Looking for ideas for a simple clip indicator

[Bill Mountain]

It could be something crude.  I have visions of a transistor hooked up as a switch between the battery and an LED but I'm not sure how to turn it on at a certain signal level.  I've also though about splitting my signal and sending one side to an opamp set for enough gain to clip an LED when my through signal reaches a certain threshold.  I've seen some more convoluted circuits with 555's but I didn't fully understand them at the time.  If I could remember where I saw them I may have a better understanding by now.  Any other ideas out there?

I'm actually thinking of using this in a dirt pedal because I want some consistency between guitars with different outputs.  I have a sensitivity control and I want a visual reference for when I'm adjusting to different instruments.

[R.G.]

Look at the LM3916.

[toneman]

Look at the datasheets for an LM311 comparator.

Also, look at some of craig andertons circuits.  some contain a clip indicator led circuit.

Also, here:

http://www.pavouk.org/hw/clip/en_index.html

http://partselectronic.blogspot.com/2009/03/audio-clipping-indicator.html

http://www.siliconchip.com.au/cms/A_30810/article.html

or just google:
"clipping indicator circuit"

afn
T

[Bill Mountain]

Thanks for the links.

I've actually googled it several times and everything seems more complicated than the pedal itself so I was looking to see if there were more basic designs than the standard ones I keep finding.

So now I need to read and digest.  I think I'm still going to try my extra opamp idea as well.

Thanks!

[Bill Mountain]

I think this takes the cake for simplicity:

http://www.siliconchip.com.au/cms/A_30810/article.html

[R.G.]

It's really easy to set something up to turn on an LED when a signal reaches a certain size (N.B. Generally voltage, but there are current-size sensors as well). This can be as simple as a single NPN transistor - or JFET, or MOSFET, or whatever. The problem only comes when you want to specify *what* signal size makes it come on, especially if that's a small signal, and you also want to have that be a variable small signal. It gets a bit touchier if you also don't want to load that signal, and want to specify a minimum on time for the LED so that the human eye can see it when it just starts to come on.

The Silicon Chip article has the advantage of working with signals where the drop across a 4.7V zener is the fundamental threshold. The signals it senses are significantly larger than that. If your pedal deals with 10-20V signals, sure, a threshold detector is trivial. If you're trying to sense the difference between an 85mV signal and a 100mV signal, a 4.7V zener and the uncertainty in its conduction voltage are going to make this a challenging task. Sensing signal thresholds in microvolt signals is a whole different world. Sensing signal thresholds in AC power voltage only requires a neon bulb and two resistors.

The spare-opamp (which will wind up being used as a comparator) or the comparator approach is probably the cheapest compromise, especially if you already have a spare opamp.

There is a cost of precision, a cost of adjustability, a cost of non-interference (i.e. non-loading the signal being sensed) and a cost of dealing with small signals.

[Bill Mountain]

It's really easy to set something up to turn on an LED when a signal reaches a certain size (N.B. Generally voltage, but there are current-size sensors as well). This can be as simple as a single NPN transistor - or JFET, or MOSFET, or whatever. The problem only comes when you want to specify *what* signal size makes it come on, especially if that's a small signal, and you also want to have that be a variable small signal. It gets a bit touchier if you also don't want to load that signal, and want to specify a minimum on time for the LED so that the human eye can see it when it just starts to come on.

The Silicon Chip article has the advantage of working with signals where the drop across a 4.7V zener is the fundamental threshold. The signals it senses are significantly larger than that. If your pedal deals with 10-20V signals, sure, a threshold detector is trivial. If you're trying to sense the difference between an 85mV signal and a 100mV signal, a 4.7V zener and the uncertainty in its conduction voltage are going to make this a challenging task. Sensing signal thresholds in microvolt signals is a whole different world. Sensing signal thresholds in AC power voltage only requires a neon bulb and two resistors.

The spare-opamp (which will wind up being used as a comparator) or the comparator approach is probably the cheapest compromise, especially if you already have a spare opamp.

There is a cost of precision, a cost of adjustability, a cost of non-interference (i.e. non-loading the signal being sensed) and a cost of dealing with small signals.

These are all great points.  My idea was basically to set the pedal up to work best with about 500mV's.  Then I would tap the signal after a boost and multilpy that (by about 4) to get it to clip an led.  With the Silicon Chip idea, couldn't I use smaller Zeners or maybe lower threshold diodes?

[R.G.]

My idea was basically to set the pedal up to work best with about 500mV's. Then I would tap the signal after a boost and multilpy that (by about 4) to get it to clip an led.

OK. So the signal you're already sensing would be 500mV times (a boost) and ?already buffered? That gets you up into the volt-or-two range, which is handy to work with.

Doing a simple amplification to turn on an LED is possibly workable, but it may not give you what you're looking for. LEDs turned on very softly put out very little light, and they may be on for so short a time that they're hardly visible, even if they're bright. I would expect the scheme you're describing to give a very soft fade-in to the LED. You might like that, but it's not going to be a threshold for signal size. It'll likely be a kind of an offset-zero brightess-indicating meter. It will appear to fade in, I think.

