Help with this Schematic

[John Lyons]

Hey there.

I can't figure out what +Vb means on this schematic.

http://www.geofex.com/Article_Folders/EQs/parmet.gif

+9 would be the battery positive tab but +vb stumps me. Maybe + volts bias? If it's bias where does this bias originate? I'm trying to build this as a 2 band Parametric eq pedal but there are a couple "missing" pieces in the puzzle.

Also how does the resistor and cap assortment  (mid left of schem) tie in  to the rest of the schematic? Does the +Vb with the two 10K resisitors connect to the other +Vb points as a bias voltage?


Thanks in advance

John

[Peter Snowberg]

You are correct about the +Vb being the bias voltage.

That network of two resistors and the cap with the ---> +Vb marked with the arrow is the bias generator. It works by dividing the rail voltages in half with a simple resistive divider, and then the cap filters out noise.

Connect that +Vb point to all the other places in the circuit that have +Vb next to them and there you go.

When you run opamps on split voltage (a + and a - supply), the bias voltage is almost always the ground. When you use a single ended supply, you can't use ground anymore so you have to make a new reference point that's half way between power and ground.... thus the divider and cap.

I hope that helps.

Take care,
-Peter

[smoguzbenjamin]

Ok, except for maybe a little bug that there's a new topic, here's the deal

You use the Vb to bias the opamps. Hence the name 'bias voltage' What it does is float the signal on 4.5vdc so that the opamp can 'swing' its output in 2 directions. If the signal is referenced to ground, the opamp can't go below 0v and it won't put out the lower half of the waveform.

The two resistors are acting as a voltage divider, to split the voltage into 50% of the supply. The cap filters the bias voltage out and makes it stable.

As far as the schematic goes, connect all the points marked Vb together.

[smoguzbenjamin]

Dang, peter beat me to it :mrgreen:

[Peter Snowberg]

You had a good complimentary answer Ben. No worries mate.

Take care,
-Peter

[John Lyons]

Ok so far I get it. I'll get busy on this and see what I come up with.

Thanks a million.

John

[David]

Peter:

There's an easier way, remember?  R.G.'s trick of using a 386 with no inputs connected is a cheap, cool, small and easy way to create bias.  Works great.

[smoguzbenjamin]

Oh yeah! Forgot about the 386 trick, I gotta try that sometime.
You're welcome John

[Peter Snowberg]

The 386 trick is a neat one when you want (or need) the impedance to be really low, but for a lot of circuits it won't matter very much and the current draw may be 10 times what a pair of 10K resistors draws. If you use the Ohms law I=V/R and solve for 9 volts and 20K, you get I=9/20000 or I=.00045A or I=0.45mA. The idle current for the 386 is much higher, but with the lower output impedance and the 386's active output drive, the 386 a sure winner in terms of stability. The resistive divider falls down when you have any noise on the ground rail because of the way it's filtered.

Either will work fine in this application, but if you have a whole bunch of bias connections, it shouldn't take too long before the 386 is the clear winner in terms of stability.

TI has some rail-splitter chips (or is there just one?) that do the same job with a super low current draw, but with a big price tag. There's not much free lunch to be had anywhere.

Thanks for reminding us about the 386 trick David.

Take care,
-Peter