Green Mujina

[stonerbox]

I been away for a long time and in the meanwhile this baby badger has entered its adult stage of life. 

Been thinking long and hard about how I want the pedal to behave and at the end of the day sound, and I am fairly positive that I have finally nailed it.

It can do bubbly, fat high gain but also crunchy dynamic saturation. It has been a constant balance to never let it get too flabby, dull or flat but never too sharp & stingy either. I am no fan what so ever of that raw eighties high gain. I strive for great mid range with warmth

Alright, enough of my unimportant rambling, let see what has been changed and added. I will post video/sound samples soon and a brand new layout next week (asap everything is figured out).

0.9v Updates:
1. I rolled back the version of the schematic from 1.7 > 0.9. My thought process was greatly flawed, something can't be over 1.0 until it is a fully functional final version.
2. The Grain is totally reconfigured. It is now a wide parametric band boost with sweepable frequency range. It boost frequencies so that the character of the saturation in Q4 and Q5 alters greatly. It ranges from deep bass up to the mid range (around 700-900Hz).
3. Several gain stage adjustments in order to not over do the saturation, there are however still plenty of it.
4. Swapped Volume 1 from 100K > 50k. Both to limit saturation and increase the fine tuning and smooth out the curve from semi clean > crunch > high gain.
5. Lowered the values of several coupling capacitors greatly to avoid an inferno of bass. No dreary and dull mud shall pass in this build.
6. Added a high self pass filter near the end of the circuit to revitalize the high end. Pushes from and above 4,2Khz if my calculation is correct? 1/(2*PI*R32*C25).

Full sized version of schematic: http://tropicalhippo.com/GreenMujina/Green-Mujina-Full-size-0.9v.jpg


Tone stack response:

[Click for full size]


Help
I would like to optimize and standardize many resistors values. R22-24 could perhaps be recalculated to only two resistors? But how? And if I would change R2 to more common value and make it into a voltage divider how would I do that and still keep the same impedance? I don't want to alter the sound. When the voltage divider is figured out I'll swap C1 so the low pass stays at 10Khz.

Maybe the last two opamps could be merged into one?

Perhaps some coupling caps could be removed too, C18 and C26 maybe? Educate me!

One more thing. Thank you QuietRob, Richard and ElectricDruid for helping me with the oscillating Wien bridge /parametric filter in that other thread.

[stonerbox]

Help
I would like to optimize and standardize many resistors values. R22-24 could perhaps be recalculated to only two resistors? But how? And if I would change R2 to more common value and make it into a voltage divider how would I do that and still keep the same impedance? I don't want to alter the sound. When the voltage divider is figured out I'll swap C1 so the low pass stays at 10Khz.

Maybe the last two opamps could be merged into one?

Perhaps some coupling caps could be removed too, C18 and C26 maybe? Educate me!

Maybe it got buried in the wall of text so I'll echo it.

[robthequiet]

Yes, I'm not clear on the purpose for R24 -- it would seem to drop a lot of signal to ground.

[stonerbox]

Yes it does!  Without R24 or with a lager value the parametric band circuit (on breadboard=long alligator cables left and right) starts to oscillate badly.

[robthequiet]

R2:

This might work, if your supplier stocks it:


http://www.mouser.com/ProductDetail/Xicon/270-715K-RC/?qs=sGAEpiMZZMsPqMdJzcrNwnK90ix8JRCrrDvXfvOeEX4%3d

[stonerbox]

My explanation of R2 was perhaps a bit flimsy. I was hoping that some kind of calculation could be used convert the large and clumpsy resistor value (713k) into a voltage divider that would present the same resistance. Is there a calculation for this?

[robthequiet]

Simple answer, if you have a divider with two resistors, each resistor will drop a voltage proportional to its value, in a simple ratio. That helps to determine the voltage at the junction of the two resistors. The total voltage drop will be the same. You simply need to find the right voltage you want at the split, then make a simple ratio for the drops and the resistor values.

If you had a string of ten resistors with +10V at one end and 0V at ground, each resistor would drop one volt, and you would have a "ladder" with 1V steps.

[stonerbox]

Excellent Rob, thank you! So I simply measure the voltage drop over R2 then calculate and implement a voltage divider with the same drop and then adjust C1 so the low pass stays at 10Khz. Too easy! I should have solved this one on my own but I got lazy and was uncertain if the conversion to a voltage divider would alter something in the loading of the input signal.

[stonerbox]

By the way guys, are my schematics too big for your screens? I am on a 4K monitor (3840x2160) but I am trying to keep the sizes manageable for smaller monitors too.

[EBK]

I'm enjoying them on a 1920 by 1080. 
The green color makes me think of engineering paper.

[stonerbox]

I'm enjoying them on a 1920 by 1080. 
The green color makes me think of engineering paper.

That is great to hear! I never really seen engineering paper in real life but I thought the color would match this project. Good to hit some nostalgia though!

[duck_arse]

I'm having a problem with your preamp/power stage switch. OPA2134A is floating when the switch as drawn. I'd reverse that pole to have common to the opamp, and the empty switch postion taken to V/2, as that stage has no bias when bypassed. C24//C13 appear to be in series, so one of them is backwards. also, I think one is redundant, it might become clearer when the switch is reconfiggered.

R10//R11//R12 - R12 is swamped by the value of R11, so can be deleted. if you multiply the value of those 10k resistors by 10? 22?, you can then reduce the C11 by the same factor. [I think this is how those filter things work.] R6 is also swamped by the 1k and pot, it can be deleted.

