Diodysus design stage

[kierkegaard]

hey my friends

had an idea today whilst messing around with a breadboard, following a few tutorials and learning about a simple amplification circuit and had an idea, probably not news to you guys but id like to see it through.

https://imgur.com/a/7EQFLzf

here is the schematic, i basically took the simple amplifier and then ran it through two different pairs of germanium diodes to hear the difference, boosting again after to get some volume, my idea is to add maybe 5 or 6 different diodes on a multi switch so i can have access to lots of different subtleties in one box.

i would like to have a pot that controls the gain of the diodes, right now im driving them about half as much as i would like too, but im very green and have not much of a clue where to start with this. I read about putting the Diodes in series will decrease voltage loss (more volume) but also soften clipping (less gain)

i was also looking at the OG klon circuit and thought that maybe i should use TL072 instead of JFETS but that feels like a whole bunch more to learn about, maybe its unnecesary?

anyway, hope someone can see what im going for and give me a few pointers in the right direction

[antonis]

Diodes have no gain..

[kierkegaard]

the Klon has a gain pot though  . . . . thats using germ diodes in the same way right?

i dont know enough about schematics to look at one and work out how they did it

[toneman]

put a pot in series with the diodes, before the switch.

probably around 50K-100K

at full resistance, diodes are escentially out of the circuit.
as resistance is lowered, the diodes start rectifying.
I've never seen this in any pedal...U heard it here first!!!!
jfets have the "smooth" clippling; diodes more harsh.

[niektb]

Funnily, I designed a very similar circuit not so long ago!
Maybe it can give you some pointers! 
https://www.diystompboxes.com/smfforum/index.php?topic=124298.20

[kierkegaard]

Funnily, I designed a very similar circuit not so long ago!
Maybe it can give you some pointers! 
https://www.diystompboxes.com/smfforum/index.php?topic=124298.20

ah ace yeah this looks cool!

what is R17 doing and is C4 important?

so looks like you are basically sending the pot to ground through the switch? i did try putting a pot on before the switch but it didnt work properly, i dont really understand schematics and first time using breadboard so youll have to be patient with me!

so saturation dials in the diodes and gain dials in the transistor OD?

[kierkegaard]

put a pot in series with the diodes, before the switch.

probably around 50K-100K

did try this but ended up giving me bare signal in the middle and silence at both ends

had it hooked up as: Lug 1 - ground   Lug 2 - right side of C2   Lug 3 - centre pin of switch

I've never seen this in any pedal...U heard it here first!!!!

not sure if this is sacastic or not

[FiveseveN]

had it hooked up as: Lug 1 - ground   Lug 2 - right side of C2   Lug 3 - centre pin of switch

That's not how you do it. See AMZ Warp Controls.

[kierkegaard]

That's not how you do it. See AMZ Warp Controls.

ah ok ace thanks for that, so ive simplified my circuit so i can work out whats going on, run before walk etc

https://imgur.com/a9zsWD8

i believe i now have the pot in series properly, but its not acting how i would expect, its acting like an on/off, at full it sends through the diodes but if i easy off on the pot slightly it drops back to clean signal

also i feel like the diodes can be driven alot more than they are, what would i add to the circuit to give it more oomph going through them

ill start messing with the warp controls once i get this bit down i think

[kierkegaard]

okaaaay quick update

i used a 50k pot and now it works as ablend which is coool!

im eventually going to have 8 different clipping stages that i can switch between

they are all gonna have different relative volume and someone mentioned to me about having 8 seperate amp stages and switching between those

anyone got any thoughts on the best way to do that? also still keen to know drive these diodes a bit more

[KahunaSapien]

Have you got a smaller value pot then the 500k? The range of resistance required to turn on the diodes is a small fraction of the 500k

A visual example would be a LED running off 9v would have maybe a 470 up to about 1500 ohms depending on how bright you want it.  it's probably off at a around 10k

The turn-on "knee" area on the diode curve as it switches on needs to be stretched across  the range of the pot to give the gradual increase in current as the signal voltage rises 

