# Sallen-Key Filter problem

Discussion in 'General Electronics Discussion' started by Yoa01, Oct 22, 2012.

1. ### Yoa01

214
0
Jun 18, 2012
Hi All,

I feel like I'm asking too many questions about my little synth projects -- just tell me when I should shut up

So I took a standard Sallen-Key filter design and modified it slightly (take the one you find on, say, Wikipedia, and multiply resistors and caps by 10). I also made the resistors pots.

My understanding says this should act like a low-pass filter, not a capacitor-driven waveshaper. Here is a chart of three waveforms: the output of the 555 oscillator, the filter fully closed, and the filter wide open.

Note that I have pin 3 of the 555 going into this filter.

Can someone explain to me how a filter is acting like a really neat square-saw waveshaper?

2. ### BobK

7,682
1,688
Jan 5, 2010
Those waveforms look like a high-pass filter to me.

Edited to add: And filters DO re-shape waveforms.

Bob

3. ### Yoa01

214
0
Jun 18, 2012
Prehaps I'm too used to synthesis. I went into this thinking that, as a low-pass filter, it would attenuate higher frequencies over its range, resulting in a sine wave at the most resistance.

But, I think I know what went wrong.

This is my design, as stated. Note, however, that the upper capacitor is not polarised. This lead me to think the signal was going into that then the output, totally ignoring the OpAmp. So, I tested it:

The top is the 555 output, the bottom is that output through one 10nF capacitor. Same waveforms as with the full setup.

So, not having a polarised 10nF capacitor, could I put a diode in front of the capacitor to make it work right?

4. ### BobK

7,682
1,688
Jan 5, 2010
The problem taking the output through a 10n capacitor! This makes a high pass filter. What is the load? If you are driving a speaker, the capacitor should be more like 1000uF. Here is my simulation of your circuit with a 100Hz square wave going in, and simply measuring the output of the opamp with no load and no capacitor:

File size:
68.8 KB
Views:
263
5. ### Yoa01

214
0
Jun 18, 2012
I really need a better understanding of what I'm doing, I take it. haha!

I have the output going to a jack, into a 10' Monster instrument cable, into a mixer, then into my computer. I have no idea what the load would be because I honestly can't find the proper specs to calulate it.

May I ask what program you're using there? It looks rather useful.

Thanks!

6. ### Harald KappModeratorModerator

11,651
2,692
Nov 17, 2011
Looks like LTSpice.

It is free. Only drawback is that the libraries are rather basic. But it is easy to incorporate models for other components (supplied by many vendors) into LTSpice.

7. ### BobK

7,682
1,688
Jan 5, 2010
Yep, that was done on LTSPICE. It took me about 10 minutes. I love my LTSPICE!

Bob

8. ### Yoa01

214
0
Jun 18, 2012
Huh. I use (now) Tina-9 due to a user suggestion on here. Is Spice better? I see a lot of Spice schematics available, as well as components.

I also have DipTrace, which is just easier to use than Tina.

9. ### Harald KappModeratorModerator

11,651
2,692
Nov 17, 2011
As far as I can see Tina's core uses SPICE, too. Both TINA and LTSpice (as many other analog simulators) are just different SPICE implementations with their own user interface.
Plus some add-ons as there are
- libraries
- special simulation models (LTSPice is strongly oriented towards simulating switch mode power supplies, hence the alternate naem SwitcherCAD for this tool)
- possibly integrated PCB layout
- possibly integrated digital simulator (LTSPICE includes a rudimentary digital simulator) for mixed mode simulation

If you're happy with TINA I see no need to switch. If your not so happy, give LTSpice a chance.