Frequency response vs. output power

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Mike,
Bode is the keyword you're searching for. As in Bode plot. Yes, the bandwidth shrinks as the power (interpret this as gain) increases.
Chip opamps have really stupid open loop bandwidths...like 100Hz. It's only after you apply gonzo amounts of feedback that the bandwidth becomes tolerable.
If you meditate on all the ramifications and implications of this junk you're likely to end up like me--meaning someone who thinks that 250kHz to 300kHz open loop bandwidth is a good start. I say that having just come up from the Dungeon wherein my Evil Experiments managed a whole, astounding 10kHz. Unsatisfactory. Need to stop and think things over. Bummer.

Grey

P.S.: Good talking to you. There's no one to talk audio with out here in the hinterlands.
 
Formerly "jh6you". R.I.P.
Joined 2006
GRollins said:
Yes, the bandwidth shrinks as the power (interpret this as gain) increases.


Somewhat difficult for me to understand.
I am simplified. Nevertheless, as the rolloff is hapenning
depending purely on the RC network, I understand that the
rolloff point is independant of the power level.
I.e. the bandwidth of an amp is the same at any power level.
The distortion might be increased at the power increase, though.

Where I'm wrong . . . ?
 
Babowana said:



the rolloff is hapenning depending purely on the RC network



Unfortunately, this is one of those 'ain't no free lunch" things. Yes, the RC network (I'm assuming that you're referring to the feedback loop) rolls off the frequency response, but even in a circuit with no feedback at all, the frequency response will decrease at higher gain levels.
Plays havoc on my intentions to build wide bandwidth circuits.

Grey
 
MikeW said:
How much does the frequency response change for one watt output to full rated output power on a Pass labs amp? All the reviews I read are for small signal frequency response. Is there a difference. Does anyone have some data?


Mike,
There is a concept in amplifiers called "Gain-Bandwidth product". Basically an amplifier typically has a constant value for the closed loop gain * bandwidth of the amplifier. In other words, for a given amplifier, your bandwidth changes porportionally to the gain you set (typically through NFB). In a normal amplifier, your gain will be fixed regardless of the input signal (and therefore the output signal).

So, the bandwidth, which is essentially your frequency response, should not change regardless of output power because the output power was adjusted simply by injecting a stronger input.

NPs response should validate this statement.

Typically NFB is used to adjust the closed loop gain (and therefore expand the bandwidth). Although emitter/source degeneration is typically not considered feedback, it does affect this as well. This is why, as a previous poster said, many opamps have open loop bandwidths of only 100Hz but when feedback is applied (ie, the closed loop gain is reduced), the bandwidth is extended.

Now, if you instead adjusted the output power by adjusting the through the feedback/degeneration or Rc/Rd's then you woudl be adjusting the bandwidth as well.

--
Danny
 
Formerly "jh6you". R.I.P.
Joined 2006
GRollins said:
Unfortunately, this is one of those 'ain't no free lunch" things. Yes, the RC network (I'm assuming that you're referring to the feedback loop) rolls off the frequency response, but even in a circuit with no feedback at all, the frequency response will decrease at higher gain levels.
Plays havoc on my intentions to build wide bandwidth circuits.

Grey



I was referring to the original question.

I still believe that the different output power levels change neither R nor C values, which are forming internal RC networks and their critical frequency points.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.