soft clipping

Is it as simple as I have read, a couple of dioodes in the feedback loop on an opamp or am I missing something here?

no...  not simple. Diodes are generally considered hard clipping but even that depends on how "hard" and what impedance you over-drive them with.

Soft clipping is probably trying to mimic old audio paths that went non-linear well before hard saturation. Modern solid state paths tend to saturate more suddenly. While I can't guess what exactly you are looking for.

JR

JR, do you know of any good peak limiters that can be DIY'ed?

dandeurloo said:

JR, do you know of any good peak limiters that can be DIY'ed?

Click to expand...

What do you mean "good".

I have designed several rather different ones using different strategies.

I usually roll a peak limiter into VCA based comps as a little extra crunching in the side chain.

I made one noise gate/limiter using cheap OTA wrapped around an opamp, so dry path was extremely clean, just the bare opamp. The limiter was only in circuit while limiting so distortion of the OTA was swamped by limiter gain modulation.

Cheap limiters using a single JFET and pair of transistors using Vbe for threshold is about as simple as you can get.

Most unusual peak limiter I did was for a cheap mutli-efx engine inside a powered mixer that had so much internal HF pre-empahis that it would overload internally if you looked at it cross-eyed (thank you peavey digital engineers). I connected a simple FET limiter to the overload LED output, presto giving me an extra bunch of headroom with spectral dependent limiting that responded to internal not external overloads.

JR

PS: If the digital engineers could light an LED they could reduce the pre-emphasis when the signal was hot and S/N was not a problem... Digital can be really clever if you take advantage of it's decision making capability.

Well lets see, as I understand it so far, lets say you have a sine wave going through an amp stage. There is a point where waveform distortion occurs.  Here is a photo to illustrate what I am asking.




I have seen  various schematics and trying to figure out how one would achieve such a thing. The short answer is overdrive a tube, but wondering how to get those operational characteristics out of a general purpose opamp

looks kinda like when one tweaks the bias pot on ba283.
how about a couple transistor stages cascaded with trim pots inserted to skew biasing...I have a pcb out of a shure mic mixer, 4 stages of 2? transistors could be an easy tweak. 

Hmmmm this page has a lot of information concerning the very subject. Ultimately for me I want a mic pre that has soft clipping built in.  Might give this a go and see what is what.

http://sound.westhost.com/articles/soft-clip.htm

The dbx 166 has a pretty good soft clipper circuit used for absolute peak limiting. The schematics floating around are a bit hard to read, but I've watched it working on  a scope, very nice rounded edge, hardly any odd harmonics.

pucho812 said:

Well lets see, as I understand it so far, lets say you have a sine wave going through an amp stage. There is a point where waveform distortion occurs.  Here is a photo to illustrate what I am asking.

I have seen  various schematics and trying to figure out how one would achieve such a thing. The short answer is overdrive a tube, but wondering how to get those operational characteristics out of a general purpose opamp

Click to expand...

You can literally make that transfer function with a diode break amplifier. Imagine a simple inverting opamp with a feedback resistor for low level part of your sine wave. Above a certain voltage diode steering put another resistor in parallel with the first for lower gain above that threshold. A similar symmetrical circuit for the negative swing.

Not sure that it would sound very good but you can make it.

JR

We were discussing the Aphex Dominator in another thread and I mentioned something about a soft clipper there.
I'm not a big fan of the Dominator itself as a band-split compressor, but the soft clipper on the end of it is actually quite good.
And yes... the clipper is just an op-amp with two red leds in the feedback loop.

I think if it's just used  to shave peaks off, then the clipper is pretty much inaudible, but if pushed then it will start to sound a bit crunchy.

pucho812 said:

Ultimately for me I want a mic pre that has soft clipping built in.

Click to expand...

It depends on what type of non-linear behavior you want.

In the simplest case you could just do 1k in series with two diodes to ground. But this will give you non-linear behavior over the whole range b/c the diodes will always conduct a little. And the clipping will be harsh. This is common in 80's guitar distortion pedals.

