[sdiy] Single supply op-amp filter biasing question

[Tom Wiltshire]

Thanks both, that's exactly what I needed to know.

To answer Roman's question; A typical two op-amp schmitt trigger/integrator square/triangle stompbox LFO produces a waveform of about this range around the centre level too, so it helps for this to do the same. While I could get a rail-to-rail op-amp to produce a clean output at 0-5V, that wouldn't actually be that useful in most stomp box circuits, and you'd probably finish up level shifting it anyway.

Tom

On 27 Apr 2017, at 21:06, Richie Burnett <rburnett at richieburnett.co.uk> wrote:

> As roman said, consider the filter from a DC perspective... The capacitor impedances go to infinity at DC, so C6 and C7 vanish from the schematic. There is also no current into the inverting input of an ideal op-amp, and therefore no voltage drop across resistor R4 so you can replace that with a wire now too. Now the op-amp circuit *is* a familiar inverting amplifier topology (from the point of view of a DC signal) with R2 and R3 setting the gain and you know the equations for this circuit.
> 
> Hope this helps.
> 
> -Richie,
> 
> Sent from my Xperia SP on O2
> 
> ---- Roman wrote ----
> 
> The bias to reach your desired 2-7V range is pure DC, so you can calculate it just like DC opamp circuit with assumption PWM is at mid level, that is 50%.
> You want 4.5V mean value at the output, while PWM mean value is 2.5V. So there is 2V across R2+R3
> which means that the R2,R3 node and opamp inverting input is higher by 1V than PWM =  3.5V. So 3.5V should also be set at non-inverting input with PR1.
> 
> Wouldn't it be easier just to use Rail-to-Rail opamp and skip the level shifting?
> 
> Roman
> Dnia 27 kwietnia 2017 20:41 Tom Wiltshire <tom at electricdruid.net> napisał(a):
> 
> Hi All,
> 
> I have the following situation:
> 
> A PIC produces a PWM output representing an LFO waveform. This is then sent to an op-amp MFB filter for smoothing. The circuit looks like this:
> 
> http://www.electricdruid.net/images/SingleSupplyLFO.jpg
> 
> If I'm running the PIC on 0-5V, and running the op-amp on 0-9V, the op-amp can't handle the raw PWM output, since it clips below about 1.6V (for the TL07x - other typical op-amps are not terribly dissimiliar). So the PWM signal needs biasing to move it to the correct range.
> 
> I found the following useful document discussing biasing, both AC and DC, for the various op-amp configurations:
> 
> https://ocw.mit.edu/courses/media-arts-and-sciences/mas-836-sensor-technologies-for-interactive-environments-spring-2011/readings/MITMAS_836S11_read02_bias.pdf
> 
> I've used this guide with success on other circuits. However, in this case I can't make theory and practice match up. For a start, the MFB filter isn't a typical DC-coupled inverting op-amp application, although that's the closest. The filter is inverting and has no overall gain, so gain = -1.
> 
> How would I calculate the required bias to get the LFO output signal (0-5V) in the centre of the op-amps range (2-7V)?
> 
> I know the answer to this question from practical experiments, but I'd like to know how I'm supposed to derive it from the theory.
> 
> Thanks,
> Tom
> 
> 
> 
> _______________________________________________
> Synth-diy mailing list
> Synth-diy at synth-diy.org
> http://synth-diy.org/mailman/listinfo/synth-diy
> 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://synth-diy.org/pipermail/synth-diy/attachments/20170427/750ca261/attachment.htm>