# Does this Sallen and Key filter work on a single supply and does the input offset matter?

This is the schematic for a Sallen and Key from the Doepfer Musikelektronik website: simulate this circuit – Schematic created using CircuitLab

I intend to run this in a synth, the oscillators use a single 12V supply and have an amplitude of +6V (I don’t know what the offset from Ground is)

Questions:

1. Can this be run from a single supply, connecting V- on the op amp to the same Ground as the capacitor?

2. And does an input signal need an offset from ground for the filter to operate correctly?

• How well will those 3 opto-isolators track? – analogsystemsrf Apr 21 '19 at 15:33
• I’m not sure I understand, but this is not a precision application – Callum Apr 21 '19 at 18:04

The answer to both questions is yes. For a more complete explanation you need to redraw the circuit with component designations.

Continuing my answer using the new schematic. Although one usually sees single supply circuits biased at one half V+, the ideal point is actually midway between the output swing of the amps. A 324 typically swings between .5 and 10.5 volts, so 5.5V would be best ; which is not that much different than 6V. In this application one can make the source oscillate around the 5.5V point, so all that this circuit needs to be single supply compatible is a capacitor in series with R3. Then when you change the gain ( and Q), by varing R4, the output remains centered around 5.5V.

Making the resistors independently varible creates a huge range of issues that are too big to cover here. The best way to do this is to keep R1 equal to R2( roughly) and vary them together to change the center frequency of the filter. I think only Qs of more than one will be useful here. So if R3 were a more reasonable 10K ( and connected to ground through a 1 uF cap), then varing R4 from 10K to 20K would change Q from one to infinity ( at which point the filter becomes an oscillator.) Also keep in mind that changing the Q this way also changes the amplitude. R4 is optocupler 3.

• Done, the opto isolator do not use phototransistors, but rather photoresistors. – Callum Apr 21 '19 at 6:19
• Thank you. This circuit is so different that I will need a lot of time to think how to word an answer. R3 is far too small for one. I don't have time now to do that and by tomorrow others will have answered. – EinarA Apr 21 '19 at 6:31
• Yes, there is a multimode filter on the same page which I am using as a value reference, however it does not include that resistor. doepfer.de/a1011_tec.htm – Callum Apr 21 '19 at 6:59
• I have updated my answer. – EinarA Apr 23 '19 at 2:43

Yes and yes.

Yes, almost any opamp can operate with a single power supply voltage; but some are way better at it than others. Some opamps require a minimum voltage between the two power pins of 10 V (+10 V, -10 V, +/-5 V, whatever), so single supply operation is possible, but not at 6 V. Also, be sure the check the datasheet for the input common mode voltage range and output voltage range. These can reduce signal voltage range by 6 V peak-to-peak or more. This could mean that an opamp operating on +10 V and GND has a max output voltage of only 4 or 5 Vp-p.

Yes, the input signal needs to be centered within the input common mode voltage range. a good starting assumption is that the midpoint is also the power supply midpoint, or +5.0 V if the part is running on 10 V and GND. This means you need an input coupling capacitor and a high impedance DC voltage to pull the non-inverting input to +5 V when there is no input signal. The impedance of the bias voltage should be at least 10 x the value of R so it has minimal effect on the filter characteristics. 100x would be better.

• I edited the post, the supply is 12V, the amplitude of the waveform is 6V. I am using LM324 op amps which are single supply, and will add an amplifier after it. – Callum Apr 21 '19 at 4:30