Question on opamp power supply bypass

PJ · « **on:** 30 Sep 2003, 01:37 am »

Hi guys,

I found a low-noise preamp which I will be using for a uni project (not audio) and it has the opamps bypassed like this...

The opamps are OPA37.

The three stage preamp (three cascaded OPA37's) has a very low noise of around 4nV/rtHz.

What I dont understand the purpose of in the above diagram is the presence of the 100R resistors. Do the 100uF caps see them and they simply enhance the RC filter effect? Maybe I am just retarded today

Thanks!

_scotty_ · « **Reply #1 on:** 30 Sep 2003, 02:45 am »

Quote from: PJ

Hi guys,

I found a low-noise preamp which I will be using for a uni project (not audio) and it has the opamps bypassed like this...

The opamps are OPA37.

The three stage preamp (three cascaded OPA37's) has a very low noise of around 4nV/rtHz.

What I dont understand the purpose of in the above diagram is the presence of the 100R resistors. Do the 100uF caps see them and they simply enhance the RC filter effect? Maybe I am just retarded today

Thanks!

This is a guess but they are probably there to damp ocillation in the circuit. Caps alone won't do this depending on the bandwidth of the circuit.

eico1 · « **Reply #2 on:** 30 Sep 2003, 04:02 am »

here's an idea:

http://www.elecdesign.com/Articles/Index.cfm?ArticleID=1478&pg=3

steve

PJ · « **Reply #3 on:** 30 Sep 2003, 05:35 am »

Thanks guys:

Scotty:

I was thinking something along those lines but was unable to find any info on it.

Eico1:

That was one of the webpages I read when I was trying to find out...but that makes logical sense to me, as it forms a 6dB LP filter. I think the one I have shown does something similar, but just need someone to expain to me how it works...

I am pretty sure there is something simple that I have overlooked.

randytsuch · « **Reply #4 on:** 30 Sep 2003, 06:41 pm »

PJ,
Here's another link for you
http://www.audiocraftersguild.com/AandE/Supply_decoupling.pdf

Guido Tent wrote it, and it also talks about resonance on a IC's power input. He also talks about controlling it with an inductor.

He goes into a little more detail and talks about implementation

randy

audioengr · « **Reply #5 on:** 3 Oct 2003, 01:27 am »

You are on the right track. When you see resistors in series with the power feeds, it is usually to damp overshoot. Sometimes this is an artifact of the opamp design and it's bandwidth/transient response. Sometimes its due to poor layout on the circuit board. Really good opamp designs don't need this to be critically damped, however, placement of the parts is as important as their values. Remember, if you have three inches of wire/trace in the way, this is an inductor and will form resonant tank circuits as well as limiting bandwidth.

peranders · « **Reply #6 on:** 3 Oct 2003, 08:05 am »

Deleted

peranders · « **Reply #7 on:** 3 Oct 2003, 08:05 am »

Opamps fed with +- supply should allways be decoupled to ground -> 100nF pin 7 - Ground, 100 nF pin 4 - ground. Why? The load is related to ground.

You can also use small resistor values 10-100 (maybe a little more) in order to isolate the opamp from the rest of the circuitry. But if you have resistors I also recommend to have 10-100 uF in parallel with those 100 nF.

PJ · « **Reply #8 on:** 4 Oct 2003, 06:47 am »

Thanks guys!