Blendable Tone Stack...?

[ch1naski]

Is it feasible to make a tonestack with a blend pot, without losing too much signal?
I know the BMP tonestack pulls some signal out of the circuit, so is there a way to split the signal before the stack, say, with a potentiometer, so you could kind of balance the un-filtered signal with the signal going thru the stack? Of course, i guess the tonestack wouldn't be as effective when blended, but that would be ok....

curt

[Vallhagen]

To split a signal in two paths is "fairly easily" (well...) done, as well as mixing the signals together after the signal processing in one of the paths. Your problem will be to handle the phase shift between the filtered and unfiltered signal.

Take a look at my Blüe Monster (http://www.madbeanpedals.com/forum/index.php?topic=6968.0) if you like to see how i solved a rather similar task. Borrow or steal whatever parts you need:)

Cheers

[RobA]

If what you are looking for is a Big Muff style tone control but without the insertion loss, then you could try an active tilt control. The PDF image below is lifted from a circuit I was working on. I liked the way it worked. Although I deleted the bypass switch circuit (just take out the switch and R3, R7) because the center position of the pot pretty much does that anyway. I also added a DPDT that could switch in caps in parallel with C2 and C3. This lets you move the tilt point for the control.

You can do the first op amp stage as an inverting unity gain stage so that it preserves phase.

Also, the version here uses split rails but I'm pretty sure you could convert it to a single supply design

The image looks a bit weird in the preview, but if you open it in a new window it's a pdf and should be rescalable and easier to read.

[midwayfair]

Some thoughts:

The BMP tone stack does reduce output level but ... have you HEARD the BMP? I mean, does it NEED more output level? :/

A tone bypass would add another pot -- is this more desirable than, say, simply using AMZ's "Body" control (to bypass the mid scoop) or simply using a TMB tone stack?

The BMP tone stack is panning between two different filters. There are ways to rejigger the tone stack so that with the tone knob rolled in one direction will essentially be a bypass.

You can make the switch in this a large pot and it will work as a tone bypass:
http://www.muzique.com/lab/tbypass.htm

It won't be as perfect as a switch even with a 1M pot, but you're not likely to notice a huge difference.

Purely theoretical, but works in my head: You could use a dual-gang pot in a kind of weird way. Hook up one variable resistor to the collector of Q3, and the other side of it to the coupling cap post-tone stack. Disconnect the wiper of the tone stack and hook up the other gang of the bypass pot to it, in series with the coupling cap. Wire them so that when you turn up the pot, the resistance between Q3 and the coupling cap input to Q4 decreases, and the resistance between the Q4 coupling cap and the wiper of the tone stack increases. This would simultaneously bypasses the tone stack altogether and prevent "backwash" into the tone stack. Probably needs refinement ... or it might not work at all!

[ch1naski]

I knew I was asking in the right place.
I can always depend on the helpful people here.

Looks like I've got some breadboarding to do. (I hate breadboard)

I didn't think about the phase issue until after I posted the question. Splitting the signal automatically makes it into a somewhat larger project, I may just go with bmp/amz setup and "rejigger" it so it's somewhat bypassed. Space inside the box is the issue, and adding a daughter board or making more real estate available on an etch may not work out.

I really would prefer the split signal.....but it's opening a can of worms, isn't it?
 I figure I'll try the easiest method first, and if that doesn't workout-_- then I'll start torturing myself by breadboarding the other two that Vallhagen and RobA have suggested.

Thank you, guys.


Sent from my Nexus 7 using Tapatalk HD
&IMgKSJS>@IMgKSJS>

[BuGG]

Sorry to dig up an "old" thread, but could someone point me to some info about (or explain) the issue of phase shift when blending split filtered signals back together?

I'm particularly interested in battling the negative effects of splitting a signal into two bands (hi/lo), overdriving/distorting them separately, then mixing back together.

I can't quite wrap my head around what Vallhagen is doing to correct the phase differences in his BLÜE MONSTER, and I don't see any of this extra circuitry on the Quadrafuzz (for example).

Any info would be greatly appreciated!   

[Vallhagen]

I'm particularly interested in battling the negative effects of splitting a signal into two bands (hi/lo), overdriving/distorting them separately, then mixing back together.

I saw my name

...and the task you look for ("...distorting them separately [...] mixing back...") is exactly what i have done in the BLÜE MONSTER, as you have noticed.

