In Reply to: Re: calling all amplifier guys..jensw had a good question.. posted by Dan Banquer on August 28, 2003 at 02:53:16:
What I was wondering was..For an amplifier, is it possible that even with a high bandwidth, it is capable of having transient delays in the 10 to 20 uSec range?
The particular scenario I speak of is this:
Two sounds on the audio..vocals centerstage, and a major bass guitar right channel..
The right channel output transistors have a big dissipation problem..the speakers I'm using (fictional, but go with it) have a large reactance at one particular note.. So, when that note is hit, the load line for the outputs becomes very reactive, forcing that channel into 4 quadrant operation. That means, there will be times when the output node is full negative, but the load is trying like hell to pull that point up. IE, a circular V/I characteristic.
Given the time constants of the output silicon to the heatsink,(10 uSec through the silicon, 10 milliseconds to the xistor case, 100 mS to the heat sink) it is possible to bring the die junction temp(remember, the junction is in the first coupla hundred microinches of the silicon.) up, as the output bias circuitry can only track either the heatsink, or the case of the xistor.
My conjecture is this: can the load reactance force the output die to heat up sufficiently to drop the bias current in the outputs, and temporarily cause transient delays?
Since my scenario has only one channel's outputs being stressed that way, any channel dependent delays would cause the vocal image to modulate, or "blur".
Your thoughts?
Cheers, John
This post is made possible by the generous support of people like you and our sponsors:
Follow Ups
- Hi Dan.. - jneutron 05:19:47 08/28/03 (5)
- Re: Hi Dan.. - Dan Banquer 09:52:59 08/28/03 (0)
- I forgot to mention.. - jneutron 05:59:03 08/28/03 (3)
- Re: I forgot to mention.. - Dan Banquer 09:47:03 08/28/03 (2)
- Re: I forgot to mention.. - jneutron 10:24:16 08/28/03 (1)
- Re: I forgot to mention.. - Dan Banquer 11:50:31 08/28/03 (0)