Passive tilt EQ with potentiometer?

[Vivek]

From that lovely tool TSC IN THE WEB:

Merlin Boneray:

https://www.guitarscience.net/tsc/boneray.htm#RIN=38k&R1=470k&R2=470k&RM=1M&RT=1M&RL=1M&C1=220p&C2=1n&C3=4.7n&C4=4.7n&RT_pot=Linear&RM_pot=LogB

Big Muff:
https://www.guitarscience.net/tsc/bigmuff.htm#RIN=1k&R1=39k&R2=22k&RM=100k&RL=1M&C1=4n&C2=10n&RM_pot=Linear

[mdcmdcmdc]

Rob Strand - thank you for those schematics, that's really helpful!

I'm curious if you have any suggestions on where to start tweaking them for different response? eg, adjusting the centre frequency and increasing boost/cut/insertion loss.

Also, just so I'm understanding the schematic correctly - there's the "POT" label in each that looks like it's sitting between two 100K resistors. I'm not clear if there's a pot with an unmarked value in there, or those two resistors represent one pot or...? Apologies, just want to make sure i understand what's going on in the schematic.

[mdcmdcmdc]

Also, I think tilt EQs tend to work well in instances where you want just a bit more overall bass or treble without changing the fundamental character of the audio too much. The passive tilts I'm familiar with in studio use are usually stepped in 1/8dB increments from flat to +/- 0.5dB. That's obviously way too subtle for "tone shaping" per se, but somewhere between 3-6dB at the extremes seems useful if you just want to give it a slight tweak switching between guitar/bass use etc.

The issue I have, personally, with trying to combine mid control with treble and bass a la mod'd BMP tone controls is that I don't usually want the mid notch to be in the same place as the centre frequency. It's easy enough to get a curve at 12:00 that sounds really nice (slight notch at ~250Hz, slight shelf boost above ~1K or whatever), but the adjustment on either side often sounds pretty bad. Splitting the EQ into two sections - a treble:bass ratio adjustment and then a fixed notch, makes sense to me for the way I like guitars to sound.

Your mileage, of course, may vary!

[amz-fx]

A passive tilt EQ is part of the Big Muff's sound.  Jack at AMZ 'took it apart' and did some mods, and describes the basics of it here:
http://www.muzique.com/lab/tone3.htm

A different version from 14 years ago:

http://www.muzique.com/lab/atone.htm

regards, Jack

[Vivek]

I'm curious if you have any suggestions on where to start tweaking them for different response? eg, adjusting the centre frequency and increasing boost/cut/insertion loss.

Maybe you will have to enter it into one of the free online circuit simulators

The NATIONAL SEMICONDUCTOR AUDIO book around 1978 had equations for 2 Band Baxandall, but said that there was too much interaction and loading with 3 bands, and it is best to use a simulator to tweak tone controls that interact with each other.

[mdcmdcmdc]

I suppose this is as good an excuse as any to start poking around LTSpice...

[ElectricDruid]

ESP had some pretty damning words to describe the Tilt EQ over at https://sound-au.com/articles/eq.htm#s10

I'm not entirely sure why anyone thought this was a good idea, but it's part of tone control history, so is included.
There are many possible tweaks that can shift the centre frequency or provide asymmetrical response, but these are generally as useless as the circuit itself. I expect it seemed like a good idea at the time, but it's really a rather pointless waste of parts.

Why was it considered unworthy ?

Which one knob tone control is better ? In which aspect ?

He's thinking about it from a hi-fi point of view, in which case boosting one bit of spectrum by 10dB while simultaneously cutting the other bit by 10dB is pretty much bound to mess up your sound! In hi-fi, tone control is *supposed* to be "equalisation" e.g. intended to equalise the response of different listening environments so you get the same natural sound in every situation.

Of course, we're not doing sound *reproduction*, we're doing sound *creation*, so it's a completely different ballgame. We're not striving for a flat response, but instead trying to deliberately manipulate the frequency response to achieve a certain sound. In terms of getting a lot of "bang for the buck" (in this case, lots of control for only one knob on the panel) it's not a bad circuit at all. No tone control suits all tastes or can do all jobs, but it's a good one to know about for pedal builders.

[Vivek]

Very valid points, Tom !!!

Are there videos on the HARD BARGAIN, showing the effect of the Tilt EQ ?

[PRR]

There are a fair number of commercial recordings which don't sound right. Some of them DO clear-up with a 1dB or 2dB tilt. All the highs got pressed a bit fainter than all the lows, or vice versa.

So yes, part-dB steps to a few dB max. Not ten dB. That's where a James shines.

That's music RE-production.

I see less justification for a music Production system having a wee bit, or a heavy hunk, of tilt. You want separate control of each ingredient.

[ElectricDruid]

I see less justification for a music Production system having a wee bit, or a heavy hunk, of tilt. You want separate control of each ingredient.

That's just a "more controls are better than less controls" argument, which is true as far as it goes. But panel space isn't unlimited  and sometimes one knob that can tweak the overall balance is good. Sure, if you've got room, go for it - stick a full Bass, Mid, Treble set on there. Or go parametric! There are so many options for tone controls, and they're all valid and have their uses. There's a balance to be struck between versatility (more options) and ease of use (less controls). The "one knob" controls are clearly at one extreme end of that spectrum, but nonetheless have found a place in pedal history. Where would we be without the BMP tone control?!?

