Question about voltage dividers for input biasing

[twoheadedfetus]

On all the schematics I look at input biasing is done this way. My question is, if the desired voltage of 4.5V has been achieved using resistors R33 and R32, whats the purpose of R5 and R9?

[ashcat_lt]

Well, for AC signals, that 4V5 point is the same as ground.  Without those resistors, you'd be shorting the audio signal that you're trying to pass.

[Groovenut]

Without R5 and R9 there would be no bias voltage for the opamp. R5 and R9 function two fold. They provide the DC bias voltage to the non-inverting input of the opamps and provide a ground reference for the AC signal voltage. The "centered" bias voltage is needed for the AC signal to swing about.

[PRR]

> whats the purpose of R5 and R9?

Literally: the opamps don't know about the R33 R32 node. R5 R9 tell the opamps to "follow this".

That plan (which is hardly "all schematics) *does* have a flub. R9 is pointless; U1 will follow U2 DC through R8.

[antonis]

You may see R5 & R9 (pointless indeed) with much bigger values (like 470k - 1M) in some other circuits..

Their specific values depend on previously set resistor (and/or capacitor) values (like C2/RA & R8), forming voltage dividers hence attenuating non-inverting input signal..

[twoheadedfetus]

how I understand it on the right is how its done in the tube screamer schematic I originally posted, on the left is how I did it on my circuit, and it works perfectly fine, am I wrong with how I understand it and is that how its actually done on the tube screamer schematic too? I dont get it

[diffeq]

how I understand it on the right is how its done in the tube screamer schematic I originally posted, on the left is how I did it on my circuit, and it works perfectly fine, am I wrong with how I understand it and is that how its actually done on the tube screamer schematic too? I dont get it

Both ways of biasing work, but the first schematic has lower impedance (half the value of two resistors used, two 1M make up for 500k input impedance) and somewhat higher thermal noise due to higher overall resistance of the network.

[amptramp]

how I understand it on the right is how its done in the tube screamer schematic I originally posted, on the left is how I did it on my circuit, and it works perfectly fine, am I wrong with how I understand it and is that how its actually done on the tube screamer schematic too? I dont get it

The version on the right allows you to add a capacitor from the junction of the three resistors to ground.  This is sometimes called noiseless biasing because it will isolate the input from any noise on the power line.

[twoheadedfetus]

Thanks for clarifying fellas

[antonis]

Just to add to both Ron & Tony said, lower divider resistors values form "stiffer" bias source..

[GibsonGM]

Just to add to both Ron & Tony said, lower divider resistors values form "stiffer" bias source..

...but they also lower the input impedance, so it is a trade-off, as with most things.

[antonis]

...but they also lower the input impedance,

Definately not in case of not directly signal coupled..

Anyway, we shouldn't bother a lot for input impedance if we only knew it's specific value..!!
(quite statisfactory statement for an edifying argument, don't you agree Mike..??)

[GibsonGM]

I am of the understanding that the opamp's input impedance must be put in parallel with that Vb resistor...is that not correct?   So typical 1M || 1M Vbias would give 500k, a VERY nice input Z, for sure.    But one could use lower values at Vb and make that lower.   And add noise. 

Or not, please do explain if say, R5 is not taken in parallel with input Z?   This only matters if DC coupling???   

[antonis]

 
I've just wanted to tease you a bit, but you know confused me..

A pair of 1M directly coupled bias resistors should add more noise than a pair of 1k in series with a 500k coupled to input..
(due to noise implementation proportionality to current flowing..)

Of course, current flowing through 1k resistors is 500 times greater that that through 500k resistors but what we actually mind is current flowing through final bias resistor..

[GibsonGM]

Oh, sure, Antonis!    I was not clear - it is the LAST resistor, Vb, that I am talking about.    It is good to have a "stiff current flow" in the divider, of course!   

And then find a good 'balance' for that last resistor.     I am sure Two Headed understands what we are talking about!    Unless we are going to draw it out, we should leave him to think about the early part of the topic, so now he knows what is going on, before we get him turned around!   (Or ourselves...) 

[twoheadedfetus]

Id like to ask some stuff actually. The impedance for what current exactly is half of one of the resistors used(looks like a current that passes through both of those in parrallel)? Can someone maybe point me in the direction of an article that explains all the benefits of input biasing like this for an absolute beginner, and the best ways to choose resistor and capacitor values. The way I understood it now, plus with some reading online, the third input resistors added to the divider are supposed to limit current flow from the power source?

[antonis]

Let's get it backwards..

(Ideal)Op-Amp inputs draw no current so there isn't any current flow from voltage divider, through bias resistor, to inputs hence no voltage drop across it..!!
Bias resistor "carries" the exact divider voltage to inputs..!!
(I know it seems like some kind of magic but let it be for the moment..)

So, the only current that matters is the one flowing through voltage divider resistors (which forms the desirable voltage on resistor junction with a help from power supply and GND..)

For input bias other than that of incoming signal, that resistor can be omited..!!
For same with incoming signal input bias, it should be included to "isolate" the -usually low- impedance voltage divider with its high value..!!

The low enough value (1k, say) bias resistors seen for bias no signal inputs are used for "equalizing" input bias current offset..
(but we know deviate from "ideal" op-amp and will discuss it later on..)

[twoheadedfetus]

I know how voltage dividers work and I understand that there shouldnt be a voltage drop across the bias resistor, but I dont think i exactly understand the purpose of the bias resistor itself. Increasing impedance? But for what? the input signal or noise from the power supply?

[antonis]

I dont think i exactly understand the purpose of the bias resistor itself. Increasing impedance? But for what? the input signal or noise from the power supply?

Impedance only matters when "seen" from signal point of view..
So yes, it increases impedance when it serves as bias resistor for signal input..

It could be omitted for a high value voltage divider(without cap, of course) if only divider was "quiet"..
But that kind of divider is noisy so we make it more quiet lowering its resistor values and isolate those low values resistor with a big bias one..

Big value bias resistor is quiet 'cause there isn't any current flowing through it..

[twoheadedfetus]

But how does the bias resistor increase impedance for the input? And why is bigger impedance good?