op-amp bias resistor?

[blackieNYC]

The runoff groove Tri-Vibe has a 1Meg resistor delivering the bias voltage to the non-inverting input of the input op amp.  All other points where Vbias is applied have series resistors of 470 ohms (that's the lowest value I've ever seen), or 100K. but the output stage op amp non-inverting input just goes straight to Vbias.  With no resistor.
I've read a nice article at Geofex about calculating values for design purposes, but this (pin 5 on U1b) doesn't seem right.  I feel like I should stick a 100K there.  What do you think?
http://www.runoffgroove.com/tri-vibe.html

[R.G.]

1. How much current flows through that resistor into the biased inputs?
2. How much offset voltage does that input current flowing through the input resistor cause on the biased inputs?
3. How much thermal noise does the bias resistor contribute to the input?
4. How much DC gain does the biased stage have?
5. How much current can the bias voltage source source/sink without changing it (and all the other biased points too)?

Datasheet time.

[TejfolvonDanone]

All other points where Vbias is applied have series resistors of 470 ohms (that's the lowest value I've ever seen)

U4 is an LM13600 which is not really an op-amp but an operational transconductance amplifiers. Which works differently. If you look at this datasheet: http://users.ece.gatech.edu/~lanterma/sdiy/datasheets/ota/lm13600.pdf and turn to page 8 you will see a recommendation for  a 1k pot for biasing. In the text you can find that they also say that two 510 Ohm resistors can be good instead of the 1k pot. In the tri-vibe it is biased to Vref not ground.

but the output stage op amp non-inverting input just goes straight to Vbias.  With no resistor.

For bias currents in an inverting op-amp configuration you have to put a resistor between Vref and +In equal the DC resistance seen by the -In. In this case it is a 47k paralel with a 47k or 100k depending on the switch position. For a textbook result it should be 23.5k or 32k.
From a DC point it is a stage with -1 or -0.47 gain. So the bias currents and offset voltages don't really make a difference that's why +in goes straight to Vref.

[samhay]

>For bias currents in an inverting op-amp configuration you have to might consider putting a resistor between Vref....

As TejfolvonDanone said/implied, this resistor helps offset DC offset errors on the output, which is the topic of another current thread. It gets tricky if the resistance in the feedback loop changes - e.g. if there is a switch or gain pot there - and you can sometimes/often/perhaps/maybe get away with omitting this resistor and directly connect the [Edit - sorry (+) input] to your (virtual) ground, particularly if the output is AC coupled to the next state/output.

[merlinb]

The runoff groove Tri-Vibe has a 1Meg resistor delivering the bias voltage to the non-inverting input of the input op amp.  All other points where Vbias is applied have series resistors of 470 ohms (that's the lowest value I've ever seen), or 100K. but the output stage op amp non-inverting input just goes straight to Vbias.  With no resistor.
I've read a nice article at Geofex about calculating values for design purposes, but this (pin 5 on U1b) doesn't seem right.  I feel like I should stick a 100K there.  What do you think?
http://www.runoffgroove.com/tri-vibe.html

The opamp doesn't need a resistor to receive bias. It just wants the bias voltage, yum yum. You only put in a resistor when something else needs it.

For example, at the input you can't connect the non-inverting input straight to Vref because then the incoming guitar signal would disappear into Vref! You need a resistor, and the guitar likes to see about 1Meg.

The output opamp, on the other hand, has nothing else going on at the non-inverting input. It's not used for anything else, so you can feed Vref straight into it, no muss no fuss, yum yum.

The intermediate stages are operational transconductance amps. Here the 470R resistors actually form part of potential dividers (together with the 10k resistors) to reduce the signal level going into the chips to prevent clipping. The non-inverting inputs are given matching resistors to improve DC offset, but they probably aren't really necessary -a bit of offset would probably pass unnoticed. These weird chips also have an extra 3k9 resistor which is supposed to reduce distortion, but in reality is fairly useless.