Big bag of Ruby mods [renamed]

[dano12]

So I finally figured out the wildly complex Big Muff tone control I've also mapped out the Ruby amp and actually spent time with the 386 datasheet in an attempt to understand what each part is doing. I searched this forum to figure out how to insert the BMP tone control into the Ruby, and came up with a couple general ideas but nothing concrete. What I've learned:

1. The tone control will result in a loss of signal--that's the nature of the beast.
2. A post here suggested the following order for adding a tone control to the ruby: buffer->tonestack->buffer->volume->386->output
3. The first transistor and related components in the Ruby represent an input buffer.

So I've drawn my own version of the Ruby schematic, inserted the tone control, added another buffer stage, and added a switch and cap in the feedback loop of the opamp.

Here it is so far:



Questions:
1. Am I close?
2. Will my second buffer stage make of for the signal loss of the tone control.
3. The LM386 datasheet says that a 10uf cap placed between pins 1 and 8 will yield a gain of 200. Will my switch/cap (in red) accomplish this, or is the 1kohm pot doing essentially the same thing when it is fully open?

Thanks very much in advance for any help with my first attempt at modifying an actual schematic!

[B Tremblay]

Questions:
1. Am I close?

Very close!  The only issues are an extra power supply filter cap (you only need one 100uF from +V to ground), the redundant gain switch  (which you suspected was unnecessary in question 3), and there should be coupling caps before and after the tonestack.

2. Will my second buffer stage make of for the signal loss of the tone control.

No.  A buffer has at most, a gain of 1.  The buffers are in place to prevent the loading of the guitar and amplifier.  When builders have asked me about adding a tonestack to the Ruby amp, I usually suggest adding a small gain stage before the tonestack and then the stock Ruby.  You could use a Fetzer Valve, the NPN Boost, or the output stage of a BMP.

3. The LM386 datasheet says that a 10uf cap placed between pins 1 and 8 will yield a gain of 200. Will my switch/cap (in red) accomplish this, or is the 1kohm pot doing essentially the same thing when it is fully open?

Correct.  When the Gain pot is at maximum, pins 1 and 8 are connected, providing full gain.

Your schematic is very attractive, as is your website.  Nicely done!

[bancika]

Nice looking schematic, which software are you using?

[dano12]

Very close!  The only issues are an extra power supply filter cap (you only need one 100uF from +V to ground), the redundant gain switch  (which you suspected was unnecessary in question 3), and there should be coupling caps before and after the tonestack.

Wow, I'm close! This is exciting stuff. Regarding coupling caps, would the existing 47 nf in the Ruby schematic be the coupling cap? So I could just add another 47 nf?

No.  A buffer has at most, a gain of 1.  The buffers are in place to prevent the loading of the guitar and amplifier.  When builders have asked me about adding a tonestack to the Ruby amp, I usually suggest adding a small gain stage before the tonestack and then the stock Ruby.  You could use a Fetzer Valve, the NPN Boost, or the output stage of a BMP.

Thanks much, this cleared up a large mystery for me. I'll look at the options and redraw schematic. Hmmm....maybe this is a great place for your awesome ToneMender circuit....

Your schematic is very attractive, as is your website.  Nicely done!

And thanks to you--I've built quite a few things from runoffgroove.com.

[dano12]

Nice looking schematic, which software are you using?

I use Visio for pretty much everything. The electronics schematics library is a bit weak, so I've created my own templates for various things like FETs and opamps. But Visio covers just about everything I've thrown at it. I also use to do all my wiring diagrams so I've built up a library of graphical representations of pots, switches, jacks, etc. Highly recommended.

[Mark Hammer]

Not meant as a criticism, but it seems a little more complicated than necessary to me, and at the same time insufficient in some ways.  Your FET input stage provides all the buffering needed, and a second one is overkill.  What the input stage needs to do, though, is provide some gain to offset the passive loss of the tone control.  Keep in mind the BMP-style tone control works by bleeding signal to ground at both the high end and low end, and what comes out the wiper of the tone pot is what's left.  That's why the BMP needs a gain recovery section after it to be able to output something hot enough to achieve noticeable boost and produce additional overload in subsequent sections.

