4-Way Guitar Selector /w Gain -- Plz help me by reviewing this design

Thread Starter

jeghers

Joined Nov 14, 2008
7
All,

Here is my first draft for the design of my 4-Guitar Selector Switch with Gain. I am using OPA2134 op-amps for the audio buffering and CD4066 analog switches for the guitar selections. I am considering using a better analog switch to reduce THD. The design goals are described at the end of this posting.

I would be most grateful for any review/comments/suggestions you all might have. If there is a way I can improve signal purity, simplify, conserve power, whatever, I'd love to know anything I've overlooked.

Also, there are specific questions I am wondering about - if anybody has answers to these questions I'd be most grateful again.
-----------------------------------------------------------------------------------

Here are links to download my schematics, both in BMP and JPG format. JPGs are a little over 200k, BMPs are close to 400k. The schematic is in two parts: the audio path and the digital control...

Audio Path:
http://www.t4p.com/ftp/pub/3-Way Guitar Selector with Gain - Audio path.bmp
http://www.t4p.com/ftp/pub/3-Way Guitar Selector with Gain - Audio path.jpg

Digital Control
http://www.t4p.com/ftp/pub/3-Way Guitar Selector with Gain - Digital control.bmp
http://www.t4p.com/ftp/pub/3-Way Guitar Selector with Gain - Digital control.jpg
-----------------------------------------------------------------------------------
Here are questions I am hoping someone can help me with:

- Are my electrolytic cap polaities reasonable?

- I hear CD4066 switches still introduce some THD (enuf to negate the benefit of nicer op-amps), any other analog switches that might do better?

- Any better place in the audio path to place the analog switches? Any better way to turn the guitars on/off instead of CD4066s?

- Unity gain buffers offer hi-impedance inputs for the guitars, final op-amp mixer provides gain. Reasonable?

- Is the biasing of the op-amp inputs ok?

- Any decoupling caps that could be dispensed with? Any more needed? Are the values reasonable?

- I plan bypass caps between both power rails and ground, what values would be good?

- I plan to support both battery and extrnal power, would that affect what the bypass cap values should be?

- Would it be worthwhile to try using a single 9v power source, not bipolar (+9v/gnd/-9v)? I'd have to introduce a Vref with a power-draining resistor network, which I don't like.

- How much gain could I count on getting before the signal clips using just +9 and gnd? How much gain using bipolar (+9v/gnd/-9v)?

- Any ideas to reduce noise? To keep op-amps stable? To reduce power consumption?

- Any obvious flaws in the digital control? In OR mode, only one guitar at a time can ever be selected, in AND mode, any combination of guitars can be toggled on/off.
-----------------------------------------------------------------------------------

Here are the design goals in a nutshell:

- 4 guitars in, one low impedance output to a pedal/amp/whatever.

- Op amps allow high impedance guitar inputs so guitar tone is not adversely affected.

- The device is meant to NOT affect or "color" the tone of the guitars in any way.

- Each guitar input has a gain control (e.g. a volume knob)

- One more volume knob boosts overall gain somewhere between 3 and 12 db (I've not decided how much yet).

- Non-popping (debounced) solid-state switching with analog switches (probably CD4066s).

- Powered by either external or battery. While I am willing to design for 2 9v batteries, I'd kind of prefer just one 9v battery if I can get away with it. Power consumption is important but not paramount.

- Digital logic (with capacitor-debounced Schmidt triggers will provide with exclusive (1-of-4) or multiple (more than one combined) guitar selection.

Thans so much in advance,
/Mark
 

KL7AJ

Joined Nov 4, 2008
2,229
The 4066s are more likely to exhibit IMD at very LOW signal levels than at higher ones. You might want to try a little voltage gain before the 4066s if this is a problem. Using voltage followers like you have is a good idea for isolation...but you might want to make some provision for gain, as well, so you can optimize IMD and noise figures.

Best trick is to build it and see!

eric
 

KL7AJ

Joined Nov 4, 2008
2,229
I like your exclusionary logic for the guitars. I also like your option to make it an AND, as well. Very well executed!

eric
 

Thread Starter

jeghers

Joined Nov 14, 2008
7
You might want to try a little voltage gain before the 4066s if this is a problem.
I am wanting to avoid gain in the first stage because that would require adding gain-ratio resistors, which would have to be very high values in order to keep the 1Meg input impedance. That in turn would mean lots of high-value resistors in the audio path, and thus more resistor noise.

/Mark
 

Thread Starter

jeghers

Joined Nov 14, 2008
7
A nice project but too many questions in one post will leave most of us wondering where to start.
Sorry, dude, don't mean to be overwhelming :rolleyes: ... I just didn't know a better way to break it all apart.

I guess I'd welcome any observations that come to mind, even if they are fragmented or small.

/Mark
 

KL7AJ

Joined Nov 4, 2008
2,229
I am wanting to avoid gain in the first stage because that would require adding gain-ratio resistors, which would have to be very high values in order to keep the 1Meg input impedance. That in turn would mean lots of high-value resistors in the audio path, and thus more resistor noise.

/Mark
I'd ADD gain amplifiers after the buffers...not INSTEAD of.

eric
 

Thread Starter

jeghers

Joined Nov 14, 2008
7
I'd ADD gain amplifiers after the buffers...not INSTEAD of.

eric
Hmmm ... that'd cost me at least two more DIPs added to the PCB. Something to think about, though. This is to reduce low-level IMD thru the 4066s, right?

Alternately, is there simply a better analog switch I could use?
 

lespaul

Joined Jan 30, 2008
49
Hey Mark, good to see another guitar player out on this board. You have a very nice circuit going here, but the 4066 will have a bit of leakage current which may or may not rob some high end tone from your guitars. Everyone in the pro-music industry has been going to true-bypass switching and moving away from logic/active switching for this very reason. The best way to ensure that your tone stays true is to breadboard your project and make adjustments as required. Excellent choice in the OPA2134, but take a look at some discrete mosfet gain stages as they have a more musical tone in their output. Look at Jack Ormans effects pedal design sight for some examples. http://www.muzique.com/
 

nomurphy

Joined Aug 8, 2005
567
See attached markup.

Use the MAX333ACPP+-ND (available from DigiKey), and ground the normally-closed side of the switch, and then add a 1K at each of the mux outputs. This will ground the channel output when not used and prevent noise into U4.
 

Attachments

Last edited:
Top