Hi,
The ACE bass system is often mentioned as a servo system but its not. A servo corrects the motion based non linearity whereas the ACE bass system puts the resistance of the voice coil in the feedback circuit thus minimizing the effect of the impedance of the voice coil.
So, what are the real tangible advantages of the ACE bass system and at what bandwidth?
Thanks and Regards,
Goldy
The ACE bass system is often mentioned as a servo system but its not. A servo corrects the motion based non linearity whereas the ACE bass system puts the resistance of the voice coil in the feedback circuit thus minimizing the effect of the impedance of the voice coil.
So, what are the real tangible advantages of the ACE bass system and at what bandwidth?
Thanks and Regards,
Goldy
Amplifier distortion is far too low to be of any problem. It has its own loop that does that already. Trying to correct that by driver current is too complicated a solution and probably of questionable performance in a real system.
Driver distortion cannot be sensed with driver current. I mean it can be sensed but it will not be of good quality. The current is dependent on BL linearity which itself is not stable across excursion. This is why some designs use a separate coil that is more linear than the voice coil, no?
US Patent for Loudspeaker lower bass response using negative resistance and impedance loading Patent (Patent # 4,118,600 issued October 3, 1978) - Justia Patents Search :
The invention is intended to provide an extended frequency range and lower distortion in the bass register in hi-fi-reproduction.
Patent number: 4118600
Abstract: An apparatus and method for improving the bass response characteristics of an electrodynamic loudspeaker is described. The loudspeaker normally exhibits actual mechanical parameters such as damping, compliance, and mass which normally determine the bass response and lower cut-off frequency of the loudspeaker. The method and apparatus of the present invention cause the loudspeaker to exhibit apparent mechanical parameters which differ from the actual mechanical parameters to substantially change the effect of the actual mechanical parameters on the bass response. The apparatus in a preferred embodiment includes an electrical network through which electrical energy corresponding to the sound to be reproduced is applied to the voice-coil of the loudspeaker. The electrical network has an effective output impedance including a negative impedance in series with a plurality of impedances connected in parallel.
Type: Grant
Filed: March 23, 1977
Date of Patent: October 3, 1978
Inventor: Karl Erik Stahl
The invention is intended to provide an extended frequency range and lower distortion in the bass register in hi-fi-reproduction.
Patent number: 4118600
Abstract: An apparatus and method for improving the bass response characteristics of an electrodynamic loudspeaker is described. The loudspeaker normally exhibits actual mechanical parameters such as damping, compliance, and mass which normally determine the bass response and lower cut-off frequency of the loudspeaker. The method and apparatus of the present invention cause the loudspeaker to exhibit apparent mechanical parameters which differ from the actual mechanical parameters to substantially change the effect of the actual mechanical parameters on the bass response. The apparatus in a preferred embodiment includes an electrical network through which electrical energy corresponding to the sound to be reproduced is applied to the voice-coil of the loudspeaker. The electrical network has an effective output impedance including a negative impedance in series with a plurality of impedances connected in parallel.
Type: Grant
Filed: March 23, 1977
Date of Patent: October 3, 1978
Inventor: Karl Erik Stahl
TNT,
"The method and apparatus of the present invention cause the loudspeaker to exhibit apparent mechanical parameters which differ from the actual mechanical parameters to substantially change the effect of the actual mechanical parameters on the bass response."
how is that an advantage, how does it reduce distortion, make excursion more linear etc?
"The method and apparatus of the present invention cause the loudspeaker to exhibit apparent mechanical parameters which differ from the actual mechanical parameters to substantially change the effect of the actual mechanical parameters on the bass response."
how is that an advantage, how does it reduce distortion, make excursion more linear etc?
I think you need to read the patent. Being in control of , and able to modify, the mechanical properties of used driver is an advantage. Negative output impedance may in some certian situation be able to reduce distorsion as I undertand it. The configuration of using 2 drivers where one is turned in the opposite direction but phase reversed so to play in unison with the other, phases out 2nd order distorsion. Its the total system taking size, cost etc into consideration that makes up for the advantage in the end - not necessarily the top performance of any parameter....
//
//
Mr. Erath designed his circuit when transistors were still newish and loudspeakers were crappy.