So, what you need here is a hotter input stage that gives 9db or more of gain. Then, what comes off the wiper of the tone pot can simply be fed to the 47nf cap and the volume pot directly.

Alternatively, the 386 provides a number of ways to get boost at the top and bottom ends, and these may well be more of what you want.  Keep in mind that the BMP tone control is really more oriented at "revoicing" a fuzz, then EQ in the classic sense of compensating for speaker or signal source.  That MAY be what you intended...but maybe not.

There is a feedback loop between pins 5 and 1 that can be exploited for bass boost, and the more familiar feedback loop between pins 8 and 1 can be modified to use different capacitor/resistor feedback combinations in parallel (just like the Rat) to yield differential gain for different frequency ranges.  I was tuned onto this a couple years ago by a Nuts & Volts article/project that used much smaller value caps than 10uf to provide a big boost in the upper treble.  The project was a pair of powered extension speakers that older people with high-end hearing loss could plug into the earphone jack of their TV and park beside them on the couch/sofa.  By sticking a smaller than normal cap in that loop, the higher gain that comes from having a low-impedance path in the 8->1 path was applied only to high end, with more modest gain applied to the rest of the spectrum. 

Another thing to consider is the use of a "bright" cap on the volume control so that top-end is preserved as you turn down.

But again, all of this assumes your goal for the tone control is to provide compensatory sonic adjustment, rather than tonal overhaul.

[B Tremblay]

Wow, I'm close! This is exciting stuff. Regarding coupling caps, would the existing 47 nf in the Ruby schematic be the coupling cap? So I could just add another 47 nf?

Yes, though you could probably go a little bigger than that after that first stage so you're providing the tonestack with a nice full-range signal, maybe 100n or 220n.  Though, like Mark suggests, that first stage should have some gain.

[dano12]

Thanks very much for the pointers. I chose the BMP tone control simply because it was simple. However, given the feedback, I decided to go back to the drawing board and do some more research. I've found a JFET based boost, tacked on (what I think is an accurate rendition of) the Stupidly Wonderful Tone Control, and sent it all of to the input of the stock Ruby.

Does this look better? Is the stock Ruby buffer in the correct place? Can I use an MPF102 instead of the J201 for a more managable boost? Why do I ask so many questions?

[jrc4558]

You dont really need a buffer in your second attempt. just a booster, a tone control, and LM386

[cab42]

Why do I ask so many questions?

Because they are very relevant questions!! I have followed this thread with great interest as I have been thinking in the same lines.

BTW: I use Visio on a regular basis, but I have never thought of using it for drawing schematics.

Regards

Carsten

[Mark Hammer]

The SWTC is what you might call a "something for nothing" control, in that it gives some tonal adjustment without having to pay too huge a price.  Note that it will only trim top end, and not provide any boost.  If all you want is to mellow things out, it'll work fine.  If you want some brightness added, you need to look elsewhere.  Similarly, if you want some bottom boosted, you'll need to look elsewhere.

One of the clever things Joe Gagan packed into a number of his excellent Fuzz Face extrapolations was a variable bottom control on the input.  Here, the signal is faced with two possible capacitors to pass through.  One of the caps is fairly small in value, giving a thinner input signal.  The other is a larger value with a variable resistance in series to adjust how "easy" it is for the bass to pass through.  When the resistance is turned up, the signal "prefers" to pass through the smaller cap and there is less bass content.  When the resistance is turned down, the higher-value cap now becomes an easier path for low end to proceed along and more bass content is introduced.