On the Hard Bargain, I used a one-knob tilt control, but I also included a mid cut/boost, so in some ways it's a different take on the Bass/Mid/treble control, but with only two knobs and some different options (because they interact differently).

Endless experiments to do, and no real right answers, just different flavours....

[Rob Strand]

Rob Strand - thank you for those schematics, that's really helpful!

I'm curious if you have any suggestions on where to start tweaking them for different response? eg, adjusting the centre frequency and increasing boost/cut/insertion loss.

Also, just so I'm understanding the schematic correctly - there's the "POT" label in each that looks like it's sitting between two 100K resistors. I'm not clear if there's a pot with an unmarked value in there, or those two resistors represent one pot or...? Apologies, just want to make sure i understand what's going on in the schematic.

The two resistors are one 100k pot and no other resistors.  I've tried to clarify this in this post.

There's a whole lot of technicalities to deal with.

Here's an example of a 12dB step (ie. +/- 6dB tilt).

The basic calculations only works when the step is reasonable.  For a 6dB step (+/- 3dB) tilt the cap values will be off - for example compare the calculations with my previous example.  It's possible do calculations for small steps but it's very messy.  The calculations are messy enough because they have a stack of approximation fixing the effect of the pot.

It's a bit laborious coming up with approximations for this stuff.  I could easily have made some small errors in the approximations.  Nonetheless you can see the LTSpice sim looks reasonable with the calculated values.

Calculations,


Sim Example:
I've also added some notes about manual tweaking.   If you get the hang of the manual tweaking you don't need to do accurate calculations.



FYI:
You can see the tilt is no longer symmetrical about middle of the maximum boost when the boost is maximum.  It's symmetrical about the -3db point instead.   I need to see if changing the impedance scaling of the arms can fix it.  I thought I got a symmetrical 12dB working before when R1 was 10k.  In the example I set R2 to 10k instead.

[mdcmdcmdc]

Heck, thanks so much for putting that together!

[Rob Strand]

Heck, thanks so much for putting that together!

No worries.

You will have to see how it sounds.   12dB is still pretty hefty.  Maybe better to start with 10dB.
Also, it's worthwhile moving the center from 1kHz.  A lower frequency will sound more even.

[mdcmdcmdc]

Agreed - when I've tried bax/james controls I've usually put the centre closer to 600.

[mdcmdcmdc]

...speaking of which, apologies for the potentially obvious Q, but in the notes in the schem you reference "ka" and "kf"; what do these refer to? Checking the formula spreadsheet, I'm not seeing a simple answer.

[Rob Strand]

...speaking of which, apologies for the potentially obvious Q, but in the notes in the schem you reference "ka" and "kf"; what do these refer to? Checking the formula spreadsheet, I'm not seeing a simple answer.

They are used for *manual* tweaking in LTspice. 

kf sets the moves the center frequency up/down by a factor kf.

ka tweaks the flatness when the tilt control is centered.  It modified the caps but keeps the centre frequency the same (more or less).

The LTspice sim manipulates what is there.

The spreadsheet takes requirements and some nominated part values (R2, R3, Rpot) then produces estimates for the other ( i.e. unknown) values.    The spreadsheet is approximate.   The process would be to get an idea of the values using the spreadsheet then check them in LTspice.  Then perhaps tweak the values.

[Vivek]

I suppose this is as good an excuse as any to start poking around LTSpice...

Do start off on SPICE !

Rob already entered the TILT EQ so that saves you the hassle of data entry.

Rob and I could assist you if you get stuck somewhere in your analysis

[mdcmdcmdc]

...speaking of which, apologies for the potentially obvious Q, but in the notes in the schem you reference "ka" and "kf"; what do these refer to? Checking the formula spreadsheet, I'm not seeing a simple answer.

They are used for *manual* tweaking in LTspice. 

kf sets the moves the center frequency up/down by a factor kf.

ka tweaks the flatness when the tilt control is centered.  It modified the caps but keeps the centre frequency the same (more or less).

The LTspice sim manipulates what is there.

The spreadsheet take requirements and some nominated part values (R2, R3, Rpot) then produces estimates for the other ( i.e. unknown) values.    The spreadsheet is approximate.   The process would be to get an idea of the values using the spreadsheet then check them in LTspice.  Then perhaps tweak the values.

Ahh, got it. LTSpice on a mac seems to be a very painful experience, so I may just stick to tweaking this on the breadboard for now.

[ElectricDruid]

LTSpice on a mac seems to be a very painful experience

We Mac LTspice users like to charitably describe it as "quirky"...

[Rob Strand]

Ahh, got it. LTSpice on a mac seems to be a very painful experience, so I may just stick to tweaking this on the breadboard for now.

The user interface is a little awkward at times.     If you are used to using the control keys for editing (like Ctrl+X, Ctrl+C and Ctrl+V) you can re-program a lot of keys to make it more like other editors.  I reprogrammed all the function keys to be Ctrl+x sequences.   Some only make sense because other software uses similar keys, for example Ctrl+R = rotate part.  You change the settings in the Control Panel.

You don't have to use LTspice, any spice sim will work.    The only thing you might have to change is how the variable parameters are done.  At worst you would loose the parameters and put in a single number, like on a schematic.