A good bit of time later and I am using it on Hypex class-d amps and Emenince drivers.
Someone recently said and I completely agree:
If you hear a sub that is distortion.
This seems to be a topic that never goes away. ACE Bass and Erath's mystery system are simply current-sensing motional feedback.
The quote TNT provides from the patent is a beautiful example of the double-talk lawyers create to hide the fact that the requested patent approval has nothing new in it.
Despite wonderfulaudio's armchair musings, current sensing works pretty good in sensing cone motion, albeit a few steps of relationship from the sound reaching your chair (a shortcoming shared by all motional feedback systems including accelerometers)... or so 60 years of mostly DIY experimentation can attest.
The apparent impedance of the VC depends on the copper and on the back-EMF. If the back-EMF doesn't match the input signal due to the cone misbehavin', you get an error message to the amp. The amp does something weird: it acts like the negative of the voice coil impedance. So when the VC goes "zig", the amp goes "zag" and the misbehaviour is reduced.
Before signing off here I must add: if your sub isn't in a feedback loop, you aren't really trying to have the best bass.
B.
The quote TNT provides from the patent is a beautiful example of the double-talk lawyers create to hide the fact that the requested patent approval has nothing new in it.
Despite wonderfulaudio's armchair musings, current sensing works pretty good in sensing cone motion, albeit a few steps of relationship from the sound reaching your chair (a shortcoming shared by all motional feedback systems including accelerometers)... or so 60 years of mostly DIY experimentation can attest.
The apparent impedance of the VC depends on the copper and on the back-EMF. If the back-EMF doesn't match the input signal due to the cone misbehavin', you get an error message to the amp. The amp does something weird: it acts like the negative of the voice coil impedance. So when the VC goes "zig", the amp goes "zag" and the misbehaviour is reduced.
Before signing off here I must add: if your sub isn't in a feedback loop, you aren't really trying to have the best bass.
B.
Last edited:
This RMS acoustics whitepaper dismisses ACE bass kind of current drive of subwoofer. The results are in page 21.
https://rmsacoustics.nl/papers/whitepapervelocityfeedback.pdf
Can somebody explain why such minor improvements in their findings, at some frequencies the results are worse than without current drive? Pls note, my understanding is that negative impedance and current drive are same thing.
https://rmsacoustics.nl/papers/whitepapervelocityfeedback.pdf
Can somebody explain why such minor improvements in their findings, at some frequencies the results are worse than without current drive? Pls note, my understanding is that negative impedance and current drive are same thing.
Have you tried it?
After reading the RMS document they use the same principal Mr. Erath used. Only ~50 years after he designed it..
I just rebuilt my units he made with more modern technology (I stopped short of SMD components)
and am amazed at how little you actually HEAR good bass. Now most of what I hear are upper harmonics through the satellites while the depth and fullness are there but totally silent since there is next to no distortion (mechanical or electronic)
Mr.Earth designed this in the very early transistor days and when speakers were still crude. It helped clean up both the amp and the woofer distortion.
My units use Hypex amps and Emenince drivers. More than likely my unit ADDs distortion to the Hypex but still cleans up the Emenince for a net positive.
Mr. Erath was ahead of his time... Turns out it was by about 50 years.
You've been touting Mr. Erath's accomplishment for years in this forum. He was pretty secretive about the details. It now seems time for you to post (or re-post) the circuit you are using or at least the general typology and guidance about circuit action.
Deserves it own thread. Deserves more than your repeated inflated claims in the absence of objective measurements.
I agree with you that "Mr. Erath" was 50 years ahead of our time today. We should catch up with motional feedback. But he was 25 years behind the research Werner (RCA Labs) published in the '50's.
Thanks.
B.
Deserves it own thread. Deserves more than your repeated inflated claims in the absence of objective measurements.
I agree with you that "Mr. Erath" was 50 years ahead of our time today. We should catch up with motional feedback. But he was 25 years behind the research Werner (RCA Labs) published in the '50's.
Thanks.
B.
Last edited:
Once again, not my place to give out someone else's IP. His patent is expired and he has passed. Very little research / effort and you can build his units.
