Resistor from Non Inverting Op Amp terminal to Ground - or not?

Some designs/designers put a resistor from the Non Inverting terminal to Ground - while others don't

My understanding is that this resistor can help balance the effects of bias currents on DC offsets and help reject common mode noise.

I've also read that this can potentially cause as many problems as it can solve (on certain op amps?...)

Using the TL074 in both DIL and SOIC in +/-12V circuits running at around audio frequency - is there a 'general' rule or reason to go with one or the other (with resistor or without)?

I can't say that using it has ever 'fixed' anything that I've noticed... but much smarter people than me do decide to use them.

Thanks.

Dear juz_ad
Hi
Where is your circuit? ( your schematic?) . though probably to provide input bias of transistors of 1st stage of opamp.
but for more information , it is important to attach your circuit.
Best Wishes
Goldsmith

A TL071 has FET input transistors so the input current is very small (go and check) and can be ignored for most aplication. Because of its very high impedance any leakage through external components, such as coupling capacitors will result in a DC shift through the amp.
If you use an op-amp with junction transistors at the input, they WILL draw their bias currents through their resistors, so a resistor must be used on the non-inverting input to ground and it should be about the same as whatever is connected to the inverting input ,to minimise drift with temperature.
Frank

juz_ad said:

My understanding is that this resistor can help balance the effects of bias currents on DC offsets and help reject common mode noise.

I've also read that this can potentially cause as many problems as it can solve (on certain op amps?...)
.

Click to expand...

I think, your understanding is completely correct.
And - I must confess that up to now I have never experienced any problems using this dc bias balancing scheme.
Where did you read about some "problems" - and which kind of problems?

chuckey said:

A TL071 has FET input transistors so the input current is very small (go and check) and can be ignored for most aplication. Because of its very high impedance any leakage through external components, such as coupling capacitors will result in a DC shift through the amp.

Click to expand...

That answer seems to suggest that I can safely ignore it for most circuits using the TL072 (not TL071 - that was my mistake) - unless there is a biasing or offset-related problem that needs fixing, then - the resistor from non-inverting to ground is one potential solution/safeguard? Is that a correct interpretation?

LvW said:

I think, your understanding is completely correct.
And - I must confess that up to now I have never experienced any problems using this dc bias balancing scheme.

Click to expand...

So - are you saying that you generally *do* use a biasing resistor from the non-inverting terminal? Your answer reminded me of a circuit I was shown recently (for input protection) using the TL072 where it was advised that using the biasing resistor would be a good decision - that's advice that I accepted.

LvW said:

Where did you read about some "problems" - and which kind of problems?

Click to expand...

I can't quote anything directly - sorry. I *think* the assumption was that unless a component was needed - it was potentially just adding more noise and the potential for more error.

I don't mind making it a standard thing for best practice, but, especially when using the TL072 - if it's not doing anything specific - does it really need to be there?

Thanks for the input - appreciated.

The extra resistor can certainly increase thermal noise but also help reduce common mode interference. In transimpedance configuration the noise is usually important so I wouldn't use the extra resistor then.

Keith

keith1200rs said:

The extra resistor can certainly increase thermal noise but also help reduce common mode interference. In transimpedance configuration the noise is usually important so I wouldn't use the extra resistor then.

Keith

Click to expand...

Yes, I agree with Keith.
However, as indicated in my last posting, I would not consider this possible noise increase as a "problem". Why not?
For my opinion, each circuit modification with the aim to improve one particular performance parameter (in this case: output offset) will cause a degradation of another property resp. performance parameter (in this case: noise). I think, this is a general rule that applies to all cases - perhaps with very few exceptions only.
This means: Each good circuit design in electronics finally results in a trade-off between some conflicting requirements.
And each designer has to be aware of this.
Basic example: BJT stage in common emitter configuration. Increase of dc collector current is good for high gain but increases power consumption and decreases input resistance.
It would be interesting to hear (resp. read) the opinion of other forum members to this "rule".
Regards to all.
LvW

So - are you saying that you generally *do* use a biasing resistor from the non-inverting terminal?

Click to expand...

I generally don't use bias current compensation resistors with FET OPs or input current compensated bipolar OPs. Because most modern bipolar OPs have input current compensation, I effectively don't use it at all.

In case of TL074 JFET input stage, there's still a systematical bias current present, in so far, a compensation resistor may be reasonable in special applications. It should be used only, if it brings a noticable improvement in overall offset voltage and drift and the noise increase can be tolerated.