Try soldering a small value resistor between the pot wiper and the top lug that is not connected to the diodes  this way when the wiper is at the bottom end the resistor will be in parallel with the 500k and reduce the value of the pot.  Try a 1k to start.  If the rotation range where the diodes "turn on" is still too small try a smaller resistor or vise versa. Experiment till you get a good range of control over the diode clipping  then replace the 500k pot and resistor combo with a pot of the same or close value (or just leave it in)

[patrick398]

If you want to drive it harder you could try use a simple op amp distortion set up, look at the MXR 250+, use a pot instead of fixed resistor in the the feedback loop of the op amp for variable gain

[antonis]

Replace 500k pot with a 10k one..
Personally, I wouldn't use anything higher than 5k, at least for a single pair of diodes..

You see, diode dynamic resistance is considered very low so you can safely ingore it even for 1% pot turn (5k)..
(vast majority of pots accuracy falls well above 1%..)
So, clipping level "off-set" (voltage added on diode forward voltage drop due to additional voltage drop on pot resistance) is big enough, even or small wiper travel, to bring clipping level up to clean signal one..

You can confirm it by replacing diode(s) with a 600mV (Si) or 250mV (Ge) voltage source, FET Drain with an about 4.5V voltage source and apply KVL between Drain and Ground..

[kierkegaard]

If you want to drive it harder you could try use a simple op amp distortion set up, look at the MXR 250+, use a pot instead of fixed resistor in the the feedback loop of the op amp for variable gain

ah ok yeah i see! so the feedback loop is the 9v to drain?

if ive found that at above 10k the transistor starts to cut out, would i use a B10k pot?

or does it act differently, i feel liek ive already got the transistor pushed as high as it will go in terms of that drain resistor, so wouldnt adding a pot just give me control rather than the ability to push it harder?

[niektb]

Funnily, I designed a very similar circuit not so long ago!
Maybe it can give you some pointers! 
https://www.diystompboxes.com/smfforum/index.php?topic=124298.20

ah ace yeah this looks cool!

what is R17 doing and is C4 important?

so looks like you are basically sending the pot to ground through the switch? i did try putting a pot on before the switch but it didnt work properly, i dont really understand schematics and first time using breadboard so youll have to be patient with me!

so saturation dials in the diodes and gain dials in the transistor OD?

Yes to your last question. R17 lowers the resistance when the diodes are fully turned on. Very useful if you want to prevent volume drops.
Also, put a DC-cap between the output of the FET and the diodes. The output from the FET is floating around 5V DC (round and about) but you try to clip to ground.

500K pot is way too big, that ain't gonna work.
You could lower R1 (but not lower than 33k) to get a little bit more volume. You could also turn the 1Meg into a pot to control gain

[patrick398]

If you want to drive it harder you could try use a simple op amp distortion set up, look at the MXR 250+, use a pot instead of fixed resistor in the the feedback loop of the op amp for variable gain

ah ok yeah i see! so the feedback loop is the 9v to drain?

if ive found that at above 10k the transistor starts to cut out, would i use a B10k pot?

or does it act differently, i feel liek ive already got the transistor pushed as high as it will go in terms of that drain resistor, so wouldnt adding a pot just give me control rather than the ability to push it harder?

I mean using an op amp, like the mxr 250 does, no transistors in that circuit. Varying the resistance in the feedback loop (output of opamp back into one of the inputs) varies the gain of the amplifier.

https://www.electronics-tutorials.ws/opamp/opamp_2.html

Look at the clipping stage of the mxr 250:
https://i.stack.imgur.com/qJTtf.png

You basically want to amplify and then clip with diodes to ground. Except you want to be able to switch between different diodes sets, you could do this with a rotary switch or three position toggle if you wanted less options.

You could use a dual opamp, the first for gain as in the mxr 250 and the second as a recovery stage to account for losses in certain diode set ups. Using 8 different amps as you suggested would be madness ha.

