feedforward clipper

I'd like to use +/-5 volt supplies and clip at about 3 volts and stay non-inverting if that's possible

I guess without a stable reference voltage none of the clipper topologies can be independent of the supply...need to get more sleep

The threshold is set by R2 and R5 (for negative going input) and R3 and R7 (for positive going input). You are correct in that the supply rails set the threshold. The action begins when the input voltage divided by

10k (R5/7) exceeds 15V across 47kohm (R2/3) which is about 3.2V or 2.25 Vrms.

The feedforward bit is that R6/8=R5/7 and R1=R4 so once the threshold is reached the opamps send a current through R4 that balances the current through R1.

Note that this circuit loads the input and should be fed from a low impedance output.

HTH

piglet

Assuming the opamps can swing their outputs close enough to the rails then just change R2 and R3 to 16.67kohms. Using 16kohm would be 3.125V thresholds which may be close enough.

piglet

Got it, thanks!

Actually, even if the opamps outputs cannot get close to the rails you could still use this topology by lowering the thresholds to (say) 1.25V which is then in the territory of bandgap references and making up the circuit overall gain by increasing R9.

Still inverting overall - simple solution is just follow by unity gain inverting stage. Plenty of non-inverting versions surely exist, have a look at the active clamp shown in fig 4.61 of AoE 3rd Edition. :)

piglet

piglet

It's an interesting circuit, but for my purposes I think a more traditional feedback clamp like

would work best. I guess the feedforward circuit has less distortion for audio than the feedback versions? I'm not dealing with audio in general though so I'm not sure there's any advantage to the higher parts count...

why would it give less distortion? It will give different behaviour above the gbw of the opamp.

NT

** That is an odd circuit.

With no input, A1 draws current through D2 to cancel the current from R2 at its negative input while D1 removes the -0.6V DC offset created. A similar argument applies to A3 so there is no current in R4.

Meanwhile, A2 simply inverts any input signal.

Nothing changes until the input exceeds + or - 3.2V when R7 and R5 respectively are able to bias A1 or A3 into linear operation as unity gain inverters and deliver current to A2 via R4 such that it exactly cancels current via R1 from the input.

The 3.2V threshold is set by R2 & R3 ( plus the supply voltages) and they could be changed to cover the range from about 0.1V up to 13V.

... Phil

Above the clamp threshold any of these circuits will generate terrible distortion! :>

piglet

Thanks Phil. The circuit is from the book "Small Signal Audio Design" by Douglas Self, 2010. There are some interesting circuits in there, but it feels sometimes like there are a few that are really overcomplicated and from my perspective would offer rather hard to justify sonic improvements over more simple circuits given the parts count - for example like using a complicated arrangement of unity gain followers and discrete transistors to bootstrap the power supplies of an audio opamp to reduce common mode distortion.

It's not a bad book, I guess just some of it feels to be of limited practical value compared to say, AoE and kind of goes off the rails in places...

I don't think that circuit was ever wired up and checked. Note that D1 and D3 tie the OA outputs together such that, during "clipping", the OAs fight each other.

The feedback in the "clamp" is not in the signal path.

Set upper to +3V

Remove lower from GND, set at -3V

If you're dealing with frequencies that approach the Slew-Rate/Power-Bandwidth, there are patches to improve the "clamp" OpAmp's recovery time. ...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| San Tan Valley, AZ 85142     Skype: skypeanalog  |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 
              
I love to cook with wine.     Sometimes I even put it in the food.

If you want a clipper, why not just use diodes or transistor emitters to clamp a signal? Soft clipping should be better for audio.

Which reminds me of a clipper that I invented when I was a kid.

The zeners and the lower resistor are optional, in case opamp windup matters. Should work with mosfets.

There are some RRO opamps around these days that clip to their power rails and recover clean and fast. Same idea pretty much.

--

John Larkin         Highland Technology, Inc 
picosecond timing   laser drivers and controllers 

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com

The circuit I'm working on isn't handling audio directly, but it's a mod for another circuit involving control voltages; though these control voltages themselves may be at audio frequencies. There might be control voltages from 0 volts to 5 volts nominal to...some pathological voltage coming in from a broken piece of gear, but it needs to interface onto a uC bus where the uC is running from a 3.3 volt rail, so that CVs coming in don't blow the uC outputs connected on the bus.

So I'm thinking of the chain being like a 100 ohm/100p RC to filter any RF interference, then a series resistor and schottkys to the rails to clamp any pathological spikes, and then the positive opamp clamp to keep nominal signals within say a 0 - 3.3 volt range

Other solutions on the S.E.D/Schematics Page of my website...

...Jim Thompson

-- | James E.Thompson | mens | | Analog Innovations | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | San Tan Valley, AZ 85142 Skype: skypeanalog | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at

| 1962 | I love to cook with wine. Sometimes I even put it in the food.

I think the circuit is probably fine. D1 and D3 cannot conduct simultaneously as the input can never be simultaneously negative and positive.

piglet

Set the input to -10.

A3 out is clamped at +0.6

A1 out is trying to go positive, but if it does, D1 and D3 conduct and the opamps fight.

--

John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com

** Yep, a resistor is missing.

There needs to be another 4k7 that mirrors R4 so the outputs of A1 & A3 are not linked by diodes.

Likely just a drafting error, as with the pins numbers on A2.

.... Phil