2. tech-talk
  3. 1104517282

Current source/Voltage source, and like questions

In another thread there has been some discussion of power amps configured to deliver current proportional to input signal, rather than the more common voltage proportional to signal. A voltage amp's current varies with the load (speaker) impedance. Is this really bad? After all the speaker designer developed and evaluated his product with this effect in mind.

It seems to me that a simple network could be used to make a voltage amp function as a current amp. The input signal would be compared with the voltage across a current sampling resistor in the return leg of the speaker circuit, and the resulting modified signal used to drive the amp. Furthermore, the amp could be made to function as voltage or as current, and anywhere in between, by use of a potentiometer. This would provide hours of fun for curious audiiophiles.

In a similar vein, I recall so-called "ceramic" phono pickups, which had a piezo crystal to generate signal instead of a coil of wires. Most of these were very cheap items used in low cost record players, but there was at least one model that was comparable to good MM pickups. The interesting thing is that the piezo pickup signal is a function of stylus deflection (position) whereas a dynamic pickup signal is a function of the rate of change of stylus deflection. However, either type of pickup could play the same groove and yield similar sound. Which one is "right"? Perhaps neither.
eldartford
timpani
40 posts
 
you may be confusing amplifier gain with current and voltage feedback. Most Power Amps employ global voltage fb but current fb is becoming increasingly popular due to better bandwidth and higher slew rate. Without going into the nitty gritty, the differences between the two are in mainly in the configuration of the differential input stage.
Other than that, all power amplifiers must (and do) amplify the applied input voltage by a fixed amount (usually 30dB), AND as an ideal voltage source are able to supply a load current inversely proportional to the load resistance (impedance), and proportional to the output voltage.
If you need further info on current or voltage feedback design then a google search is a good start.
eldartford OP
4,262 posts
 
Timpani...The "simple network" that I suggested is the "current feedback" which you mention.
timpani
40 posts
 
Eldartford,

Then, yes your idea for current fb is quite valid. Current is sensed to the load and compared at the diff input. As with voltage feedback though, current fb also comes with its own set of penalties. Some of these include potential instability into inductive loads (opposite to Vfb) and input sensitivity to load impedance. The latter implies that a lower drive impedance is required. I'm not sure if it's been done yet with power amplifiers, but it would seem to me that the most optimum amplifier design would incorporate both voltage and current fb, and specifically, utilise Vfb at low and mid frequencies and Cfb at hf. This would achieve both speed and stability which are usually mutually exclusive in Vfb only designs.

Food for thought.
Timpani
eldartford OP
4,262 posts
 
Timpani...Your suggestion that current feedback should be frequency dependent sounds right, but I think it should be the other way around. It is the woofer that exhibits that big impedance hump at resonance, where current control would be effective. Also, stability issues would be minimal at LF.

I recall that at one time there was a brand of loudspeakers that had some kind of network in them with feedback to the power amp. I wonder what that was all about. It was supposed to be a substitute for mounting an accelerometer on the cone driving a servo (like Velodyne and others). I guess that the servo approach probably does the same thing as a current amp, and better.
More to discover