Crude Input Limiter

[Bill Mountain]

Last question for the day.  I promise!

I will not be able to physically work on any circuits for a little while (much needed housework) but I was wondering about a crude limiter idea I have.



This would only really work with high output instruments and I'd have to use Schottky or Ge diodes.

What effect would the pot under the diodes have on the input impedance?

[R.G.]

What effect would the pot under the diodes have on the input impedance?

Before the diodes begin conducting - no effect at all. That's a simplification, but a useful one. When the diodes begin conducting, the resistor acts as though it's connected directly to the input signal, so the impedance drops by that amount in parallel with the input.

Note that this happens on an instant-by-instant basis.

[FiveseveN]

Many circuits have this kind of protection. Blackstar pedals and the HT-5 (the latter with 12V Zeners), various preamps, CMOS inputs etc..

[CynicalMan]

For it to work all the time, you'll want a resistor in series with the input. If the previous stage can drive the diodes and resistor, you won't hear much of an effect. However, if you only plan to use this after a passive guitar, then you don't have to worry about that.

[Bill Mountain]

For it to work all the time, you'll want a resistor in series with the input. If the previous stage can drive the diodes and resistor, you won't hear much of an effect. However, if you only plan to use this after a passive guitar, then you don't have to worry about that.

I plan to use it after a passive bass guitar but I'll want to experiment with other effects.  What does the resistor do?  Can it be small enough to not really effect the output volume of the bass?

[R.G.]

Many circuits have this kind of protection. Blackstar pedals and the HT-5 (the latter with 12V Zeners),

Note that in those instances, the clipping is after some series resistance at the input. The input resistance limits the currents in the clippers, while the clippers limit the voltage on the inputs; exactly as CM says.

I plan to use it after a passive bass guitar but I'll want to experiment with other effects.  What does the resistor do?  Can it be small enough to not really effect the output volume of the bass?

If the stage following the series resistor has a high enough input impedance, it has essentially no effect at all. The input series resistor could be, say, 10K. The input to the next stage, perhaps 1M. The overall loss from that is 1M/(1.01M) so the signal to the input is 0.9900990099... of the original signal. I'd be willing to bet that a human can't hear that amount of difference reliably.

[Keppy]

But wouldn't the input resistor create a voltage divider once the diodes begin to conduct? That's what I thought the idea was, creating a voltage divider that wouldn't kick in until the diodes conduct.

[Bill Mountain]

This idea intrigued me too much to let it go so I breadboarded a quick transistor boost and put two 5158 Schottky's with a spdt switch to select between ground or a 10k resistor.  I must admit it sounded just fine and it lightly compressed my signal and kept it from clipping the transistor (there was of course distortion from the Schottky's depending on the guitar I used and how hard I played). The switch allows me to select between two gain/threshold levels.  I tried a 10k pot but it was mostly noticeable at the extremes so I just put in a switch instead.  I guess the impedance dropping when the diodes clip is not audible because the signal is already being squashed when it clips.  So I call it a success!

[R.G.]

This idea intrigued me too much to let it go so I breadboarded a quick transistor boost and put two 5158 Schottky's with a spdt switch to select between ground or a 10k resistor.  I must admit it sounded just fine and it lightly compressed my signal and kept it from clipping the transistor (there was of course distortion from the Schottky's depending on the guitar I used and how hard I played). The switch allows me to select between two gain/threshold levels.  I tried a 10k pot but it was mostly noticeable at the extremes so I just put in a switch instead.  I guess the impedance dropping when the diodes clip is not audible because the signal is already being squashed when it clips.  So I call it a success!

Great! Engineering is all about what works. If it sounds good, do it. 

[Johan]

if you ad a series resistance on the input, your pot (in the schematic) becomes a ratio pot with the threshold set by the diodes forward voltage drop. lets say 10k series R and a 10k pot.
if you look back about a year and a half, I had an overdrive project up where the last stage was similar for some soft clipping/compression effect
J

[armstrom]

If you want something crude could you use a comparator to switch on an LED coupled to a LDR serving as a variable resistor to ground on the input stage of the power amp? You would have to provide a reference voltage to the negative input of the comparator equal to what you want your threshold to be (allowing for some fudge factor). Again, just off the cuff here.. but if you used only one comparator you would only be detecting over voltage for half of the waveform, you could either rectify the wave or use two comparators (one for each half of the waveform). The general idea is that when the signal exceeds the threshold voltage the comparator brings its output high, turning on the LED which in turn decreases the resistance of the LDR shunting more of your signal to ground, limiting the voltage input to the next stage. Any signal that does not peak above the threshold will have no resulting change to the output (the LED will remain off). Once the signal breaks the threshold voltage it will begin pulsing the LED at a rate proportional to the audio frequency. However, low frequencies will have fewer, longer pulses while high frequencies will have more, shorter pulses. If the input to this whole thing is taken from a point AFTER the LDR connection to ground it should behave in a closed loop system attempting to bring down the signal level until it is below the voltage threshold.

At least that's the theory it's very crude and I have never tried it. I know some older opto-compressors use a similar system but I believe they still try to make use of an RMS level detector. This idea may not work at all
-Matt