Clean Blend Question

[bufferz]

Before I start designing a pcb for this i wanted to make sure I did this right.

I am implimenting a clean blend into a modified ts9 type circuit with some ideas borrowed from the paisley drive in the tone circuit.



cheers

[bufferz]

just noticed 2 errors, "load" pot should be 1m, not 100m and gain should be 250k

cheers

[psychedelicfish]

I would recommend sticking a buffer in there, otherwise you'll get feedback of some kind. Maybe you could use an amplifier instead, and rather than a blend control, you could have a clean volume and a distorted volume control.

[bufferz]

I intend for the tl072 to be a buffer, just before the split

[psychedelicfish]

what I meant was a buffer/amplifier between c15 and r28

[bufferz]

I could use the other side of the tlo72

[psychedelicfish]

That would be good. What I was saying earlier was that you could make the buffer between c15 and r28 an amplifier, and rather than a blend control, you could have a clean level control, which would give you the same flexibility as a blend knob, and you wouldn't have a mismatch in signal levels.

[Bill Mountain]

I would think the blend pot would need to be a higher value than the volume pot.  The series resistance from the volume pot will play a part in the blend pots functionality.

[bufferz]

Thanks to everyone who has chimed in.

I have attached two revisions based on what I am hearing.

Blend control:


seperate clean and effected controls:


Please let me know if I mis-applied the advice that was given,

cheers

[bufferz]

I would think the blend pot would need to be a higher value than the volume pot.  The series resistance from the volume pot will play a part in the blend pots functionality.

Good point. a 1uf cap before the blend pot should be added.

[bufferz]

That would be good. What I was saying earlier was that you could make the buffer between c15 and r28 an amplifier, and rather than a blend control, you could have a clean level control, which would give you the same flexibility as a blend knob, and you wouldn't have a mismatch in signal levels.

Thank, if I were do choose this method, woud i just send signals to be summed at the output? Also, would they be in phase?

[Bill Mountain]

I would think the blend pot would need to be a higher value than the volume pot.  The series resistance from the volume pot will play a part in the blend pots functionality.

Good point. a 1uf cap before the blend pot should be added.

That won't make a difference.  I need to look at your design this weekend when I get some free time.

I think you're making it harder on yourself than you need to.

Have you looked at the schematics for other TS's with clean blends?

[bufferz]

Yes, you are prob right. Thank you for your help and patience

I wanted to avoid the sparkle drive dual gang method and I was using these as a reference:

http://www.sabrotone.com/?p=1003

http://www.jmkpcbs.com/JMK_PCBs/DIY_Projects_files/Buffer-Splitter.pdf

http://www.beavisaudio.com/techpages/blocksfragments/

It seems as though they do not require the additonal buffer / booster that is being refernced as these are supposed to be able to work as standalone units with other FX. Perhaps I am just incorporating it wrongly.

(i dont belive i have permission to add the schems here, but the links are to pages easily accessable via google)

[Bill Mountain]

DOD made a bass distortion based on a tube screamer that had a simple clean blend implementation.  The Sparkle Drive is more complicated but that is how proper active mixing is done.  There is no chance for crosstalk and various resistance problems you can run into with passive blending.  You can also look at the Runoffgroove blend pedal.  That design is simple-ish but quite effective because it doesn't suffer from some of the issues that may or may not be present in your design.

Here's the DOD pedal's schematic:

http://www.sugardas.lt/~igoramps/article16/fx91.gif

[bufferz]

Vey cool.

Interesting, seems similar to the others, however the dod does the split pre-tonestack. I have updated my schem based on the DOD method. It still seems to have the series resistance prior to the blend pot so I imagine if the dod was funtional, it is not a huge concern:

[Bill Mountain]

I don't know why they picked the values they did for the resistors before the blending but with the 100k blend pot being 10x's and 100x's higher than the resistors I doubt you would notice anything decreased functionality.  I have also seen examples of putting a resistor to ground right before each side of the blend pot.  Something like 100k would work.  I do not know the methodology (it might be related to something in here: http://www.geofex.com/Article_Folders/panner.pdf ) but it seemed to improve the blend function by having less of the other signal present when the blend pot is turned to the extreme (a problem with passive blending).