which poor amp does have difficulties to control a loudspeaker and its devilish back-EMF ? Maybe only a very ancient one with stability problems of negative feedback in low frequencies.
The driver back-EMF at the resonance implies a high impedance seen by the amp. No modern amp with voltage output ever complains with high impedance loads.
What can happen is the problem of instability of the driver, as mentioned by some authors and recently reminded in this file by Kgrlee.
Current drive can take care of it, I think it is its main virtue, but driver instability does not seem to be a major default in current production nowadays.
Forr is correct as to the effect of high Qms on back emf. No decent amp should gave any problem with it, but tube amps witb highish Zout will show deviations from a flat fr.
Don't just accept it from me but thinkh about it yourself. Mechanical friction in a driver is bad, bad, bad. Firstly, it lowers efficiency. Secondly, mechanical friction is never linear, so it introduces distortion. And thirdly, mechanical friction typically produces noise, so it introduces rub and buzz sounds. Therefore, you want drivers with an as high Qms as possible, greater than 10 is desirable. Unfortunately, loudspeaker manufacturers typically keep these drivers for there own final products.
The only upside of low Qms is that it lowers Fs, but it is the worst way to accomplish it.
Don't just accept it from me but thinkh about it yourself. Mechanical friction in a driver is bad, bad, bad. Firstly, it lowers efficiency. Secondly, mechanical friction is never linear, so it introduces distortion. And thirdly, mechanical friction typically produces noise, so it introduces rub and buzz sounds. Therefore, you want drivers with an as high Qms as possible, greater than 10 is desirable. Unfortunately, loudspeaker manufacturers typically keep these drivers for there own final products.
The only upside of low Qms is that it lowers Fs, but it is the worst way to accomplish it.
Well friction has several forms, some are therapeutic in the right dose - such as viscous drag, or in the case of speaker drivers losses in the elastomers which form the suspension.
Similar intentional damping crops up in phono cartridge mechanics, where there is a trade off between force required to move the suspension versus damping provided - in that case Q ends up typically between 2 and 3, which in practice provides for perfectly adequate stability and damping of a resonance typically near 10Hz. In phono cartridges, mechanical damping is all there is, so Qts is effectively between 2 and 3 typically, without issue.
So there is a successful audio precedent for Qts near 2, as would be the case for a high Zo current drive amplifier and a well chosen standard speaker driver.
As I say, seems to work well in initial tests.
Similar intentional damping crops up in phono cartridge mechanics, where there is a trade off between force required to move the suspension versus damping provided - in that case Q ends up typically between 2 and 3, which in practice provides for perfectly adequate stability and damping of a resonance typically near 10Hz. In phono cartridges, mechanical damping is all there is, so Qts is effectively between 2 and 3 typically, without issue.
So there is a successful audio precedent for Qts near 2, as would be the case for a high Zo current drive amplifier and a well chosen standard speaker driver.
As I say, seems to work well in initial tests.
Qts of 2 is indicative of a crappy driver, no precedents known. All friction in drivers is bad, full stop. The only allowable way to lower Qts on purpose is electrical.
It is actually the opposite. A high Rms indicates high resistive loss in the suspension of the moving mass = low propensity to resonate (good damping) = low Qms.
My B&C 10PS26 does not have any "rub and Buzz" sounds, despite a low Qms of 2,9.
Scan Speak Illuminator 18WU874T00 (Qms 3,31) and Accuton C173-6-090 (Qms 3,75) does not either?!!! Why would Scan Speak and Accuton design very expensive and acknowledged drivers for high end equipment if low values of Qms was soo devastating for soundquality and efficiency?
Please tell my why modern high output drivers like the B&C 18IPAL, 12NW100, 18SOUND 18NLW9401, RCF LF15N451 among others have a low Qms and high Rms? Please compare the latest generation high output bass drivers with comparable drivers from 10 years ago, and you will see a distinct drop in Qms. Is this some kind of regression of performance that several manufacturers of drivers have agreed to suddenly make?
I can't speak for industrial robots but I think you will find 'enhanced' stability controls on the new fighters have very high mechanical Qs with the required damping obtained by 'electromagnetic' and other electronic means including DSP.
With cars, the mechanical damping has to be sufficient when the 'enhancement' is switched off.
The new fangled 'electrical' damping is usually much more consistent than any mechanical damping and this holds in speakers too.
Have you got links to the modern units which have low Qms? There are ways to achieve this consistently but they are expensive. You don't want friction but viscous damping.
Actually in low Qms systems, the requirement is usually to have reduced Bl^2 / Rdc. This has the units of Ns/m ie mechanical ohms. The high mechanical resistance means less electromagnetic resistance is needed.
I think some revision of Thiele/Small 101 might be useful here. Or perhaps just playing with something like Unibox. 🙂
Last edited:
Found B&C 10ps26
It is certainly $$$ but the Qms isn't all that low for a 'high quality' unit. 🙂 Neither are your Scan Speak or Accuton examples. In these 3 cases, the mechanical damping is about (or less than) 10% of the electromagnetic damping. They would be considered 'average' even in the 1970s.
Last edited: