Author Topic: Which Op Amp Supply Bypass Technique for Low Noise?  (Read 2499 times)

0 Members and 1 Guest are viewing this topic.

Offline ShredheadTopic starter

  • Regular Contributor
  • *
  • Posts: 50
  • Country: us
  • Yup
Which Op Amp Supply Bypass Technique for Low Noise?
« on: June 18, 2014, 05:51:47 pm »
Hey guys, I was reading that there are two ways to bypass the supply pins of an op amp.  Just so we're on the same page, I'm using .1uF ceramic caps.  The 2 ways:

1. -Use 2 caps per op amp, one goes from the positive supply pin to ground and the other from negative supply pin to ground.

2. -Use 1 cap per op amp, one end goes to the positive pin, the other end to the negative pin.  Not involving the ground.

The application is for audio EQ (powered with +/-15V from linear regulators) so the lower the noise the better.  I read that using method #1 is more of a risk because if there is any noise on your power supply, it will inject that noise into your ground.  I have been using method #2 but I don't know which is better.


Has anyone out there ever tested which of the two methods is better?  All opinions welcomed.   :-+
 

Offline kizzap

  • Supporter
  • ****
  • Posts: 477
  • Country: au
Re: Which Op Amp Supply Bypass Technique for Low Noise?
« Reply #1 on: June 18, 2014, 06:13:55 pm »
you could run LP filters onto the power inputs ;)

Or even ferrites if the frequency is high enough!

That said, I believed preferred way to do it is to decouple each of the rails to ground.
<MatCat> The thing with aircraft is murphy loves to hang out with them
<Baljem> hey, you're the one who apparently pronounces FPGA 'fuhpugger'
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 16545
  • Country: us
  • DavidH
Re: Which Op Amp Supply Bypass Technique for Low Noise?
« Reply #2 on: June 22, 2014, 04:01:48 am »
Analog Devices has an old but excellent application note discussing this.  The decoupling capacitors, one for each supply, go between the power pins and the ground return for the output stage of the operational amplifier which may not be the same as the signal ground.  RC or low Q LC filters may be appropriate for the power supply decoupling in some cases.
 

Offline VK5RC

  • Supporter
  • ****
  • Posts: 2672
  • Country: au
Re: Which Op Amp Supply Bypass Technique for Low Noise?
« Reply #3 on: June 22, 2014, 10:48:34 am »
If you use the one cap version might you be coupling some noise onto the negative rail?
Whoah! Watch where that landed we might need it later.
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 16545
  • Country: us
  • DavidH
Re: Which Op Amp Supply Bypass Technique for Low Noise?
« Reply #4 on: June 22, 2014, 12:53:47 pm »
If you use the one cap version might you be coupling some noise onto the negative rail?
Even worse, the noise from the positive rail will be coupled directly into the output if the operational amplifier because one of the supplies is used as the reference for the integration stage.  Notice that most operational amplifier datasheets show a lower power supply rejection for one power supply rail than the other.  The ones that do not probably measured it incorrectly and at least in my experience, SPICE is useless for modeling this although a that is a macromodel problem.

Linear Technology application note 41, Questions and Answers on the SPICE Macromodel Library:

http://www.linear.com/docs/24958

"Signals having frequency components above the amplifier closed-loop bandwidth will be transmitted from V- to the output with little or no attenuation."

Analog Signal-Handling for High Speed and Accuracy - PDF Page 5:

http://www.analog.com/library/analogDialogue/bestof/pdf/11_2a.pdf
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf