Low Frequency Preamp

Thread Starter

ritchie888

Joined May 18, 2012
10
Multisim shows (severe distortion) that a lousy old 741 opamp (45 years old) does not work when its supply is less than about 10V.
Yeah, unfortunately I had to use a different op amp as the 971 isn't in Multisim. Rather than search for a low noise, low power op amp in their database I just went with the first one that came to mind!

I'm still getting a bit of a jump/error in and around 2Hz at the start, but it sorts itself out fairly quickly to a near perfect state. I figure that's either down to the op amp being used or some time that is needed to charge the capacitor, does that seem right?

Having used:

Gr = 20 * log10((w * R * C) / (sqrt(1 + (w * R * C)^2)))

and a range of frequencies from 1-100Hz (w being 2*pi the frequencies) and plotted in MATLAB, I'm fairly certainly the component values are correct now.
 

Audioguru

Joined Dec 20, 2007
11,248
With such a low frequency response, the capacitors take a long time to charge. Until the capacitors are charged then the output will be distorted.

A capacitor charges a little through a resistance in a time that is R x C. That is ONE TIME CONSTANT. The capacitor is almost fully charged after 5 time constants.
 

#12

Joined Nov 30, 2010
18,224
Just for reference, a supplement for what AG said, the time constant for (2) 150k resistors in parallel and a 220 uf cap is 16.5 seconds. 5 time constamts will be 82.5 seconds. Be ready to accept that much time for the circuit to "settle down".
 

Ron H

Joined Apr 14, 2005
7,063
Just for reference, a supplement for what AG said, the time constant for (2) 150k resistors in parallel and a 220 uf cap is 16.5 seconds. 5 time constamts will be 82.5 seconds. Be ready to accept that much time for the circuit to "settle down".
And the -3dB corner is at .01Hz, which is about 20 times lower than it needs to be. A 10uF cap will result in the response being down about .05dB at 2Hz. The settling time will go down to about 4 seconds.
 

Thread Starter

ritchie888

Joined May 18, 2012
10
And the -3dB corner is at .01Hz, which is about 20 times lower than it needs to be. A 10uF cap will result in the response being down about .05dB at 2Hz. The settling time will go down to about 4 seconds.
Ah, interesting and useful, thank you.

Any suggestions to cut even more time off would be most appreciated.
 

#12

Joined Nov 30, 2010
18,224
First reaction: Build it the way I showed you.

It should be obvious to you by now that fast settling time costs you some phase shift. You are in a typical engineering decision. How much phase shift can I accept versus how fast do I need it to stabilize.

Play with the capacitor size and make a decision.
 
Top