If you decide to stick with the SWTC in some form, here's a suggestion:

• First, wire up the SWTC pot directly to the volume pot
• Second, replace the .15 cap ahead of the SWTC with a smaller (say, .0033-.0068uf) cap
• Third, wire up the .15uf cap in series with a 100k pot set as variable resistor and put it in parallel with the small cap

These series caps will provide the needed DC blocking, sao the 47nf cap you show becomes unnecessary.  The SWTC will roll off high end, and the parallel pot/cap will let you dial in low end.  These two should permit you to achieve thinner brighter tones, bright full-range tones, and darker tones, as well as other combinations.  You may well need to tinker with cap values to suit whatever speaker you settle on, but you'll have a nice range of possibilities, a simple gain structure, and a fairly small board.

[dano12]

If you decide to stick with the SWTC in some form, here's a suggestion:

• First, wire up the SWTC pot directly to the volume pot
• Second, replace the .15 cap ahead of the SWTC with a smaller (say, .0033-.0068uf) cap
• Third, wire up the .15uf cap in series with a 100k pot set as variable resistor and put it in parallel with the small cap

These series caps will provide the needed DC blocking, sao the 47nf cap you show becomes unnecessary.  The SWTC will roll off high end, and the parallel pot/cap will let you dial in low end.  These two should permit you to achieve thinner brighter tones, bright full-range tones, and darker tones, as well as other combinations.  You may well need to tinker with cap values to suit whatever speaker you settle on, but you'll have a nice range of possibilities, a simple gain structure, and a fairly small board.

This is most helpful. Thanks for the the lead on Joe's circuits--I'll have a look.

I followed your suggestions and although I was a wee bit lost on step three, I think I have it right. Think its ready for the breadboard?

[B Tremblay]

Looks good!  I'm still a bit slow when it comes to impedance issues, so I tend to err on the side of caution; hence my suggestion of the keeping the stock Ruby input buffer.  However, I'll gladly defer to those with a better grasp!

My only suggestion is to remove the 1M resistor from the first stage's gate to ground.  It's unnecessary since you have that 1M pot there.

[Mark Hammer]

First, since I haven't said it yet, these are really nice drawings.  Kudos.

Second, BT is bang on about the redundant 1M resistor.  The only question mark for me is whether you need the pot there or can simply do with a 1M fixed resistor.  I take it that the input stage doesn't add THAT much gain, so the input pot and volume control don't really fulfill any real sort of gain/master function, like they might with a hotter gain structure.  On the other hand if you anticipate feeding it potentially overloading signals, then the input pot can be used to pad down such signals and make them more manageable.

Looks just about ready.  Keep in mind that the assorted tone-cap values are merely starting points and you may find other values more suitable for your needs.  Also, keep in mind that 12v worth of penlights (an 8-pack) can make 386s peppy little things with a nice punch and longer battery life than with a 9v.  You obviously can't stick it in an Altoids container, but you'll appreciate what a beefier supply can do.  If you go with 12v, make sure to use an 8 ohm speaker.  386s don't mind higher supply voltages, but they'd rather encounter low-impedance speakers/loads when using more modest supply voltages.

[dano12]

My only suggestion is to remove the 1M resistor from the first stage's gate to ground.  It's unnecessary since you have that 1M pot there.

Good catch. Thx!

[dano12]

First, since I haven't said it yet, these are really nice drawings.  Kudos.

Second, BT is bang on about the redundant 1M resistor.  The only question mark for me is whether you need the pot there or can simply do with a 1M fixed resistor.  I take it that the input stage doesn't add THAT much gain, so the input pot and volume control don't really fulfill any real sort of gain/master function, like they might with a hotter gain structure.  On the other hand if you anticipate feeding it potentially overloading signals, then the input pot can be used to pad down such signals and make them more manageable.

Looks just about ready.  Keep in mind that the assorted tone-cap values are merely starting points and you may find other values more suitable for your needs.  Also, keep in mind that 12v worth of penlights (an 8-pack) can make 386s peppy little things with a nice punch and longer battery life than with a 9v.  You obviously can't stick it in an Altoids container, but you'll appreciate what a beefier supply can do.  If you go with 12v, make sure to use an 8 ohm speaker.  386s don't mind higher supply voltages, but they'd rather encounter low-impedance speakers/loads when using more modest supply voltages.

