[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: USB sound cards



Series resistors are often put in for reasons of stability, for example when driving capacitive loads. An example is shown in figure 2 of the attached PDF.

 

I also always believed that a low output impedance is optimal for precisely controlling a "reactive" load like a loudspeaker, although this is of course not the most power-efficient design (-> see impendance -matched transmission lines). At least that is what most texts on amplifier design suggest - but anyway, that might be wrong and hey, I'm only a psychologist, not an electrical engineer ;-)

 

[Although I actually had to build a low-noise headphone amplifier for my PhD experiments, because back then my lab wanted to save the money for something like a TDT device… Interesting experience.]

 

Best

 

Daniel

 

 

PD Dr. Daniel Oberfeld-Twistel

Johannes Gutenberg - Universitaet Mainz

Department of Psychology

Experimental Psychology

Wallstrasse 3

55122 Mainz

Germany

 

Phone ++49 (0) 6131 39 39274

Fax   ++49 (0) 6131 39 39268

http://www.staff.uni-mainz.de/oberfeld/

https://www.facebook.com/WahrnehmungUndPsychophysikUniMainz

 

From: AUDITORY - Research in Auditory Perception [mailto:AUDITORY@xxxxxxxxxxxxxxx] On Behalf Of James Johnston
Sent: Thursday, December 18, 2014 6:16 AM
To: AUDITORY@xxxxxxxxxxxxxxx
Subject: Re: USB sound cards

 

Not sure of your point, Jont. Of course it is the current through the voice coil that actually moves the diaphragm and thus makes the pressure.

None the less, the back EMF is key, especially for low-frequency system designs, and the back EMF is converted to current by both the internal and external resistance.  This is a key part of the filter design problem that makes the system something approximating flat at low frequencies.

I'm not sure who you think is putting resistors in series, one generally avoids that, except for the obligatory resistance of a crossover if one is using a passive crossover system.

 

On Wed, Dec 17, 2014 at 3:51 AM, Jont Allen <jontalle@xxxxxxxxxxxx> wrote:

Jim + Auditory-list

While it is true that speakers are "designed" to be driven by the voltage, our detailed research has shown that it is the
current that controls the pressure (i.e., force on the cone). If you want the details I already sent the links, but I'll send them once
more, just in case you missed it.

The theory of how the current drives the force is given by our ASA presentation from last month, at the special session on historic transducers:

http://hear.ai.uiuc.edu/public/Allen/NK_ASA_Indianapolis.pdf
* Kim and Allen (2014) Historic transducers: Balanced Armature Receiver (BAR), ASA meeting INDY, IN, Oct 30, 2014 pdf


The full theory is in a Hearing Research article:

http://hear.ai.uiuc.edu/public/KimAllenMar20.13.pdf
* Kim, Noori and Allen, J.B. (2013). "Two-port network Analysis and Modeling of a Balanced Armature Receiver, "Hearing Research, special MEMRO 2013 issue (pdf)

Ms Kim's (i.e., Noori's) PhD thesis has the most complete story, and is available directly from her (noorimail@xxxxxxxxx).

Because it is the current that determines the pressure, via the electrical input impedance, and because this impedance can be (i.e., is) reactive (have inductance), it can be tricky to figure out what is going on. One must know (i.e., measure) the load impedance, or at least the current. Also the distortion is mostly determined by the nonlinear flux vs current (i.e., B-H curve), as described by Hunt, for example.
Hunt mostly quotes Mott and Minor's BSTJ Jan 1951 article, adding little new material to their story.

In the interesting case of the Etymotic ER-3 (insert headphone), they place a 10 ohm resistor across the input terminals.
So if you try to beat this with a series resistor, you will not be successful in changing anything, because all the current is going into that 10 ohm resistor.
I opened up one of my ER-3 and cut out the resistor, and then I could see what was really going on. It worked better too.

When that ER-3 went back for repairs once, they were shocked to see the resistor was missing. I had to admit I cut it out, and wanted it left that way.

In general, adding a series resistor to your circuit, is working blind. You will likely get a random result, and random is never better.
Thus I dont advise it unless you are willing to measure the input impedance and the resulting frequency response of the headphone. While there
are some interesting things you can do with a series impedance, a series resistor is not one of them.

Jont Allen



On 12/16/2014 06:12 PM, James W. Beauchamp wrote:

I agree with Bob. Most loudspeakers are designed to be driven by
an ideal voltage source. Under this criterion they attempt to get 
the flattest possible response overall. This is not necessarily 
the most efficient response. In fact, some loudspeaker designs, 
like the Small-Thiele closed-box non-vented design, are very 
inefficient.
 
Jim
 
Bob Masta wrote:
From: Bob Masta <audio@xxxxxxxxxxx>
Date: Sun, 14 Dec 2014 09:15:19 -0500
To: AUDITORY@xxxxxxxxxxxxxxx
Subject: Re: USB sound cards
 
On 13 Dec 2014 at 21:59, Richard F. Lyon wrote:
On Sat, Dec 13, 2014 at 5:52 AM, Bob Masta <audio@xxxxxxxxxxx> wrote:
 
It's the other way around:  Adding resistance in the
driving circuit gives poorer damping.  "Damping Factor" for
a power amplifier is the reciprocal of output impedance.
 
Bob, I wasn't aware of that definition.
I was thinking of http://en.wikipedia.org/wiki/Damping_ratio
rather than http://en.wikipedia.org/wiki/Damping_factor
Oops, you are correct... I should have said it is 
*proportional* to the reciprocal.
 
Is there an understanding of why high "damping factor" would be good?
Jont's findings suggest otherwise (I believe he's saying the current is
typically more relevant than the volage).
The second reference you cited covers it under 
"Explanation".  Basically, a conventional electrodynamic 
speaker is both a motor and a generator.  Imagine that the 
speaker receives a momentary voltage pulse, after which is 
is instantly disconnected from the source.  The speaker 
would ring at its resonant frequency, damped only by 
friction.  The generator would be creating a voltage, but 
no current, so no load to add damping.
 
However, if instead of disconnection the leads were 
*shorted* after the pulse, the generator would be driving 
all its current into the zero-ohm load, giving a maximum 
damping effect.  
 
Conventional loudspeakers are designed to be driven by 
voltage sources, not current sources. The current may be 
more "relevant" (in the sense of force generation), but not 
for getting a flat frequency response from a conventional 
speaker design.  (Although there have been occasional 
attempts at current drive, the ones I recall required 
special dedicated amplifiers.)
 
Best regards,
 
Bob Masta
 
           D A Q A R T A
Data AcQuisition And Real-Time Analysis
          www.daqarta.com
Scope, Spectrum, Spectrogram, Signal Generator
   Science with your sound card!

 



--

James D. (jj) Johnston

Independent Audio and Electroacoustics Consultant

Attachment: design_tames_highspeed_opamp_an257.pdf
Description: design_tames_highspeed_opamp_an257.pdf