ACE Bass DIY

The ACE-BASS circuit uses feedback around the driver, but it's not "motional feedback" per se. Although in principle the circuit can produce benefits in terms of extension of the bass AND distortion reduction, the latter is not so great in reality because of voice coil heating. There is no magic to the bass extension - to do this the circuit is boosting the level similar to other equalization circuits e.g. Linkwitz Transform, however, the ACE-BASS does have the ability to work with vented boxes.

If you want to play around with it, just make the simplified single op-amp version for a driver in a closed box. You can find it in the literature (patent or paper, I forget which). See if you like it. Make sure to pay attention to the input/output polarity of the amplifier you use because of the feedback!
 
The ACE-BASS circuit uses feedback around the driver, but it's not "motional feedback" per se

Semantics again it might be, but I don't agree... :)

The ACE-BASS amplifier relies on coil current sensing for feedback: It is motional feedback therefore in all cases except where the driver motion is perfectly blocked. Stahl's original AES paper rightly distinguishes it from velocity sensing feedback, but what distinguishes ACE-BASS from the prior art (concerning amplifier negative output impedance) is an additional coil current feedback path that obviates the need for any "pre-calibrated" filtering. And "per se", that motional impedance modifier I reckon qualifies it as motional feedback?

But aside from the semantics, (as we have discussed before) voice coil heating issues can be circumvented if a self-balancing bridge is adopted, such as the elegant solution described by Birt in his AES papers and patents (which also provides an exceptional means for adding genuine voice coil rms power limiting too) - and in which amplifier output impedance can be modified as in the ACE-BASS system or as otherwise desired.
 
Member
Joined 2008
Paid Member
Semantics again it might be....but what distinguishes ACE-BASS from the prior art (concerning amplifier negative output impedance) is an additional coil current feedback path that obviates ...
I'll just quote one bit from soundbloke's very worthwhile post.

"Prior art" - as we all know - is patent-office language. Folks make all kinds of semantic somersaults in order to falsely claim something is new and hence patentable. That's the point, as soundbloke says.

BTW, the bridge approach is an easily implemented trick that also compensates for VC inductance.

B.
 
BTW, the bridge approach is an easily implemented trick that also compensates for VC inductance.

Indeed, but where coil-sensed motional feedback is employed, we also need be concerned with how pure an inductance the driver provides. Non-linearities and effectively parallel semi-inductances (with their own non-linearities) are pitfalls that await those unaware of the prior art :)
 
I have always wondered why, in the balanced bridge implementation, why the driver is being balanced by a resistor. Why not make the balancing circuit more similar to the equivalent circuit of a driver that has the desired F and Q. This could more closely follow the impedance at higher frequencies. For an example, the attached 1958 paper by Werner that uses a balanced bridge has several options for the balancing element, some of which incorporate a rising impedance using an inductor or a bypassed capacitor. Even back then he got pretty good results.

So why is this not done?
 

Attachments

  • Werner1958 - Loudspeakers and negative impedances.pdf
    884.3 KB · Views: 67
Last edited:
Member
Joined 2008
Paid Member
I have always wondered why, in the balanced bridge implementation, why the driver is being balanced by a resistor. Why not make the balancing circuit more similar to the equivalent circuit of a driver that has the desired F and Q. This could more closely follow the impedance at higher frequencies.....

So why is this not done?

Fair analysis.

And in Werner's paper, which you link, I think he does so, just using a capacitor in the correct leg. In practice, inductance (or semi-inductance) isn't a big factor and esp for woofers (which is the only band where MFB seems to be applied.... not sure why).

BTW, everybody interested in subs should read the Werner article because it says (and measures) just about all the main topics you'd want to know about MFB and that mysterious unit, the amp with negative output impedance (no kidding).

I think Werner was dreaming about full-range speakers and - like Philips - trying to help RCA-labs (Olson) make a silk purse out of a mere sow's ear using feedback. For folks at DIYaudio, we want to take great woofer drivers and make them rattle our dentures with dramatic transient performance.

B.
 
Last edited:
Fair analysis.

And in Werner's paper, which you link, I think he does so, just using a capacitor in the correct leg. In practice, inductance (or semi-inductance) isn't a big factor and esp for woofers (which is the only band where MFB seems to be applied.... not sure why).

I think Werner was dreaming about full-range speakers and - like Philips - trying to make a silk purse out of a mere sow's ear using feedback. For folks at DIYaudio, we want to take great woofer drivers and make them rattle our dentures with dramatic transient performance.

B.

See on second page of article, Figure 2. He writes an equation for the impedances in the bridge, like Z1/Z2 = Zvc/Z3. You can balance it in several ways, as he shows, to include the rising VC impedance. I always found that interesting.