From: "Tom Faloon"
Subject: Re: Choosing Feedback Resistors for Op Amp
Date: Tue, 1 Oct 2002 01:48:50 +0100
NNTP-Posting-Date: Tue, 1 Oct 2002 00:45:53 +0000 (UTC)
X-Newsreader: Microsoft Outlook Express 5.00.2615.200
The minimum value is decided by the amplifiers current drive capability.
The feedback resistors, plus any load, must not exceed the maximum output
current specified in the data sheet.
It is less easy to specify a maximum resistor value.
As a rule of thumb, don't use more than about 1M, otherwise stray
capacitance become significant, and you run a risk of spurious oscillations.
You can use much higher values, but they need careful circuit layout.
The above defines limits for the sum of the two resistors in your non
inverting amplifier. But the gain is defined by the ratio of the same two
resistors, so you need to choose something which matches both requirements.
This leaves many options, but these are often limited by other
You may want to use large resistor values to minimise power in a micropower
or small values because low noise, or voltage offsets are important
In your case, the OP547 is capable of driving high currents at high
voltages. 0.5A, ~60V
This may put a limit on component values. e.g. If you were using 0805
surface mount resistors, you could easily exceed their maximum power
Good engineering design is about finding the best compromise.
Rileyesi wrote in message
> Are there any guidelines for selecting the feedback resistors for a
> non-inverting op amp? I couldn't find anything in the data sheet or in my
> electronics texts.
> Specificially, I am using a Linear Technology OPA547 by Texas Instruments.
> am measuring a slow moving 0-4 VDC. I want a gain of one for one amp and
> gain of two for another. The formula is Vo=Vin(1+R2/R1). For a gain of
> simply set R2 to 0. For a gain of two, I simply set R2=R1.
> My question is this. What are the ramafications for the absolute resistor
> values? Should I use a 10 ohm resistor? 100 ohms?? 10,000,000 ohms?!?
> the past, I used what ever I had handy (usually in the 1,500 ohm range)
> problems. I was just wondering if there is a better way of selecting the
> resistance values.