# How to regulate a signal down to tens of mV?

#### MS&T

Joined Jun 17, 2010
4
I need to attenuate a sine wave down to 10mV peak from 500mV peak. We want to use something that doesn't require adjusting, so we don't have to use an attenuator box. We were going to put schottky diodes to ground, but they have a turn-on voltage of 300mV, which would still be too big.

Any suggestions?

Thanks.

#### retched

Joined Dec 5, 2009
5,208
Resistors?

#### bertus

Joined Apr 5, 2008
21,850
Hello,

What is the output impedance of the source?
What is the input impedance of the circuit to be driven?

With these parameters it is possible to calculate an attenuator.

Bertus

#### bloguetronica

Joined Apr 27, 2007
1,521
I think a simple resistive voltage divider network is your answer. However, it will use some of the source's driving capability. If your load has less impedance than the output impedance of the source or if the voltage divider is somewhat influenced by the load, use an op-amp as a unity gain amplifier (or voltage follower) next to the voltage divider.

A good rule of thumb: the total impedance of the divider network (from source to ground) should be at least 20 times lesser than the input impedance of your load. However, it should be 20 times greater than your source impedance. This will allow an error with a tolerance tighter than 10%.

Then again, as Bertus suggested, what is the output impedance of your signal source and what is the input impedance of your load?

#### Wendy

Joined Mar 24, 2008
22,776
All good suggestions. I would use a variable gain op amp circuit myself, since it would have low impedance out (be careful of loading!), and would isolate the input signal from the output signal.

As Bertus suggested, we need more info to be able to show schematics.

#### bloguetronica

Joined Apr 27, 2007
1,521
All good suggestions. I would use a variable gain op amp circuit myself, since it would have low impedance out (be careful of loading!), and would isolate the input signal from the output signal.
Just out of curiosity: How can you achieve fractional "gains" by using the feedback network of an opamp? By doing this, shouldn't the gain be always greater than one?

#### tom66

Joined May 9, 2009
2,595
Just out of curiosity: How can you achieve fractional "gains" by using the feedback network of an opamp? By doing this, shouldn't the gain be always greater than one?
An inverting op-amp can be wired up for (absolute) gains of less than 1, but can be unstable in such a configuration.

Remember: $$V_{out} = -V_{in}\times\frac{R_{f}}{R_{in}}$$, if you have an Rf ≤ Rin then your gain is between -1 and 0.