In your schematic, the left op-amp is used to buffer the voltage produced by a pot, and is used to bias the right hand op-amp. The left op-amp circuit is not needed. Replace it with two 10K resistor between Vcc and gnd. Their junction point will be 1/2 of your Vcc. Use that to bias the right op-amp. You have been shown examples of this by several posters. Please look at the circuits shown to you. They are all good ways to do what you want to do.

 

The reason I am using a second op amp as voltage follower is because I thought this would give a more stable virtual ground, what is wrong with my circuit? I thought that because the op amp does not draw current through the inputs the total power consumption of the circuit would be reduced because of the higher resistance voltage divider.

Thanks for all the replies!

To answer your question, there's nothing WRONG with doing this. But your circuit doesn't need it...this isn't a high-current application or one where the impedance of your bias-ground makes a lot of difference. A pair of resistors (and maybe a cap) will do fine. I fix battery-powered audio devices on a regular basis (guitar stomp-boxes), and I basically never see anyone use the op-amp buffered ground scheme. But using it will not stop your circuit from working.

 

For starters, the 1meg Pots are way too high in resistance for LM324 opamps. Their input bias current is too high to use such high resistances.

Lose that useless simulator; it shows ideal behaviour only.

Thanks, as I said I am new and learning a lot. I thought 5V/1m = 5000na right? Which is 100 times more than 45nA(specsheet input current biassing), which is why I thought it would be ok. Is there a usefull simulator?

In your schematic, the left op-amp is used to buffer the voltage produced by a pot, and is used to bias the right hand op-amp. The left op-amp circuit is not needed. Replace it with two 10K resistor between Vcc and gnd. Their junction point will be 1/2 of your Vcc. Use that to bias the right op-amp. You have been shown examples of this by several posters. Please look at the circuits shown to you. They are all good ways to do what you want to do.

The reason i had the left op amp there is because I thought it would result in lower power consumption, having just one high resistance voltage divider and a voltage buffer.

 

Thanks, as I said I am new and learning a lot. I thought 5V/1m = 5000na right? Which is 100 times more than 45nA(specsheet input current biassing), which is why I thought it would be ok. Is there a usefull simulator?

The reason i had the left op amp there is because I thought it would result in lower power consumption, having just one high resistance voltage divider and a voltage buffer.

The max input bias current for the LM358 is 500nA. To get an accurate Vcc/2, the current out of the tap on a voltage divider should be less than 1% of the current down the divider, so the total resistance should be 5/(100*500n) = 0.1meg, or 50K above and 50K below the tap.

If this goes into an Arduino that draws like 45mA, why worry about another mA?

Another question: If the bias divider drives an input pin of a LM358, does it matter if it drives a voltage follower or the amplifer itself? You would only win if you really needed a much lower impedance Vdd/2 reference...

 

@ MikeML-That circuit you posted gave me a good idea for something I'm working on (involves that 5V Arduino input problem).

 

I built the circuit, with two 10k voltage dividers one for the non inverting input and one for the inverting input to offset the voltage.

The problem I found is that my 10k resistors are different values, they are only 5% accurate resulting in a variation in offset. The gain is 6 because I am using a 100k feedback pot resistor. And because of this the resting position measured by the arduino floats around 2.2V-2.4V.

 

Is there a usefull simulator?

I have used multisim because that is what we were using in school but LTspice works pretty well but that is just my opinion (there are other spice simulators out there). Oh I forgot it's free so have at it