Frequency band splitting and phase correction is a common task in (HIFI) speaker design, so its among the hifi folks i found most of the info. This is some great Reading, Linkwitzlab almost cover it all: http://www.linkwitzlab.com/filters.htm

If you take a look at the BLÜE MONSTER schematic, the signal is splitted in two bands, "hi" and "low". the stages immidiately after the input stage are highpass resp lowpass filters, with the same cutoff frequency. So far nothing spectacular? This stages cause timedelay/phaseshift. Hi and Lo are 180deg out of phase. Here's where we need phasecorrection. The next stage does this, designed as "all pass / phase shift"-stages. These correction stages are - in turn - 180deg phaseshifting but in "opposite direction" to the earlier hi- and lo-pass filters. So we are back at zero, task solved:)

There might be other ways to solve this, but i found this being the best way. Hint: it was when i played around this filters i noticed i needed bi-polar Power to get the math vs reality right. I also recommend simulation software if you like to experiment.

And you are right, Craig Anderton does have a different approach. the Quadrafuzz is based around bandpass filters, not hi- and lo-pass ditto. So Andertons design gives you notch artefacts. As you have read Craigs Words about his Circuit he saiz something like "it takes time to dial in the perfect sound".

Hope this were of some help:) Feel free to ask additional questions, i know i tend to be "short" when i explain stuff.

Cheers
Bengt

[BuGG]

Thanks for the info!   The info at Linkwitzlab looks very helpful, I can see now I need to dig out my old textbooks and read up on active filters.   

What defines the 180 degree phase shift?  Is it the fact that they are opposing filters (HP/LP) or actual component values?

Would the degree of shift be different if one of the filters had a different cutoff frequency?

Can you recommend a decent circuit simulation software that isn't terribly intimidating?

My apologies if these are elementary questions, I'm certainly no EE and have been out of school for quite a while now.   

[Vallhagen]

*1* What defines the 180 degree phase shift?  Is it the fact that they are opposing filters (HP/LP) or actual component values?
*2* Would the degree of shift be different if one of the filters had a different cutoff frequency?
*3* Can you recommend a decent circuit simulation software that isn't terribly intimidating?

1) Opposing filters (actual kind of HP and LP) gives, "by nature or by physics laws" 180 deg phaseshift. It doesnt matter which cutoff frequency you choose, you will Always have 180 deg as long as your R and C values are the same in both filters.
2) yes, if you choose different cutoffs for the LP than for the HP, the phaseshift will be something different than 180.
3) I did some research last year, and i have chosen Tina-TI. It is freeware and does the job. I know that LT-Spice is popular around here and on DIYSB, but i found it harder to get into than Tina-TI.

Cheers:)

[BuGG]

Thanks for pointing out Tina-TI!   I grabbed it earlier today and was almost immediately tinkering with active filter circuits.   Excellent tool!

Is there any particular reason you chose to use two phase shift correction circuits rather than a single inverting stage after either of the two filters?

[RobA]

It's not so much of a correction to the phase shifts as it is bringing them into line with each other so that they won't add destructively.

The LP and the HP filters each alter the phase for frequency response. There is a 180° shift in each signal when comparing the phase above and below the cutoff frequency of the filter. If you look at the frequency response of the filter section in your spice program, you should be able to see it graphed.

The problem is that the LP and the HP shift the phase in different directions. The two all pass filters align the two sides of the split signal to have the same phase shift and then they can add constructively.

[Vallhagen]

It's not so much of a correction to the phase shifts as it is bringing them into line with each other so that they won't add destructively.

This is actually a better explanation of the correction stages than my explanation above. Thanx RobA.

Is there any particular reason you chose to use two phase shift correction circuits rather than a single inverting stage after either of the two filters?

I tried a few other approaches, among other a single inverting stage. But nothing gave a satisfying result. Also, when trying to simplify, even if you succeed you dont win many components.

(On the other hand, this "fx world" is funny when it comes to soundshaping. We are actually - in opposite to the HIFI guys - destroying sound. We dont have to follow the rules, something fun can come out if we do it the wrong way. I mean; unwanted phaseshift might sound cool in the right context. But this is a parenthesis, right now we are searching for "something perfect")

I'm glad Tina-TI did it for you:). I like it as well.

Cheers

[BuGG]

It's not so much of a correction to the phase shifts as it is bringing them into line with each other so that they won't add destructively.

That makes sense.    I was able to see this phase shift happening in Tina-TI, which is really cool to be able to visualize what a circuit will do to a signal before actually building it.   I'm a complete beginner to circuit simulation software but I can see it's going to prove very useful.   

I tried a few other approaches, among other a single inverting stage. But nothing gave a satisfying result. Also, when trying to simplify, even if you succeed you dont win many components.

Ahh yeah, I certainly wasn't trying to "improve" your circuit or tell you something you didn't already know!     I was just curious to know if "technically" it would work.    Much like your comment about not following the rules, I realize quite often there is a difference between what should work in theory and what actually works in reality.     

Thanks to you both for the replies, I think I have enough info now to hunt down the dirty details.   

[ch1naski]

All way above my skill level and comprehension.

Sent from my SPH-L710 using Tapatalk 4