• Welcome to ASR. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Motional Feedback (MFB)?

DanielT

Master Contributor
Forum Donor
Joined
Oct 10, 2020
Messages
6,947
Likes
7,948
Location
Sweden - Слава Україні
It seems to be a smart system, right? Common in constructions nowadays?

In the art of audio sound reproduction it is well-known that the dynamic loudspeaker is more nonlinear and generates more distortion than all the other system components combined. This is particularly true at low frequencies which require large cone excursions where the stiffness of both the inner spider and the outer surround increases rapidly as the cone approaches its peak displacement, resulting in a nonlinear suspension compliance generating high distortion.
For example, in a typical high fidelity sound system at a frequency of about 35 Hz the total harmonic distortion of the amplifier might be of the order of 0.01%, whereas the distortion of the loudspeaker might range from about 3.0% to about 50.0%, depending upon the loudness. If this cone motion can be sensed and given as a feed back to the earlier stage of the amplifier, this distortion can be reduced dramatically.

Motional Feedback (MFB) was a speaker system developed in the early 1970s by Philips Holland. It introduced a feedback system to the woofers of HiFi loudspeakers, enabling them to achieve a more extended low frequency response in a relatively small enclosure. The key benefits are a very controlled bass response. Any distortion induced by the enclosure or the woofer itself is immediately corrected by the feedback. These hand-built speakers were sounding very good and were quite expensive
.



Aha! Separate solutions for it are sold! Maybe overkill for a beginner in DIY, but still:

 
Velodyne uses a servo for this.
1.png

2.png

3.png

4.png
 
It seems really smart, MFB . Reasonable too. Amplifiers and speakers will work together in practice. What is written on the paper, technical data for each separate unit, amplifier and speakers is one thing, but I am at least more interested in how the boat performs after it has left the windless harbor and comes out on the open sea, so to say.:)
 
 
Would be great to see some objective measurements of these DIY kits. In theory, it should linearize the response and bring down distortion significantly. I'm also wondering what the effect would be in case you have two identical woofers, where one of them uses the feedback. How well does it then work for the second one?
 
In theory, it should linearize the response and bring down distortion significantly.
One important goal of MFB is not distortion reduction (distortion can't be any lower than that of the sensor used), rather stabilize everything and make the piston cone motion fully predictible and free from chaotic effects, drifts, etc.
For best results wrt stability in woofers/subwoofers you need to monitor cone excursion, absolute position. That ususally leads to a mixed sensor design as position sensing only is fraught with resolution issues at higher frequencies and that's why velocity and/or acceleration sensors are used.
 
I had a pair of Philips MFB loudspeakers way back when, and very good they were too. Gave them away to an impecunious friend who used them for a good while.

The principle works very well at low frequencies, but can't cope with cone breakup, as the unwanted cone movement doesn't get reflected in the accelerometer as that's attached to the coil former. Provided the woofer isn't operated above any breakup frequencies, then it's pretty much 'perfect', but that's hard for a two-way. Still, very much an improvement over a small conventional 'speaker, and at little extra cost except for some circuit complexity and adding a second coil as a sensor.

S
 
but that's hard for a two-way .... okay but can be used in full range? Although it might be difficult to get working in a proper way.

Is it possible to use MFB over the entire register? 20- 20 000 Hz ?!
Just thought it was for Subwoofers? Hm well why not full range, if possible.

Even more interesting in that case.:)
 
Last edited:
One important goal of MFB is not distortion reduction (distortion can't be any lower than that of the sensor used), rather stabilize everything and make the piston cone motion fully predictible and free from chaotic effects, drifts, etc.
Guess what the result of that is...
 
Noise cancelling basically is MFB but for lower frequencies only and not only driver movement only (which MFB is)

MFB can only be used for lower frequencies because cone-breakup can not be compensated for.
 
but that's hard for a two-way .... okay but can be used in full range? If it might be difficult to get good?

Is it possible to use MFB over the entire register? 20- 20 000 Hz ?!
Just thought it was for Subwoofers? Hm well why not full range, if possible.

Even more interesting in that case.:)
I don't know about full range, but it seems you can use it for bass to the midrange at least. It will be hard to fit an accelerometer to a tweeter. The mass alone would make it impractical. Also, you can only use it in the area where the drive unit is actually mostly pistonic. So you'll need to stay away from resonances.
 
Philips didn't abandon MFB for no reason though.
Their attempt was to extend low frequency response with smaller cabinets.
The issue was maximum excursion though. The smaller ones clipped hard and audible.
There was still a lot of midrange coloration.
The larger ones were better (3 way systems).
As was already mentioned, the excellerometer (a ceramic disc) was the limiting factor.

Below the QC 35-II without NC (MFB)
dist-nc-off-r-percent.png


below the QC 35-II with NC on)
dist-nc-on-r-percent.png


It works well up to 500Hz.
 
Philips didn't abandon MFB for no reason though.
Their attempt was to extend low frequency response with smaller cabinets.
The issue was maximum excursion though. The smaller ones clipped hard and audible.
The Piratelogic amplifier kit adds a limiter for this exact reason.
There was still a lot of midrange coloration.
Why was that?
The larger ones were better (3 way systems).
As was already mentioned, the excellerometer (a ceramic disc) was the limiting factor.
Luckily technology has advanced a bit in this regard as well.
 
Why was that?

Because basically only the voicecoil movement was corrected up to a certain frequency. Cone break-up was not compensated for.

With NC the problem is delay (wavelength) and the mic position/quality.
 
Because basically only the voicecoil movement was corrected up to a certain frequency. Cone break-up was not compensated for.
Ah, the age-old adage of too high crossover in 2-ways, combined with too wide bandwidth feedback.

Here is a video of the effect of the motional feedback on the frequency response:
 
Last edited:
Back
Top Bottom