Balancing feedback loop clipping diodes amplitude?

[blearyeyes]

When soft clipping using the feedback loop of an Opamp, with 2 LEDs then switching to four IN914s, there is a large gain drop in the signal. How would you control the difference? Is there a way to maintain the clipping caricature but lower the amplitude of the LEDs side of the feedback circuit?

Dan

[stecykmi]

it depends on the topology somewhat, but you should be able to switch in a series resistor somewhere to change the gain of opamp. for example, you could switch in a resistor in series with the 1n914's to increase the gain, but switch it out for the LED's (as we know, opamp gain increases as the feedback resistor increases).

[midwayfair]

You could put a resistor in series with the diodes ... and consequently raise the forward voltage of the diodes to that of LEDs and make it pointless to use different diodes.

Or you can put a limiting resistor in to change the gain by a set amount when you switch the diodes ... and either lose the cleaner settings on the diode side or some volume/gain on the LED side, making it pointless to have the different diodes.

You could try to come up with some sort of calculation to switch the resistor on the output so that the volume of the whole pedal is lower when the LEDs are in the circuit ... but then you lose some maximum output from the LED side, which is sort of the whole point of using LEDs (so you can get more output from the pedal with less distortion).

You might see where I'm going with this. The whole point of the diode differences is to change the headroom of the circuit, so anything you do to adjust the gain is going to negate the reason you have a diode switch. And I might sound flippant, but this question comes up a lot, and I'm not sure I really understand the reason people don't just use the obvious solution: If you want the volume to be the same, adjust the volume knob. That's what it's there for.

[blearyeyes]

Well heck I guess a flippant answer is in order...In my 20 years of playing professionally I always tried to make life as simple as possible, playing in a power trio and playing keyboards and singing, the last thing I wanted to do was try to match volume in a distortion pedal using a knob on the floor... Actually I didn't use distortion pedals in the 60s and 70s, I had a batterey powered opamp in my guitar that would blow out the front end of any amp, along with all the pedals and such, I was using... Maybe I'm just odd?

So I want to understand this. The type of diode sets the headroom of the IC and the IC does all the clipping? That is not how I think it is...Or the diodes have an inherent clipping profile which includes their rise time and forward voltage which is controlling the gain of the opamp as well as supplying the shape of the clipping?

So what you are saying is that if I lower the voltage AFTER the clipping occurs with the LEDs and lower the overall gain of the opamp it will be the same as using silicon?

Please help me to understand this as I have not gotten any answer that makes sense to me... Basically all I have heard is you have to live with it. I'll go with that if that is how it is...

Thank you guys for your replies I really do appreciate it and am trying to learn...

Daniel Shattuck

[midwayfair]

Well heck I guess a flippant answer is in order...In my 20 years of playing professionally I always tried to make life as simple as possible, playing in a power trio and playing keyboards and singing, the last thing I wanted to do was try to match volume in a distortion pedal using a knob on the floor... Actually I didn't use distortion pedals in the 60s and 70s, I had a batterey powered opamp in my guitar that would blow out the front end of any amp, along with all the pedals and such, I was using... Maybe I'm just odd?

So I want to understand this. The type of diode sets the headroom of the IC and the IC does all the clipping? That is not how I think it is...Or the diodes have an inherent clipping profile which includes their rise time and forward voltage which is controlling the gain of the opamp as well as supplying the shape of the clipping?

So what you are saying is that if I lower the voltage AFTER the clipping occurs with the LEDs and lower the overall gain of the opamp it will be the same as using silicon?

Please help me to understand this as I have not gotten any answer that makes sense to me... Basically all I have heard is you have to live with it. I'll go with that if that is how it is...

Thank you guys for your replies I really do appreciate it and am trying to learn...

Daniel Shattuck

Diodes have a property called forward voltage: when you pass voltage through them, there's that much signal loss. When you pass voltage across them (your guitar signal in this case), they will clip the signal to no bigger than the forward voltage. Things do get slightly complicated when they're in the feedback loop of the diode, but this is still *generally* true.

Let's say you have a completely clean sine wave (you never do with guitars, but for example). Your guitar signal is perhaps a volt. Now you amplify it so that it's 4V peak to peak. That's still less than the power rails on the op amp (4.5V in either direction), so it's still clean.

Now you take a pair of silicon diodes. Their forward voltage is 0.6V, normal for a generic silicon diode. You put them in the feedback loop. Now, when you amplify the signal, it will never get larger than 0.6V peak to peak, because that's the limit set by the diodes. What happens to the rest? It's clipped off at the top. It's soft clipping in an op amp, so there's some weird compression stuff going on otherwise, and there's some clean bleed with a HUGE input signal, but for the most part your signal won't exceed that limit set by the diodes.

Essentially, the diodes are acting as an artificial limit on the headroom of the circuit, sort of as if you'd used an imaginary 1V or 3V battery instead of a 9V. In general a circuit simply can't reproduce a signal that's larger than the supply voltage provided by the power supply, and most of the time you're luck to get 90% of the total voltage (that's total voltage -- positive AND negative swing, so think 4V peak to peak as max on a good day). There actually is an exception to this, but it's not something you can just tack onto any given circuit. Anyway, the more your signal exceeds the headroom of the circuit, the more it's clipped, until you get a complete square wave. The clipping is heard as distortion.

Think about the difference in size of a signal that's 4V compared with one that's 0.6V -- it's almost 1/8 the size! That translates directly to a corresponding loss in volume. If you use two silicon diodes in series on either side, you'll get a little over a volt, still MUCH less volume than the diode-less signal size we were discussing.

LEDs have a much higher forward voltage, about 1.7-2V. That's three times the size of most silicon diodes, and still higher than your two diodes in series on each side. Again, that voltage difference corresponds directly with (a) less clipping at the gain levels and (b) more volume at similar levels of clipping.

In other words, if you want distortion, you are going to sacrifice volume. If you want volume, you are going to sacrifice distortion.

But let's go back to your original question. You want to "maintain the same clipping [character]" -- why are you switching the clipping diodes if you already have the clipping character you want? What exactly are you trying to accomplish? The reason for the diodes in the first place is to create the clipping. Are you switching between less clipping and more clipping?

You also say that you're trying to "keep things as simple as possible" -- are you switching the diodes in mid-set? You could mark your pedal for the two settings you use and you could set the volume at the same time you switch the diodes without having to fish around for the new setting.

But this goes back to you saying that you want to keep the same clipping sound ... the different sound of the clipping is directly related to the forward voltage and the size of the signal (gain) before it's clipped. LEDs clipping a signal three times their forward voltage will clip a signal pretty much exactly the same as a silicon diode clipping a signal three times their size, but the LEDs will be louder. The only reason to switch the diodes is (a) you need more volume at lower distortion levels (LEDs) or (b) you need more distortion at lower volume levels (silicon). Sorry if it sounds like I'm repeating some stuff, but I'm just trying to hammer home the fact that the diodes are only there to create a certain amount of distortion at a certain gain level.

Like I said, it's probably possible to change the output volume at the same time you change the diodes, but I'm not sure exactly why you'd do that -- and it's not going to be perfect in any case, because the volume change won't be identical at every gain and volume setting. You would still end up needing to change the volume most likely.

[blearyeyes]

But let's go back to your original question. You want to "maintain the same clipping [character]" -- why are you switching the clipping diodes if you already have the clipping character you want?


I want to offer different clipping character?

[blearyeyes]

not possible without volume change...right?

[jubal81]

I've found the simplest solution to this problem (two distortion levels-same volume) is a two-channel pedal. It can pretty easily be done in a typical 125 enclosure.

[blearyeyes]

Thanks David...

[chromesphere]

Thanks for the explanation Jon, you answered a question I had that popped up in my head today actually, cheers.

Hi Blearyeyes, welcome to the forum.  I think Jon's saying that the majority of the difference you will get between diodes is in the forward voltage and level of clipping that they offer (as he explained above, having the peaks chopped off).  So perhaps changing from leds to silicon diode + resistor will sound almost identical.  I assume you have to add the resistor to the silicon diodes to raise the voltage for the silicon's. I have found very little difference in clipping character from messing around with clipping diodes personally, but then I'm tonally / intellectually deaf.  If you wanted to try it I guess you could add a trimpot inline with the silicon diode to adjust the volume to exactly match the second diode set.

Paul

[blearyeyes]

Thanks Paul, I guess I thought they each had a sound apart from their forward voltage. Sorry if this has been gone over a million times..

[chromesphere]

Its a topic of debate I guess.  Some people say there is a very audible difference, some people not.  Depends on the application I guess.  Personally I think there are components that make more of a difference in a circuit and are better spent experimenting with then swapping diodes / leds but that's just my opinion.

[jubal81]

Thanks Paul, I guess I thought they each had a sound apart from their forward voltage. Sorry if this has been gone over a million times..

Different diodes with the same forward voltage can have a VERY subtle difference - much less than changing the value of a cap somewhere. That's a case where once the rest of the circuit is worked out, you can test and select which ones you prefer. Just switching diodes won't do what you're looking for because it changes the headroom - volume drop like you said.

If you want to switch between, say, a medium gain OD sound and higher gain OD sound, you can put two "pedals" in one enclosure and wire it so it switches between them, like having two channels.

There are 'shortcut' examples to two channels, like the Blackstone Mosfet Overdrive, which uses one circuit and switches pots to mimic two channels.

[blearyeyes]

Hey I want to thank you guys for your time and patients with a driven nubette.
Thanks for the excellent explanation midwayfair, I get it now. It's not about the diodes per say, it's how they interact with the circuit and a large part of the sound is based on the diodes forward voltage and when they start to conduct and where they clip. So if I modify the circuit so they are the same amplitude...then they are going to pretty much be the same. ahhh duh....

Dan 

[chromesphere]

Yeah I thinks so Dan.  Its an excellent first question by the way.  Once again welcome to (the best) forum (ever)