Enhanced Tone Control on Red Llama circuit?

[Mr. Lime]

I'm currently working on a kind of Red Llama circuit with some minor chances to implement a tone control. CMOS suffer usually from a high output impedance why tone controls maybe not work as good as expected. My idea is to take apart the lowpass and highpass filters from the BMP tone control and give each of them a driving CMOS gain stage/buffer. I am not sure if therefore the common tone control can be omitted and replaced by two 33k isolation resistors. How much lows and highs get through could be dealed in with the gain stages.
Are my considerations purposeful?
Should I keep the tone control?

Maybe someone could give some educated advice.

Thanks so much!

[Mark Hammer]

Typically, the variation in gain in such a circuit is accomlished via a common stage up front, and derivatives of the BMP tonestack simply adjust the relative balance/level of the already-distorted higher and lower bands.  Yourdrawing indicates that separate drive is being used for what gets transformed into higher and lower bands.

There are pluses and minuses to that approach.

• +One does not have to forfeit bass/mids for treble, with the normally reciprocal BMP tone control
• +The relative drive of each half means you can almost skip the master output volume control.
• -The two individual gain stages aren't really "doing" anything different.  True, as shown A3 omits the 51pf feedback cap, but that cap actually isn't doing much that is tonally audible, merely maintaining stability.  Its rolloff is well above human hearing and speaker.  A2 and A3 each amplify the same signal range, and any real shaping is done via the subsequent filter network.

My own bias is that you'd have better luck and control doing the following:

• Make A1 your variable gain stage, or better yet, make A1 a fixed gain stage and use a JFET or op-amp input stage to provide varying amounts of drive.  This means that the feedback resistors for A2 can be simply 100k.  Since the amplitude of bass content is generally much greater than that oif treble content, you'll need to compensate a bit by upping the gain of A3.  So consider making R9 something a bit higher, like 330k.  A feedback cap of 82-100pf should be suitable and help to reduce fizzies as well as some noise.
• Use C2/R3 for A2's input (although I like to use 100nf myself), but connect A1's output to A3 via a separate parallel path.  For example, use 100k but 6800pf instead of 47nf.  That will do some of the highpass filtering that seems to be your goal for A3.  As part of the same overall strategy, make C6 1000pf instead of 51pf, to roll off top end above 1500hz for the "bass" section.
• Either make the tonal balance control continuous, like a BMP, instead of a pair of 33k mixing resistors, or else consider using two variable resistances to add in or remove as much of each channel as you wish, prior to the master volume.

Having experimented with such parallel-drive things, I will simply say that one shouldn't plan around using identical drive for each path, and one should plan around using whatever interstage opportunities present themselves to do tone shaping, rather than relying entirely on output filtering to do the heavy lifting.

[Mr. Lime]

Thanks for spending your time and giving me such a extensive reply, Mark!

I see your point, A2 and A3 are almost identical and you are probably right in voicing them differently. My goal was to stay in the Red Llama ballpark but different coupling caps between the stages don't hurt for sure. Is there a specific advantage in relocating the gain control in the feedback loop to A1 only? Does it matter much if the first stage is gain controlled while the second one has 1M in the feedback loop or visa versa? Treble caps are updated now.
As I'd like to implement that circuit on a PCB kit an additional JFET or op-amp stage would be too much for my taste and when considering the Red Llama circuit does sound alright it's not even necessary.
Furthermore I'd like to try hard clipping circuit which is either connected to the low pass or high pass filter when the switch is engaged for some fuzzy tones. Maybe I should decrease the the volume pot to 50k for a better range?

Jack Orman once came with a similar approach


Here's the updated schematic:

[Mark Hammer]

Invertors have a nice unique kind of sound.  I don't know if supplementing that with hard clipping would really add anything appealing, but maybe that's just me.

As for what the advantage is in using A1 (or an earlier stage) for drive, it's simply easier and more repeatable to have one common gain control, than to have to try and get the settings of two controls "right" to re-achieve a preferred sound.  It's like a food recipe in "parts" (4 parts this to 3 parts that, whatever the total volume).  A2 and A3 have their individual gains suitably set for the band they are covering, and A1 simply gives the same push to each.  Remember that gain is multiplicative, so if (hypothetically) A1 is 5x-20x, A2 is 10x and A3 is 20x, then what comes out of A2 is 50-200x and what comes out of A3 is 100-400x.  The ratio of gains between the A2 and A3 paths is preserved.

[Mr. Lime]

Well I used to built a prototype to test the circuit and get a better understanding of the different parts.
That's my conclusion:

There are really nice sounds coming out if the lows are dealt back and the highs are  boosted. The 100k Volume control isn't really practical as the volume gets pretty low when the gain stages aren't driven at maximum. The range of the tone control isn't that awesome either to be honest. It's getting muddy pretty fast and I guess 1M for the bass side is far too much although some fuzz tones are possible.

Haven't tried the diode switch yet but I think this one wouldn't bring me further anyway.
What Mark suggested concerning the bass content was absolutely right and the separate gain stages aren't a great advantage either. The exact gain that is perfectly needed for the individual gain stages has to be tested and maybe then a common gain control on the first stage would be fine.

This is the circuit I came up with the parts I had laying around: