Hi, I've got a similar situation to this post (https://www.diystompboxes.com/smfforum/index.php?topic=91618.0) where I was testing asymmetrical clipping of one side only purely to experiment. However as the post explains I ran into the issue of accidentally creating a diode clamper circuit (when attempting to clip only one side, the signal is not clipped but shifted by a DC voltage). This is because of the capacitor between the op-amp and clipping diodes.
Any idea what my options are to actually achieve clipping instead of clamping? I had the following ideas:
1. Operate the op-amp using a dual rail supply to avoid the need of an output capacitor?
2. Increase the value of the output capacitor, add a parallel resistor to the clipping diodes and increase the series resistor. These could help slow down the capacitor charging time and ultimately not achieve a clamping circuit.
Thanks!
Could we see the complete circuit..?? :icon_wink:
P.S.
It (http://www.muzique.com/lab/warp.htm) might be useful..
How about connecting the diodes to VCC/2 instead of the ground? I mean to the middle of the voltage divider you use to bias your opamps. Then move the output cap after the diodes if there is none there.
Quote from: antonis on November 17, 2023, 05:35:14 AMCould we see the complete circuit..?? :icon_wink:
You can refer to this schematic (https://www.electrosmash.com/images/tech/pro-co-rat/pro-co-rat-schematic-parts.jpg) and consider only one hard clipping diode to ground.
Quote from: Elektrojänis on November 17, 2023, 05:41:18 AMHow about connecting the diodes to VCC/2 instead of the ground? I mean to the middle of the voltage divider you use to bias your opamps. Then move the output cap after the diodes if there is none there.
That's a simple idea which makes a lot of sense. I think that it should work. Will test it out and report back. Thanks!
If you have followed the Rat circuit faithfully, it should clip, not "clamp". But you should explain what you mean by "clamp". Keep in mind that the gain applied to the 308 chip is so much greater than it can provide, at a certain point, no matter how many diodes you use, you're not going to get any more volume.
As Mark alludes to, you may still be getting a lot of clipping from the opamp, regardless of your diodes. That said, have you tried extremely asymmetric diodes, such as 1n34a or BAT41 on one side and 3 1n914/1n4148 on the other?
This (https://www.physics-and-radio-electronics.com/electronic-devices-and-circuits/rectifier/clampercircuits.html) is what I meant by a diode clamping circuit and this only happens when I try clipping only one side of the wave. This is where it would differ to an original RAT schematic.
(http://mrotqch.web.fc2.com/stomp/Clip/asym_clp_17.png)
QuoteYou can refer to You can refer to this schematic (https://www.electrosmash.com/images/tech/pro-co-rat/pro-co-rat-schematic-parts.jpg) and consider only one hard clipping diode to ground. schematic and consider only one hard clipping diode to ground.
I there a reason for adding the volume
after the output buffer?
No reason other than buffer stage input impedance..
Quote from: Lino22 on November 20, 2023, 12:11:53 AMQuoteYou can refer to You can refer to this schematic (https://www.electrosmash.com/images/tech/pro-co-rat/pro-co-rat-schematic-parts.jpg) and consider only one hard clipping diode to ground. schematic and consider only one hard clipping diode to ground.
I there a reason for adding the volume after the output buffer?
Or why volume pot is not smaller
Or why there is a buffer at all
Quote from: Steben on November 20, 2023, 02:28:44 PMOr why there is a buffer at all
I'd say the output impedance would be pretty big with the 100k tone pot in series ... at least for frequencies the 3n3 won't pull down.
Quote from: Lino22 on November 20, 2023, 03:35:22 PMQuote from: Steben on November 20, 2023, 02:28:44 PMOr why there is a buffer at all
I'd say the output impedance would be pretty big with the 100k tone pot in series ... at least for frequencies the 3n3 won't pull down.
The volume pot will "usually" be set rather low. It's impedance in series usually high.
Quote from: Steben on November 20, 2023, 02:28:44 PMOr why there is a buffer at all
A passive circuit like this clipping configuration has a significant output impedance.. :icon_wink:
Quote from: redbagy on November 17, 2023, 10:40:11 AMThis (https://www.physics-and-radio-electronics.com/electronic-devices-and-circuits/rectifier/clampercircuits.html) is what I meant by a diode clamping circuit and this only happens when I try clipping only one side of the wave. This is where it would differ to an original RAT schematic.
All normal behaviour.
The key point is the opamp is AC coupled to the single diode via the AC coupling cap. The cap has to charge up, more or less peak detecting. The cap charges to the point where the diode no longer does anything anymore. What you get is a DC voltage shift at the diode and if the next stage is AC couple that DC shift is ignored.
If you DC couple the diode like in m4268588's pics then you can clip one-sided without the cap charging effect.
There is an "in between" case in that a DC coupled resistance, say in parallel with the diode, provides a path for the cap to discharge. The discharge time is determined by the RC time constant. A small R or C discharges quicker and is more like the DC coupled case.
You also can get transient clipping effect when the signal is initially applied and before cap charges. The time constant for that is determined by the series resistance and the AC coupling cap.
Quote from: Rob Strand on November 20, 2023, 05:48:22 PMIf you DC couple the diode like in m4268588's pics then you can clip one-sided without the cap charging effect.
Can someone repost this please? As all I can see is an empty comment for whatever reason.
Quote from: Rob Strand on November 20, 2023, 05:48:22 PMThere is an "in between" case in that a DC coupled resistance, say in parallel with the diode, provides a path for the cap to discharge. The discharge time is determined by the RC time constant. A small R or C discharges quicker and is more like the DC coupled case.
In fact this is how I solved it for now but I'm still curious about other and possibly better solutions.
This is the schematic (https://aionfx.com/app/files/docs/prism_documentation.pdf) that I'm currently modding. I've split C5 into two capacitors (also added a 470Ohm resistor after the op-amp output) and placed hard clipping diodes. If m4268588 is suggesting to DC couple the end of the hard clipping diodes to VB instead of Ground, I've tried this (in this case I left the op-amp output DC coupled) however I needed to setup an 'external VB' circuit as the diodes were shifting IC1A non-inverting terminal then. I think that something can be done about this but I haven't looked further into this so far.
Quote from: redbagy on November 21, 2023, 02:03:03 AMCan someone repost this please?
https://mrotqch.web.fc2.com/stomp/Clip/asym_clp_17.png
Yes this option will definitely work as well as I tested it. Thanks :)
What's the reason for not being able to use the circuit on the right however? To isolate from VB in this case or because it's not ideal?
Didn't you just say it was messing with the bias? "shifting IC1A non-inverting terminal"
I am sorry guys, why would i want a hard clipping stage that has DC on it, messes with it and thus needs a buffered Vref, instead of a hard clipping stage decoupled by a capacitor(s), so it really clipps just the signal? I am getting lost here.
Quote from: FiveseveN on November 21, 2023, 03:44:55 AMDidn't you just say it was messing with the bias? "shifting IC1A non-inverting terminal"
Yes, I just wanted to confirm whether the solution applies to this particular schematic only or in general.
Cheers for the help!
If diode clip in NFB of Op-Amp, bias shift has no other meaning.
If Op-Amp clipped, Vref disturbance causes unexpected results. (Gated etc...)