How to do diode selection switch

[Antonio1963]

I am building a 855 drive pedal and I wanted to add a diode selection switch but I am not real sure how to do it. If I want to switch between two types what type of switch do I use? I tried to post the schematic but I can't figure out how to do it. Any help is really appreciated.
   

[nocentelli]

The diagram below is an easy way to get three clipping options: The stock Si diodes, Ge diodes or LEDs. If you only need two options, leave off the LEDs and just connect the switch: You can use a SPDT on-on switch in that case.



You will need to identify where the stock diodes connect to the output capacitor from the second opamp and use this point as your "in" connection to the switch: The ends of both clipping pairs can just be connected together and then connected to ground.

It is sometimes more convenient to mount the clipping pairs on the switch itself, so you might choose to do it this way instead (using a DPDT on-on):

[antonis]

<If you only need two options, leave off the LEDs and just connect the switch: You can use a SPDT on-on switch in that case.>

Or a SPST activating the lower forward voltage drop diode pair.. 

[Antonio1963]

Since the PCBs has extra holes where the diodes mount, is this where the switching curcuit is inserted?

[Mark Hammer]

<If you only need two options, leave off the LEDs and just connect the switch: You can use a SPDT on-on switch in that case.>

Or a SPST activating the lower forward voltage drop diode pair.. 

I was just about to suggest that myself, but then I started wondering: if I have a pair of diodes with a 600mv forward voltage, and then place a second pair in parallel that have a 300mv forward voltage, does the clipping threshold now become +/-300mv, as if the first pair weren't even there, or is there something else going on with diodes in parallel?

That question aside, another way of doing it is to put two pairs of diodes in series.  So, for example, a pair of silicon and a pair of germanium.  The toggle, whose "common" is connected to the junction between the pairs, can then be used to bridge/shunt one or the other pair, or neither.

[stallik]

One thing I don't like is the volume change that can happen when switching in the diodes. To get round this, I often wire a resistor into the 'clean' setting. Left side of the switch changes the diodes, right side bypasses the resistor when the diodes are in circuit. Change resistor value to balance the levels as required. Switch is on/off/on

[Mark Hammer]

I did something similar with a 4049 invertor-based overdrive some years back.  I had a DPDT rocker switch that could kick in a gain change with one half of the switch, and insert a resistance in series with the input of the volume pot to lower the output level, such that higher and lower-gain sounds had fairly similar volume levels.

[Antonio1963]

The diagram below is an easy way to get three clipping options: The stock Si diodes, Ge diodes or LEDs. If you only need two options, leave off the LEDs and just connect the switch: You can use a SPDT on-on switch in that case.



You will need to identify where the stock diodes connect to the output capacitor from the second opamp and use this point as your "in" connection to the switch: The ends of both clipping pairs can just be connected together and then connected to ground.



















It is sometimes more convenient to mount the clipping pairs on the switch itself, so you might choose to do it this way instead (using a DPDT on-on):

On my schematic, is C8 the cap you are referring to? I am trying to finish the rest of the board so I can hook it up to a breadboard so I can figure out the diode switch mod.

[idy]

The PCB you show in the photo has extra holes so that you can put two diodes in series, for a higher clip, on one or both sides. Either use just the outside holes for two diodes, use all four holes and four diodes. The "outside" holes are also connected, so one "top" hole and one "bottom" hole will go to your new fancy switching setup. There really are lots of ways to do this, you should grab some diodes and a socket and play with some options including assymetric: one diode on one side, two on the other, or one vs none, or a MOSFET with two legs tied together (it clips in both directions at different voltages...)

[antonis]

<If you only need two options, leave off the LEDs and just connect the switch: You can use a SPDT on-on switch in that case.>

Or a SPST activating the lower forward voltage drop diode pair.. 

I was just about to suggest that myself, but then I started wondering: if I have a pair of diodes with a 600mv forward voltage, and then place a second pair in parallel that have a 300mv forward voltage, does the clipping threshold now become +/-300mv, as if the first pair weren't even there, or is there something else going on with diodes in parallel?

Hmmmm.. You now make me to dig for diode dynamic resistance, p-n junction capacitance, reverse breakdown behavior and throw them into a diode small signal model...
(which, I ensure you Mark, I wont do..) 

[nation]

[/quote]
I was just about to suggest that myself, but then I started wondering: if I have a pair of diodes with a 600mv forward voltage, and then place a second pair in parallel that have a 300mv forward voltage, does the clipping threshold now become +/-300mv, as if the first pair weren't even there, or is there something else going on with diodes in parallel?