I think: R22,23,24,25 and C17, 18 can be replaced by a single cap and two resistors. work out your filter resistor value for say 100nF cap, then split that value to suit your divider and take the gate from there. probably the same goes for that whole "VOL 2" section from C20 to C22.

why 713k at R2? if that is a 20kHz filter with 10pF, there is no usable difference between 713k and 680k. (most might rescale for 68k and 100pF, or as high as 220pF even.)

hard wire the 50pF across the rotary switch, and add the others in parallel. you'll always have a failsafe setting, not reliant on the switch.

master volume is drawn wrong, you want the tip-out from the wiper. I'm a bit hazy on the wiring of the OPA2134 at the end, but will leave comment to the less opamp hazy among us. I hope some of this is correct, and helps.

[stonerbox]

Good grief, so many mistakes! I am glad someone came along and shed some light on them. I really like your Pre/Pow switch idea.

That Master Volume is indeed miswired in the schematic. I am afraid it is a case of the copy/paste mishap from Vol 2.
I had already removed R11 & R12 in order to keep the signal strong enough when the rotary switch is set at 50pF. I did have my suspicions about the redundancy of R12 & R13 though but I never got to the point of investigating and optimizing these areas just yet. At the moment I am trying to re-bias the preamp section so it will not fizz and buzz when dialing in cleaner sounds. Aron noticed it (in the samples) back when I started the thread but I never got it properly dealt with. No matter how I lower the voltage (pot to ground) in between Q1-Q2 or Q2-Q3 the fizz is still there. Could it be an link to the very small caps in the rotary? I just noticed the fizz/buzz appears when set on the lower value ones (150-100-50pF). EDIT: Never mind it is the same with the larger ones.

Here is a sample of the fizz and buzz. http://tropicalhippo.com/GreenMujina/clean.wav

I'll look into all the things you pointed out for me. Also, it is a good idea to keep one capacitor in parallel by the rotary switch even though it would result in 100pF as the lowest value on the switch, right? I do not know if these switches are fragile but if it breaks we got a backup plan. Thanks!

why 713k at R2? if that is a 20kHz filter with 10pF, there is no usable difference between 713k and 680k. (most might rescale for 68k and 100pF, or as high as 220pF even.)

It is meant to lower the input signal and form a 10Khz low pass filter. I will swap that one out for two resistors and a new cap.

I hope some of this is correct, and helps.

It really does, thank you for helping me Stephen.

[stonerbox]

Closing up on version 1.0 so here is a demonstration I recorded early this Sunday morning. I use a 73' Gibson Les Paul (standard) with dead strings going into a clean Fender Blackface clone via one SM57 and one ribbon mic. In my DAW I added a small amount of top end boost (+2.7dB). That is more information than you really needed! But hey, you are welcome anyway!

I regret ever referring this pedal to be clean sounding.  

https://www.youtube.com/embed/XOCO0-fiCTk

I made a lot of fixes and reconfigurations suggested by Duck and Rob in this version (but not all yet). There was a lot of small mistakes on my part and leftovers from previous builds of the project. Some unnecessary resistors and capacitors got kicked out of there.

0.95 Update
1. Removed the output opamp (had enough level without it). I also moved the opamp configured as high pass filter from the end of the circuit to the end of the power amp section, the preamp did not really need more shine.
2. Removed a lot of redundant resistors.
3. Increased the high pass filters capacitor from 680pF to 1nF.
4. Reversed the switch to the power amp and hooked up one of the switch pins to 1/2Vcc, so the opamp would not be left floating when the power amp is bypassed. A great idea by Duck!
5. As a precaution I added a 50pF cap in parallel with the input cap switch in case it would fail. Also a good idea from the Duck.

[robthequiet]

That has a lot of character, well done! \m/

[duck_arse]

clean sounds - clean, like my sheets.

I really liked the first switch to the mid-cap position. (I think "attenuator" was the word you wanted.)

also, my comment about the 50pF on the switch, I meant to have that smallest value hard connected around the switch, so the switch will have a position with no cap on. so if the switch fails in all other positions, you still have (your fave) 50pF connected. and all the other values can be 50pF less than design value, and switch-in in parallel w/ 50pF.

also, your latest image (circuit?) shows as broken-link.

[stonerbox]

clean sounds - clean, like my sheets.

I really liked the first switch to the mid-cap position. (I think "attenuator" was the word you wanted.)

also, my comment about the 50pF on the switch, I meant to have that smallest value hard connected around the switch, so the switch will have a position with no cap on. so if the switch fails in all other positions, you still have (your fave) 50pF connected. and all the other values can be 50pF less than design value, and switch-in in parallel w/ 50pF.

also, your latest image (circuit?) shows as broken-link.

Yes, "attenuator" was the word I was looking for! I must have read you comment in a rush the first time around, moving the 50pF all together down to the board is much more logical and better idea. 
Anything above 150pF as input capacitor is overkill for me. THE BASS, oh the bass.. Probably because I used 6,8uF coupling caps in the rest of the circuit.

Forgot that I resigned from my previous hosting company but all content is up-to-date on a offline storage. Phew.. Here is a Shiba Inu to still the tingling nerves.

[stonerbox]

Sorry but I have to ask just to be sure I got this right. The ratio of the voltage divider (R20 4,7k / R21 33Ω) is around 142:1.  If I would change these to either 47k/300Ω or 500k/3,5k (also around 142:1) would the voltage divider give the same performance as the with current values?

[duck_arse]

the voltage should divide the same, yes. but you can also then scale your feeding cap in the opposite direction - R times 100, C divided by 100. if you keep the massive 6u8, you may get blocking distortions.

[stonerbox]

the voltage should divide the same, yes. but you can also then scale your feeding cap in the opposite direction - R times 100, C divided by 100. if you keep the massive 6u8, you may get blocking distortions.

I don't get it? Reverse C17, and why? What is this blocking distortion you speak of?