I'm sure RMS doesn't want to share their circuit, but the block diagrams and functions look familiar.
I will be trying to build one of theirs and see how it compares.
I heard the Grimm speakers in Munich several years ago and really liked them but it was under show conditions so no critical listening.
With today's technology and ease of access to manufacturers I don't understand why people don't experiment more.. I just had 10 prototype pcbs shipped to my door for $100.
I havent tried it but will do so soon hence so many questions trying to understand it.
The project seems to be nicely done if entirely conventional for that kind of voice coil feedback. His diagrams are good blueprints for a DIY builder. But as you say, the results are unimpressive.
With research (esp "published" by a company without peer review) it is easy to get null results and you only get a "hit" when all research bits are working right.
I can't point to a specific failing here - maybe others can. He seems to have run the speaker very loud and VC feedback goes off the rails if the drivers are over-stressed (including over-heated).
The curves that interests me as a builder are the performance of the electrical output. What does the FR look like and is the amp actually pumping out corrective distortion? He does have an FR but it looks more conceptual than empirical and no THD breakdown.
Rice-Kellogg sub speakers with the resonance inside the passband is kind of dumb and motional feedback is a crucial to making it better. But that's a struggle and very hard to have much correction in systems sold to the public without risking instability.
But 'most anyone who has tried it - commercial or DIY - loves it.
B.
Bentoronto,
Is this valid way of making a prototype?
1) Take current feedback from 0.1 ohm resistor in series with driver load.
2) Feed it back to the amp.
3) This will yield a system with a non flat frequency response. Now, Eq it to flat.
Is that sufficient to get a working prototype?
Thanks and Regards,
WA
Is this valid way of making a prototype?
1) Take current feedback from 0.1 ohm resistor in series with driver load.
2) Feed it back to the amp.
3) This will yield a system with a non flat frequency response. Now, Eq it to flat.
Is that sufficient to get a working prototype?
Thanks and Regards,
WA
Yes, good concept. Take care to use lower power amp than driver can handle because many a slip during prototype testing. DAMHIK.
Even bigger resistor may give a bigger feedback signal to work with.
There are a few corrections needed. As I understand it, the main correction is that current flow (AKA back-EMF influence) reflects velocity but driver sound output relates to acceleration (I might have that backward). So the feedback signal would unduly reduce the speaker output **unless** you boosted amp output by 6 dB/8ave (AKA one capacitor).
Few people realize that cone drivers are artfully fashioned to produce uniform sound output FR with a constant voltage. Doesn't come naturally to the raw driver (and not when you use cone motion feedback either).
Main point is: feedback catches all kinds of cone misbehaviour but needs corrective EQ.
B.
Even bigger resistor may give a bigger feedback signal to work with.
There are a few corrections needed. As I understand it, the main correction is that current flow (AKA back-EMF influence) reflects velocity but driver sound output relates to acceleration (I might have that backward). So the feedback signal would unduly reduce the speaker output **unless** you boosted amp output by 6 dB/8ave (AKA one capacitor).
Few people realize that cone drivers are artfully fashioned to produce uniform sound output FR with a constant voltage. Doesn't come naturally to the raw driver (and not when you use cone motion feedback either).
Main point is: feedback catches all kinds of cone misbehaviour but needs corrective EQ.
B.
Last edited:
Does it mean that a driver with greater back-emf (ie greater BL, higher sensitivity) is better suited for current drive implementation?
People tend to think in terms of all or nothing. That is not necessarily the correct approach in this scenario. 100% feedback from the current sense resistor is going to turn the system into a current drive, which has it's own drawbacks.
Instead, the signal from the current sense (series resistor or the resistance of the voice coil itself) should be used as an additional feedback source which, in conjunction with the drive signal would produce an error correction signal that would be added to the drive signal.
The overall drive to the driver remains predominantly a voltage source, just slightly tweaked to compensate for current related errors.
Instead, the signal from the current sense (series resistor or the resistance of the voice coil itself) should be used as an additional feedback source which, in conjunction with the drive signal would produce an error correction signal that would be added to the drive signal.
The overall drive to the driver remains predominantly a voltage source, just slightly tweaked to compensate for current related errors.