Mfb

Status
Not open for further replies.
Jump to Latest
#21
While I assume that MFB will work properly in the frequency-range where a driver is working as a solid piston I doubt that the waveshape of the sound emerging from a driver will have the same waveshape as the motion of the voice-coil or one part (where the feedback signal is taken of) of the cone as you move further up in frequency.

I think there is no way getting around using good drivers (I don't say expensive !!!!) to make good-sounding loudspeakers.

Regards

Charles
 
#22
Nelson Pass said:
All these techniques have been around a long long time:

Microphone feedback
Accelerometer feedback
Dual voice coil feedback
Current sensing feedback
(Have I missed any?)
They all work, and they all have their limitations: specifically,
instability when applied for more than a few dB.

Microphone and current sensing are my favorites, but I
have to admit that I've enjoyed all of them at one time or
another. Of the bunch, the accelerometer technique is
the biggest pain in the butt.
Click to expand...

Hi, You missed few 😉

maybe: capacitance feedback,
my experiments with laser feedback, optocoupler feedback element (like tension meter in Tascam tape).

The biggest problem is (IMHO) delay between corrective signal and moving of piston. (generating IM distortion and broken closed loop at the end, maybe)

Current sensing is most popular way (Genelec Active speaker?).

regards
moamps
 
#24
Of all the MFB versions the easiest for DIY would be the one using a microphone, but I assume it will work only well at low frequencies.
Capacitive on the other hand could be implemented easily for metal dome - tweeters (though I am not sure how one should handle the pretty nasty cone resonances).


Regards

Charles
 
#25
Actually the bandwidth limitations are not only at high
frequencies - at lower frequencies you will also run into
problems as the speaker response drops off.

The microphone technique was used in the Phantom
Acoustics Shadow (Pat # 4,899,389) worked very well
either as an acoustic black hole or as feedback for an
active woofer.

For stability we rolled the feedback with a single pole on
either side of 60-70 Hz or so.
 
#26
:: the cheapest ::

What about this approach?? [is this current sensing?] 😕
I think this the cheapest way to applicate the MFB .... 😀

Regards:
ragil.hastomo
 

Attachments

  • mfb.jpg
    mfb.jpg
    13.2 KB · Views: 716
#27
Selamat malam Mr. Hastomo

What you posted was quite often called MFB. In my opinion it is not.
It is a way to generate a negative output resistance of the amplifier. This way the Qtc of a driver can be lowered (though it would be wiser to take a stronger mangnet from the beginning).

Using current feedback, almost any output impedance (within reasonable limits) can be achieved, depending on the feedback- and forward-path functions. The ACE - Bass circuit works this way.

Regards

Charles
 
#28
Nelson Pass said:
(Pat # 4,899,389)
Click to expand...

Sure this number is correct? US4899389 is an extended Japanese patent for a pulsating noise removal device by a Mr. Kiyoshi. It seems to be a means to remove ignition noise from a low-level audio signal (wonder if it removes those techno-beats too...). Can't see how this corresponds to a MFB woofer.

Eric
 
#29
Ragil,
What about this approach?? [is this current sensing?] I think this the cheapest way to applicate the MFB ....
Click to expand...
For one thing, current feedback is not “motional” feedback, no big deal though. 🙂 The real issue with current feedback is that it has a non-linear relationship with the cones motion. This is particularly a problem in the area of the drivers resonant frequency, which would be even further complicated in a vented box alignment. There would have to be some type of compensation for the drivers impedance (an therefore the current) variations over the frequency range of interest.

Rodd Yamas***a
 
#30
Here's a recent example of the microphone feedback school of thought--the Meyer X-10 studio monitor. Eye-popping performance boasts. It credits control circuitry adapted from new avionics advancements. Guess we'll have to wait for the patent to see any specifics.
 
#31
Meyer X-10

Those seem to be really cool speakers. If they perform as one would expect from the data they supply, they might even outperform some expensive high-end speakers that are in the same price range ($ 40k/pair, fully active !).

I think the MFB thing wasnt an easy task here because they are vented :bigeyes: 😎

I have a paper (in German) about a German speaker (maybe it was by Backes & Mueller) where the Microphone is glued to the woofer's cone (as it is with the new Manger subwoofer) for a closed box. Application of MFB to closed boxes is much easier than reflex enclosures.

Regards

Charles
 
#37
I’ve wondered for some time about using an optical mouse for motion feedback on a transducer. Of course decoding the USB output through a PC or would likely add too much latency to be of any use. It seems though that a time related magnitude of change could likely be decoded without a whole big fat computer.
 
#38
I started a MFB project some times ago.. still wondering what kind of sensor I'll use, I tried IR and it worked but not the way I wanted it to.

I think i'll try using piezo a piezo element but I'm not sure how good it'll be on very low freqs.

Anyone ever did that with IR and had good results?
 
#39
Capacitance feedback and magnetic fields

Thought I'd resurrect this tread to ask about capacitance feedback. I'm fishing around for a method of instantaneous diaphragm position sensing, and this seems about the simplest. What I'm wondering is would eddy currents in the sensor plates from magnetic fields modulate the static charges and generally mess anything up? If not, what about using an aluminum voice coil and pole-mounted shorting ring as the two halves of a variable capacitor?

Bill
 
Status
Not open for further replies.
Top Bottom