Here's some words from Mr Hammer on diodes:
https://www.diystompboxes.com/pedals/diodes.html

and everything you need to know about diode clipping from Mr Orman (AMZ):
http://www.muzique.com/lab/zenmos.htm

[kierkegaard]

You could use a dual opamp, the first for gain as in the mxr 250 and the second as a recovery stage to account for losses in certain diode set ups. Using 8 different amps as you suggested would be madness ha.

amazing thank you! ill read all that this evening

yeah i thought 8 amps would be madness, but then again i was talking on a synthesizer forum and its a bit less compact over there. I have a TL072 so i could try that out, looking at the datasheet its not marked as drain/gate/source which is what ive been working by, so will need to do some more research!

the only issue i see would be big differences in volume of the different diode sets schotkky - Silicon, so was wondering if there was a way of locking them in

[idy]

One simple way to even out the level of the different clippers is to double or triple the low ones, like three silicon in series vs one LED, or two germanium vs one silicon.

The other low tech way to do it is to use a double pole switch for the diode pairs with the second pole adding a set of "padding" resistors before the volume control/output. If no volume control you would add a fixed R to ground and then choose series Rs.

The simplest way to get more gain would be making a second identical transistor amp like the one you have. Either a volume control (usually marked "gain") between them,...
...or a pot in series with a "source resistor bypass cap." (Up the source resistor to 12k, put a 50k pot in parallel with a big 10-20uf cap from wiper to ground.) Or google JFET booster...

[kierkegaard]

One simple way to even out the level of the different clippers is to double or triple the low ones, like three silicon in series vs one LED, or two germanium vs one silicon.

The other low tech way to do it is to use a double pole switch for the diode pairs with the second pole adding a set of "padding" resistors before the volume control/output. If no volume control you would add a fixed R to ground and then choose series Rs.

The simplest way to get more gain would be making a second identical transistor amp like the one you have. Either a volume control (usually marked "gain") between them,...
...or a pot in series with a "source resistor bypass cap." (Up the source resistor to 12k, put a 50k pot in parallel with a big 10-20uf cap from wiper to ground.) Or google JFET booster...

JFET booster is GREAT! crazy what adding a single cap in there can do, i dont know why its doing it but it worked

regarding getting a pot on there, i put one in series, so pin 1 - ground  pin 2 - 10uf cap   pin 3 - source resistor and it works but its only the top end of the pot that does anything, i used a 10k pot, need smaller?

regarding what you wrote, i think i put one in parallel but it jsut acted like a volume, im sorry im incredibly new, could you elaborate what you mean by 'wiper to ground' im still recognising my pins as           Pin 1 - left position     Pin 2  -  Mix     pin 3  -  right position        and i know its wrong, but its knowledge i havent yet investigated

[niektb]

One simple way to even out the level of the different clippers is to double or triple the low ones, like three silicon in series vs one LED, or two germanium vs one silicon.

The other low tech way to do it is to use a double pole switch for the diode pairs with the second pole adding a set of "padding" resistors before the volume control/output. If no volume control you would add a fixed R to ground and then choose series Rs.

The simplest way to get more gain would be making a second identical transistor amp like the one you have. Either a volume control (usually marked "gain") between them,...
...or a pot in series with a "source resistor bypass cap." (Up the source resistor to 12k, put a 50k pot in parallel with a big 10-20uf cap from wiper to ground.) Or google JFET booster...

JFET booster is GREAT! crazy what adding a single cap in there can do, i dont know why its doing it but it worked

regarding getting a pot on there, i put one in series, so pin 1 - ground  pin 2 - 10uf cap   pin 3 - source resistor and it works but its only the top end of the pot that does anything, i used a 10k pot, need smaller?

regarding what you wrote, i think i put one in parallel but it jsut acted like a volume, im sorry im incredibly new, could you elaborate what you mean by 'wiper to ground' im still recognising my pins as           Pin 1 - left position     Pin 2  -  Mix     pin 3  -  right position        and i know its wrong, but its knowledge i havent yet investigated

How need to wire it is: Pin 1- 10uF cap, Pin 2 - Pin 1 (yes, connected together), Pin 3 - JFET source