Phase inverting at input buffer vs after effect

X

xconverge

Member
I am working on a custom version (for personal use) of a blend for 2 effects based on the https://jmkpcbs.com/product/paralyzer/ that includes the ability to invert the phase of one of the signals, and also incorporates the 3DPT and led all due to space constraints. My question is, is there any issue with just swapping the phase at the input buffer like I have done here, or should I swap IC1 to a TL072 and just add a 2nd unity gain circuit after R2-T return from the 2nd effect that is switched. What I have done here seems totally fine to me

1733158602321.png
 
I think I am going to switch IC1 to a TL072 and add an inverting stage at R2-T that gets skipped or not with the DPDT, seems a bit simpler and only adds 2 resistors. I think in my diagram above R13 would cause issues when the DPDT switch is set to not-inverting. I also will have no doubts or questions if I do this (y)
 
A topic of great interest to me.

Bookmarked for future reference (will be continuing a similar project soon).
 
Last edited:
I have 5 boards on the way so we will see!

The traces and layout are SUPER sloppy but its all connected and passed design rule checks so YOLO

1733239853866.png

Pretty obvious that I am going to just slap 2 builds I have already done into a 1590xx with this

1733240325427.png

I briefly tested a parallel mixer pedal, and while functional, the sheer number of patch cables and power cables grossed me out https://www.sonicake.com/products/portal

I may not even populate the DPDT slide switch and will just use jumper wire or see if I have any 2 or 4 pin jumpers to put there instead
 
Last edited:
giovanni said:
I asked this same exact question a couple years ago and was advised to do what you did here. The first option, iirc, has a much lower input impedance in the inverting configuration which may be undesirable.

Edit: here’s the thread I was referring to:
Thread 'Why use 2 op-amps when you only need 1?'
https://forum.pedalpcb.com/threads/why-use-2-op-amps-when-you-only-need-1.10794/
Click to expand...
Thanks for the links! Even in my schematics, the 2nd option ended up being simpler to me, especially since I already had the IC (socket) sitting there waiting for it to be used. Glad to have some other answers/reasons though, so thanks!

I probably could have added a coupling capacitor to selectable the inverting buffer but I should be fine without it I would think
 
Last edited:
I dont think I will have any issues without a coupling capacitor, but if I wanted to I could use a cap as my jumper between pins 5/6 of the DPDT :LOL:
 
jesuscrisp said:
Oh you will totally have issues without coupling caps around the inverting stage...
Click to expand...
Ok, I will use a 1uf as my jumper then (output side) of the inverting stage since I am not actually populating a DPDT, does that seem ok?

What DC are we worried about in the unity gain inverting opamp buffer/phase invert section, the -VB input?

I think the input side is fine because the output of the effect has a coupling cap already (B-RETURN)
 
Last edited:
Your main issue will be misbiasing as you'll have resistors and pots on either end of the op amp going to ground (on the input only if you put another pedal in that lool) but your inverter is biased to virtual ground.
 
jesuscrisp said:
Your main issue will be misbiasing as you'll have resistors and pots on either end of the op amp going to ground (on the input only if you put another pedal in that lool) but your inverter is biased to virtual ground.
Click to expand...
Any recommendations on values? I was going to use a 1uf on the output of the inverting buffer and a 100nf on the input side
 
Not sure why you're using an inverting configuration for the input buffer(s) (IC2). As shown on the schematic in post #3, your input Z is 50k (R13||R14), which is way too low if you're going to plug a guitar straight into BLEND_IN.

Also, the buffer noise performance is significantly degraded (by more than 10dB) versus using the same TL072 in a non-inverting topology.
 
JTEX said:
Not sure why you're using an inverting configuration for the input buffer(s) (IC2). As shown on the schematic in post #3, your input Z is 50k (R13||R14), which is way too low if you're going to plug a guitar straight into BLEND_IN.

Also, the buffer noise performance is significantly degraded (by more than 10dB) versus using the same TL072 in a non-inverting topology.
Click to expand...
There will be at least 1 buffer between the guitar and this pedal, but noted! Should I change R9, R10, R13, R14 to 1M just in case?
 
I will make that change if I need a respin for other reasons that I can't rework. Thanks! I didn't put much thought into that portion since I was just referencing 1:1 from the jmk paralyzer.
 
Well, depending on what comes before, the extra noise may or may not make a practical difference. If it comes after some gainy pedals, then there might already be enough noise baked into the input signal that a bit more doesn't matter.

But it's good to know that non-inverting is always less noisy than inverting, at the same gain. The biggest difference is actually in unity gain circuits. And any resistor in series with the signal adds noise, so inverting circuits really have a disadvantage if you need a higher input impedance.
 
JTEX said:
Well, depending on what comes before, the extra noise may or may not make a practical difference. If it comes after some gainy pedals, then there might already be enough noise baked into the input signal that a bit more doesn't matter.

But it's good to know that non-inverting is always less noisy than inverting, at the same gain. The biggest difference is actually in unity gain circuits. And any resistor in series with the signal adds noise, so inverting circuits really have a disadvantage if you need a higher input impedance.
Click to expand...
Yea both effects will be quite gainy so I am not concerned about it, but I really appreciate the insight. I should do more simulating but people here are such a wealth of knowledge too!
 
For my ears, I haven't added the coupling caps, and the noise level on everything is totally acceptable. I may play with adding some later after the other stuff below is resolved.

The build came together nicely from high level standpoint, fits EXACTLY, I don't think I will need to do a rev 2 of the board, but if I did I would adjust the hole sizing to be a bit larger for the DPDT, probably switch to ovals (I used circles in case I messed up and needed to rotate the 3DPT by 90 degrees), decrease dimensions of the board by a mm or two, and add coupling capacitors for the phase invert stage.

I did have a major problem which is why I didn't post pictures yet, I used a bit too much solder on the 3DPT lugs, and it caused a bridge on the center most pin (prettttty unlucky) which was always grounding the output :( I was able to get the 3DPT off, the output pad lifted, I dont have a spare 3DPT, so for now I used jumper wire to take the place of the 3DPT. I already confirmed bypass was working.

All sounds great, when I flip the phase vs not with jumpers, I can hear the phase cancellation very clearly. The blend works great. So all in all I am happy, just waiting on a few parts to rebuild a 2nd board. For now I will use it as an always on until the tayda order comes in!

Thanks for the help and tips everyone, glad I won't be sinking tons of more time and money into getting this one right and can get back to playing!
 
Last edited:
You must log in or register to reply here.

Similar threads

J
TUTORIAL Why Your Clean Blends Suck ... or "How to (Clean) Blend"
Replies
8
Views
3K
jrhevron
J
Back
Top