Dirtbag Phase Reversal?

[eldanko]

I built a Dirtbag with MN3005s and a Cusack clickless bypass system.  (Build report here: http://www.madbeanpedals.com/forum/index.php?topic=3673.0) I was running a stereo rig the other night, with the Dirtbag running on the right side of the stereo chain and my Deep Blue Delay on the left.

When the Dirtbag is engaged, the signal phase seems to be reversed, putting the amps out of phase.... which is both cool and uncool.

Can anyone explain this to me?  Is it inherent in the effect, or caused by something in the bypass system?  Anyone else experience this?

[madbean]

Yeah, it's inverted output from the DB that is causing this.

[eldanko]

Ahhhhh.... 

[madbean]

You could fix this with a little additional circuitry. Since space is tight in the DB, you might want to do a dedicated unit for correcting phase issues if you are doing a lot of stuff in stereo.

I could draw it up for you in like 10 minutes if it's something you want to do.

[eldanko]

If you've got the time, I'd love to explore that...

[jimmybjj]

Here is a article about Polarity Reversing by RG, includes a schematic. The ROG splitter-blend includes one if you looking for another excuse to build a pedal. I had built and its fun to play with.

http://geofex.com/Article_Folders/polarity_reverser/polarity_reverser.htm

http://www.runoffgroove.com/splitter-blend.html

[eldanko]

I like this splitter blend.  Adding this to my list of to-builds. 

[madbean]

Here ya go. Eagle files are in the zip if you want to make a little PCB. If you are building in the same box as your delay, you can leave off C5 and D1. R5 can be increased to 100k, if needed.

__________________________________________
BTW: you could also use this to make a simple booster. Lower R1 to 10k, use a 1M pot for R2 wired as a variable resistor and increase C2 to 47 -100pF. It will sound really nice.

Extra credit: make R1 3k3, and use two back to back 1n914s in place of C2 and put C2 on the Gain pot. Now you gots an overdrive.

[jimmybjj]

Here ya go. Eagle files are in the zip if you want to make a little PCB. If you are building in the same box as your delay, you can leave off C5 and D1.

.....yeah you could do that if you want to do the simiplest and most logical thing duh... .  A simple opamp stage will flip the signal to where you need it.

A couple quick questions, just so I have a better understanding. What is the purpose of lowering R1, filter related? And Is the suggestion to possibly change R5 just to change the output impedance and/or to change frequency response.

[madbean]

Here's an explanation of how this circuit works:

C1 decouples the input of the circuit from the AC source. It removes the AC component and leaves the DC, which is what our circuits like to work with. The "+" end goes toward pin2 because that is the highest voltage. We know that because the op-amp is biased to about 4.5v via the voltage divider created by R3 and R4.

R1 does two things: sets the overall input impedance (along with C1 to some degree, I think) and works in tandem with R2 to set the overall output on pin1. For an inverted input like this one, the output gain is R2/R1.

To make it as close to a 1:1 ratio on input and output, we make R1=R2. This is in an ideal op-amp but it's not exact in the real world. But, close enough.

Lowering R1 for making it into a boost simply increases the output (higher gain) of the amp. Since the total gain for inverted input like this one is R2/R1, the output gain is 100 if R1 is 10k and R2 is (up to) 1M. Raising R2 to 1M means you also have to increase C2 to re-adjust the corner frequency of the filter created by R2and C2. 100pF might be too high, but 47pF is a good value. Even 22pF. It's a soft-corner, so your ear may or may not hear the difference in adjacent values. Too high, and you will for sure notice the treble loss.

Also, the formula for calculating corner frequency here as with everything else: 1/(2*PI*R*C), where R=R2 and C=C2 in this case.

((((((((((  F(c) = 1/(2*Pi*R*C) is the MOST important formula to know by heart (other than Ohms) in circuit design and analysis---learn it, live it, love it

It works the same for LP and HP filters. The slope (for a single pole) is -6dB per octave.)))))))))))

C3 decouples the output of the circuit. We have to do this because the bias voltage that makes it operate is about 4.5v. We can't have 4.5vDC coming out our circuit, so decoupling it adds the AC portion back in and effectively makes the output 0v.

R5 sets the output impedance, so yes, increasing it will increase the output imp. 10k might be enough in a lot of cases, but 100k is also a good all around value.



----------

It's early, so hopefully I haven't made any embarrassing gaffes in the breakdown