Thanks for the kudos and the help! This is going on the board tonight sans the redundant 1M.

I'll play around with cap values: we are talking about the .0033uf and .15uf parts right? I'm thinking that if I find a nice range of values, I may put several off them on a rotary switch.

I've run my existing stock Ruby on both a 9v battery and a 12v wall wart. You're right--the higher voltage makes it even more interesting.

Have you seen the movie "Men in Black"? I've decided that the name for this circuit will be the "Noisy Cricket", as in:

[dano12]

As I've been breadboarding this, I happened upon an interesting little mod:

- Remove the 100nf cap that goes from pin 7 to ground
- Add a 56nf cap across pins 5 and 7

This adds a *nice* amount of extra gain. A bit more noise too, but it sounds great.

[Ardric]

The resistor in the Zobel network, R5 in the last schematic, should be 10 ohms, not 10k ohms.  I think this is scaled to an 8 ohm speaker load, but it'll probably be fine with 4 or 16.  It's purpose is usually to protect the chip amp, but it may also help take some of the top end fizz off.

I think some series resistance on the MPF102's gate may help with RF rejection.  To be really complete you could include a dc-blocking capacitor and protection diodes.

I'd also embellish the power supply a little.  I'd put C1 as close to the 386 as possible, and supply the battery V+ to C1's + terminal.  I'd use C1's - terminal as a star ground point for IC1 pin 4 (gnd), the gnd side of the speaker cable, and the gnd for the rest of the circuit.  I'd then use a small RC filter like 100 ohms/22uF to filter the power before Q1's drain, because I'd be worried that IC1 will be pulling enough current out of the power supply to overwhelm C1 and modulate the signal on to V+, which will feed back through the PS to Q1.  Oh, and it wouldn't hurt to make C1 bigger, like say 1000uF, and include a polarity protection diode.

My last suggestion is to "valvestate" the 386 by using negative feedback, as discussed here: http://www.diystompboxes.com/smfforum/index.php?topic=38331.0

Anyways, hope that's of some help.

[dano12]

The resistor in the Zobel network, R5 in the last schematic, should be 10 ohms, not 10k ohms.  I think this is scaled to an 8 ohm speaker load, but it'll probably be fine with 4 or 16.  It's purpose is usually to protect the chip amp, but it may also help take some of the top end fizz off.

I think some series resistance on the MPF102's gate may help with RF rejection.  To be really complete you could include a dc-blocking capacitor and protection diodes.

I'd also embellish the power supply a little.  I'd put C1 as close to the 386 as possible, and supply the battery V+ to C1's + terminal.  I'd use C1's - terminal as a star ground point for IC1 pin 4 (gnd), the gnd side of the speaker cable, and the gnd for the rest of the circuit.  I'd then use a small RC filter like 100 ohms/22uF to filter the power before Q1's drain, because I'd be worried that IC1 will be pulling enough current out of the power supply to overwhelm C1 and modulate the signal on to V+, which will feed back through the PS to Q1.  Oh, and it wouldn't hurt to make C1 bigger, like say 1000uF, and include a polarity protection diode.

My last suggestion is to "valvestate" the 386 by using negative feedback, as discussed here: http://www.diystompboxes.com/smfforum/index.php?topic=38331.0

Anyways, hope that's of some help.

Much help. Thanks for pointing out the 10 ohms. I've updated the schematic.

Thanks for the additional ideas--I'll start playing with them.

[DEZREJECT]

wow... dude looking at ur site... great site, but beyond that.. wow.. amazing pedals... that tubescreamer thing.. Thats effing amazing... where do you get those look meters?!!?! i loooove those... people at gigs would be like why the hell do you nede that many switches.. i love it.. total control!!!! :-D!!!!!!!!! i deff am gunna have to try to do something like that.. isaw your site a good long time ago.. but i couldnt find it again.. i remembered the kick ass enclousers.... sorry to be off topic.. but effing amazing bro