[/quote]

Interested to know this too. And does the differential in Vf matter for pedal purposes - say switching between a Schottky and Ge? If I recall correctly there were comments that difference in Vf needs to be higher for amp purposes for the parallel switch to effectively shunt the lower Vf diode pair.

[antonis]

If I recall correctly there were comments that difference in Vf needs to be higher for amp purposes for the parallel switch to effectively shunt the lower Vf diode pair.

Could you plz post a rough diagram of what you mean..??
('cause I can't figure a parallel switch which shunts the lower Vf diode pair without simultaneously shunting the higher Vf diode pair too..) 

[amz-fx]

I was just about to suggest that myself, but then I started wondering: if I have a pair of diodes with a 600mv forward voltage, and then place a second pair in parallel that have a 300mv forward voltage, does the clipping threshold now become +/-300mv, as if the first pair weren't even there, or is there something else going on with diodes in parallel?

Technically, there will always be some very low level of current flowing through the pair with the higher threshold.

For practical purposes, the higher pair have no influence on the sound and it is dominated by the lower threshold diodes.

The forward voltage of a diode is not fixed, and this is easily seen on the datasheet for the 1N4148, for example. At 200ma, the 1N4148 diode Vf is over 1v but if the current is only 1ma then the Vf is less than 0.625v. As the current goes lower, so does the diode's Vf.

regards, Jack

[Mark Hammer]

Which I gather is the basis for different pedal designs employing different series resistance values leading up to the diode pair to ground.  It would also be the basis for some designs using variable resistors in series with a diode pair to produce "soft" clipping.

And regards right back at ya, Jack.  I'd say "warmest" but we got another few inches of snow last night.  I was out shovelling the driveway at 10:00PM. 

[nation]

If I recall correctly there were comments that difference in Vf needs to be higher for amp purposes for the parallel switch to effectively shunt the lower Vf diode pair.

Could you plz post a rough diagram of what you mean..??
('cause I can't figure a parallel switch which shunts the lower Vf diode pair without simultaneously shunting the higher Vf diode pair too..) 

Sorry, wasn't correct with my wording. Just a traditional on-off-on switch with back to back diodes on the 3 throws.

[PRR]

> if I have a pair of diodes with a 600mv forward voltage, and then place a second pair in parallel that have a 300mv forward voltage, does the clipping threshold now become +/-300mv, as if the first pair weren't even there, or is there something else going on with diodes in parallel?

You are running a kiddie-ride. For *small* kiddies.

You have a bar 600mm high, and won't admit anybody taller.

You get a smaller kiddie-ride, and a bar 300mm high to limit kiddie-height.

It doesn't matter if the 600mm bar is still there. Only <300mm kiddies will be admitted.

OK, but diodes are not steel bars. More like foam. If you push (current) 10X less, a "600mm" bar will brush a 540mm kiddie; if a 660mm kiddie pushes 10X harder he will slip through. But for your 600 vs 300 example, these "soft threshold" differences won't matter. From 300mV to 600mV on the same diode is 32,000:1 of current, 90dB of SPL, so the effect of the higher diode is insignificant.

{hmmmm.... 300mm is an awful small kiddie to be walking into a ride.....}

[Mark Hammer]

Good.  Then I have no compelling reason to use a switch to select between two sets of diodes, instead of reverting from a default pair to something with a lower Vf.  I was concerned that, in my electronic naivete, there was some detail/feature about diodes I did not completely understand.

I am comforted. 

[amz-fx]

Which I gather is the basis for different pedal designs employing different series resistance values leading up to the diode pair to ground.  It would also be the basis for some designs using variable resistors in series with a diode pair to produce "soft" clipping.

That might be the intent but it really can't make a lot of difference. Going from 1000 ohms to 500 ohms in a typical diode clipper would change the peak clipping from 0.62v to 0.65v... less than 0.5db and inaudible to most of us.

Also consider that most of the opamps used in distortion pedals have limited current sourcing abilities and it becomes even less likely to be significant to the sound.

Best regards, Jack

ps: let the youngsters do the snow shoveling! 

[antonis]

in my electronic naivete

Are you kidding, Mark..?? 

"Naivete" has nothing to do with in-depth knowledge/analysis of semiconductor advanced physics..
For DIY pedal purpose, "2nd" small signal diode model (differencial internal resistance & forward voltage drop - no reverse leakage current, no internal capacitance, no switching & recovery delay time, no etc..) is more than adequate..

let the youngsters do the snow shoveling! 

I presume, despite his age  , Paul left no room for additional snow shoveling just above..

[amz-fx]