Diodes in the feedback path of an op amp will give you soft clipping because, as you overdrive more, the clipping stays rounded over a much larger range compared to grounded diodes. This are also very popular for guitar pedals. Perhaps more popular. This is used in the classic Tube Screamer pedal. I think it is supposed to mimic  something more like guitar power amp distortion.

For a mic pre my guess would be that you don't want any non-linear behavior at all under a certain signal range and then above that you want rounded clipping. For that, I would try transistors in the feedback path of an op amp where the bases tap the junctions of three resistors in the feedback path of the op amp. That way, under a certain signal range the transistors are either completely off or back biased and so you just have a clean op amp. Use a bi-polar op amp like the 5532 with small resistors and x-large bi-polar caps. That has the potential to give you very low noise which is the most important issue when building analog circuits AFAIC.  You could make a very simple non-inverting mic pre with one half of a 5532. Or use a 5534. The NE5534 is perfect for a stupid simple mic pre. When overdriven, at some point the transistors turn on and you get a soft clipping. This config was used in the Crate FlexWave guitar amp. Google for "crate flexwave schematic" in Google Images.

Small quote from www.geofex.com

"One recent entry is the diode-connected MOSFET transistor. This connection makes for a "diode" that has about a two volt forward drop, but a "knee" that is very softly curved"

I have used this with some nice results but in an overdrive circuit and not a mic preamp.

Here is the clipping info page :  http://www.geofex.com/effxfaq/distn101.htm

I have used BS170 / 2N7000 and IRF510/IRF520 for this purpose also in combination with 1N4148's and Ge diodes like 1N34a - reported
to have a quite soft "knee" also but their onset of clipping is much lower, 0.4v IIRC

Marty.

This is a variation on Douglas Self's feedforward clipper, which he designed to be a low-distortion Hard Clipper. The addition of the extra resistors in the feedback paths of the feedforward opamps softens up the clip pretty well.

magicchord said:

"a low-distortion Hard Clipper"

Click to expand...

An oxymoron, as far as I can tell. 

Yeah ... but I think the main point of "low distortion" is that the diodes are (practically) "out of the circuit", i.e reversely biased, until they "meet" the reference voltage, so in that area below the reference, they don't "fuzz" the signal very much ... so yeah, an epilady ...

You could easily "soften" the diode action with inserting some series resistance ..

distortion is the name of the game here... if you're clipping a signal, hard or soft, it will result in a distorted wave form. Maybe he meant "low 3rd harmonic clipper"

tv said:

Yeah ... but I think the main point of "low distortion" is that the diodes are (practically) "out of the circuit", i.e reversely biased, until they "meet" the reference voltage, so in that area below the reference, they don't "fuzz" the signal very much

Click to expand...

That's the case with any half-decent clipper, but "soft" means the transition zone is much larger than with a "hard" clipper. So the zone where distortion occurs is larger in a soft clipper than a hard clipper, but overall the distortion is less harsh with a soft clipper. As always, it's a trade-off.

Soft saturation: You mean like this?

pucho812 said:

Is it as simple as I have read, a couple of dioodes in the feedback loop on an opamp or am I missing something here?

Click to expand...

I used a VCA as a modulator and made the waveform modulate itself. The Waveulator: Two VCA-based Clipper Saturators - Pro Audio Design Forum

Jump to page 4 where you'll find some demo files: The Waveulator: Two VCA-based Clipper Saturators - Page 4 - Pro Audio Design Forum

Hard clipping without hard corners:

KA-Electonics.com said:

Soft saturation: You mean like this?

I used a VCA as a modulator and made the waveform modulate itself. The Waveulator: Two VCA-based Clipper Saturators - Pro Audio Design Forum

Jump to page 4 where you'll find some demo files: The Waveulator: Two VCA-based Clipper Saturators - Page 4 - Pro Audio Design Forum

Hard clipping without hard corners:

Click to expand...

When do you think this will be available?

Thanks for your interest!
I'm putting the final touches on the circuit design now and am trying to figure out which features are the most important.
I hope to have a layout ready by late July.