Simple Phaser – Experimenting with All-Pass Filter

This is getting a bit complex for a simple Phaser 😀. After a long way coming back to the original schematic, it’s time to do some experimentation.

In this article I cover what might be useful to try out modding, I do some SPICE simulation and then I breadboard another version of the phaser so I can try my mods side-by-side with the original circuit.

As usual, you can jump straight to the video and come back here for diagrams if you wish.

What to Try Out?

Here’s a list of possible things to try out in terms of just choosing different component values. In a later video I’ll cover LFO mods, but for now, besides different JFETs which I covered in the previous post, we could try out some different values for Caps and Resistors. Let’s see what can we try changing in the schematic to play with all-pass filters:

As I marked above, we could play with: C2/C3, R6/R24/R9/R25 and possibly C11/C12. We could try out R4/R5 & R7/R8 but it’s a bit tricky because we want all pass filter to have unity gain. We’re mixing phase shifted signal with the input sound and we want them to cancel out each other – hence unity gain.

If I translate this to SPICE diagram, I could possibly visualize what each change does. Having said that, there’s quite a bit of options so if you wish to play with SPICE I added some comments to the file.

Turns out that changing C11 and C12 won’t make any difference, and there’s little point using different values for R6 and R24 for example. If I’m going to change R6, R24, R9 and R25 I’m going to use the same values for all 4 of them.

SPICE

Here’s a bit of what I did in the video.

What I did there was to remove completely LFO. Instead, I used V3 voltage source as my LFO. This way I can control my “control” signal. There are lots of different parameters. For example I used {rtest} to test changing resistance for 4 resistors in the diagram above.

In the diagram above, LFO signal is going from 3.625V up to 3.675V in increments of 25mV. And I also run a list of capacitance values for {ctest} – 22n, 47n and 68n. All together we have two sweeps and this is the diagram I get when I run the above SPICE simulation without any changes:

As you can see above (if you try really hard 😊) there are notches for 3 different control voltages (3.625, 3.650 and 3.675) across 3 different capacitance values – 22n, 47n and 68n. That’s 9 diagrams in total.

On the video I go into more detail, and I try out different stuff as well, but I hope you get the picture. Disclaimer as always when using SPICE, real world is not SPICE, this is just an indicator of what the circuit response might be when varying different components. In the end of the day only hearing it is the real measure. This is just to give us some idea and a starting point for experimentation.

Here’s the LTSpice file if you want to try it out yourself:

Using TL074

I breadboarded another version of the circuit using TL074 – a 4 op-amp chip. For all intents and purposes the circuits are equivalent. If you look at the schematic it’s pretty much the same.

Here’s the breadboard diagram I used to prepare the circuit:

Now with this out of the way we can try out getting some different sounds from our phaser.

Video

Conclusion

I’m not quite sure if I can call this a success to be honest. Two stage phaser like this is quite subtle effect, and while there are some audible differences using different components, they are not big by any means. I have a feeling more benefits could be had from manipulating LFO.

Since we’ve seen looking at SPICE (with all words of caution that come attached with it), our useful LFO swing might be just 200mV, or even less. It’s also a bit tricky to get the biasing point just right because of such a small margin, so I’ll be exploring what to do with LFO in the next video. Stay tuned 😉