Does anybody have an opinion on the 'optimal' capacitor value for the Gainclones zobel?
National's LM3875 datasheet recommend: 0.1uF with 2R7.
47 Labs: 0.01uF with 10R
Thanks,
- John
National's LM3875 datasheet recommend: 0.1uF with 2R7.
47 Labs: 0.01uF with 10R
Thanks,
- John
I used to think that the Zc did the work. But my continuing learning now shows that it is the Zr that does the work.
The job of the Zc is to determine when the resistor starts to work and to stop the Zr blowing up.
On that basis the Zc value seems to somewhat variable and certainly not critical.
The job of the Zc is to determine when the resistor starts to work and to stop the Zr blowing up.
On that basis the Zc value seems to somewhat variable and certainly not critical.
Thanks Andrew - I'll stick with National's recommended value for that then. They seem to know a thing or two 😉
Will use 10R for the Zr though, for the reasons mentioned earlier in this thread.
Something interesting about zobels and chipamps.
IC Audio Power Amplifiers and Zobel Networks: One Size Does Not Fit All
Lots of math and stuff, but you can read around it.
The ideal values are impractical because you need a very high wattage resistor .
IC Audio Power Amplifiers and Zobel Networks: One Size Does Not Fit All
Lots of math and stuff, but you can read around it.
The ideal values are impractical because you need a very high wattage resistor .
Last edited:
Interesting article - thanks! By the looks of it the standard 'rule of thumb' values can potentially do more harm than good. Perhaps it's best to just leave it out (as Peter Daniel does), or have it wired directly across the speaker outputs once proper values are calculated with the loudspeaker in mind, if the chip is close enough to them...
Still I'd be surprised if National added something meant as a safety precaution which actually made things potentially worse?!
Still I'd be surprised if National added something meant as a safety precaution which actually made things potentially worse?!
hello johnm
please bear in mind that the datasheet specifies these component values of the zobel network concurrently with a resistive load of 8 Ohms and a parallel RL in series to that load. the actual application you have in mind is a speaker or even worse a loudspeaker enclosure with a crossover most probably. the RC zobel network depends on the total load the amplifier drives and its simplification/reduction to a single RL in series. that's why the usual formula dictates the resistor of the zobel to be equal to the total equivalent resistance of the load and the capacitor of the zobel to be L/R^2 where L is the inductance of the speaker and the connecting wires. if you had an oscilloscope you could conduct some tests with the actual load connected and watch out for high frequency oscillations. better yet, if you had an impedance analyzer you could determine yourself the equivalent RL of the total load.
please bear in mind that the datasheet specifies these component values of the zobel network concurrently with a resistive load of 8 Ohms and a parallel RL in series to that load. the actual application you have in mind is a speaker or even worse a loudspeaker enclosure with a crossover most probably. the RC zobel network depends on the total load the amplifier drives and its simplification/reduction to a single RL in series. that's why the usual formula dictates the resistor of the zobel to be equal to the total equivalent resistance of the load and the capacitor of the zobel to be L/R^2 where L is the inductance of the speaker and the connecting wires. if you had an oscilloscope you could conduct some tests with the actual load connected and watch out for high frequency oscillations. better yet, if you had an impedance analyzer you could determine yourself the equivalent RL of the total load.
- Status
- Not open for further replies.