In my mind, indicator means that I get a binary or nearly so blip that says on or not. Hence the comparator thoughts. I suspect that the 555 circuits you refer to use the 555's timing setup to give a minimum on-time for one LED blip and I'm certain that it's not going to fade in. It'll be full on for the time it's triggered.

With the Silicon Chip idea, couldn't I use smaller Zeners or maybe lower threshold diodes?

Yep, you could. However, diode thresholds start getting quantized as you go lower. Zeners down to about 2.? volts exist, as I remember. Then you get LED forward voltages, silicon carbide, silicon, silicon schottky, germanium, etc. There are some exotic IV-V semiconductor setups with other voltages, but you're not likely to find those. You can also get bandgap reference diodes like the TL431 for 1.25V and up. There's a world of possibilities, but IMHO you're pushing too hard on a non-precision setup to try to force some precision on it.

It's worth tinkering with if you like the approach, but I would go for a comparator with some hysteresis, a couple of diode junctions for a reference, and a pot to set threshold. The comparator could be a spare opamp if you are careful about it.

There isn't any right or wrong here. I just prefer other approaches for better precision, human factors, and predictability. That's one reason I mentioned the LM3916. Sure, it seems complicated to use a whole IC for it, but if you hook it up in the minimal configuration, it gives you a solid LED drive with comparator thresholds and hysteresis. You don't have to use all those LEDs, just the one threshold you want. As I remember, it winds up being the LM3916, four resistors and the LED, maybe some filter/bypass caps.

[Bill Mountain]

My idea was basically to set the pedal up to work best with about 500mV's. Then I would tap the signal after a boost and multilpy that (by about 4) to get it to clip an led.

OK. So the signal you're already sensing would be 500mV times (a boost) and ?already buffered? That gets you up into the volt-or-two range, which is handy to work with.

Doing a simple amplification to turn on an LED is possibly workable, but it may not give you what you're looking for. LEDs turned on very softly put out very little light, and they may be on for so short a time that they're hardly visible, even if they're bright. I would expect the scheme you're describing to give a very soft fade-in to the LED. You might like that, but it's not going to be a threshold for signal size. It'll likely be a kind of an offset-zero brightess-indicating meter. It will appear to fade in, I think.

In my mind, indicator means that I get a binary or nearly so blip that says on or not. Hence the comparator thoughts. I suspect that the 555 circuits you refer to use the 555's timing setup to give a minimum on-time for one LED blip and I'm certain that it's not going to fade in. It'll be full on for the time it's triggered.

With the Silicon Chip idea, couldn't I use smaller Zeners or maybe lower threshold diodes?

Yep, you could. However, diode thresholds start getting quantized as you go lower. Zeners down to about 2.? volts exist, as I remember. Then you get LED forward voltages, silicon carbide, silicon, silicon schottky, germanium, etc. There are some exotic IV-V semiconductor setups with other voltages, but you're not likely to find those. You can also get bandgap reference diodes like the TL431 for 1.25V and up. There's a world of possibilities, but IMHO you're pushing too hard on a non-precision setup to try to force some precision on it.

It's worth tinkering with if you like the approach, but I would go for a comparator with some hysteresis, a couple of diode junctions for a reference, and a pot to set threshold. The comparator could be a spare opamp if you are careful about it.

There isn't any right or wrong here. I just prefer other approaches for better precision, human factors, and predictability. That's one reason I mentioned the LM3916. Sure, it seems complicated to use a whole IC for it, but if you hook it up in the minimal configuration, it gives you a solid LED drive with comparator thresholds and hysteresis. You don't have to use all those LEDs, just the one threshold you want. As I remember, it winds up being the LM3916, four resistors and the LED, maybe some filter/bypass caps.

Yeah, some of the LM3916 circuits seemed pretty straight forward but it was new to me (read: scary).  I totally missed the point about it needing to be a clear on/off, yes/no, 1/0 etc.  I was just thinking a little blinking light would give me (or another user) a clue that they were in the right signal range.  I guess I need to decide how clear it needs to be and if it is worth the effort.  Right now I have an opamp clipper with an extra pot in the feedback loop to adjust the minimum gain.  I turn it until the signal just barely (audibly) clips.  Then I use the regular gain pot (in the same feedback loop) to adjust the amount of distortion.

The new design I'm working on was going to be a basic opamp clipper with a clean opamp boost in front to set the threshold/ratio/sensitivity (I'm not sure what to call it yet).  That's when I thought I'd be able to include an indicator.

[Bill Mountain]

It's a bit off topic but since I brought it up I thought I would explain why I'm even trying to redesign.  I was fairly happy with the original design (modded Distortion +) but with the sensitivity control as I designed it, the pedal would be fairly noisy (and some oscillations) at the highest gain range for my lowest output bass.  I'm hoping by having a separate boost for input gain that it would alleviate some of these issues (by not asking a single opamp to boost up to 600x).

[Bill Mountain]

Double Post.

[wavley]

Ahh, just replace one of your clipping diodes with an led, when it lights up... you're clipping

Actually, I did do this with my D+, (not as a clipping indicator of course, but I did mount it in a bezel so I could see it) and it sounds pretty good, plus it looks neat.

[Bill Mountain]

Ahh, just replace one of your clipping diodes with an led, when it lights up... you're clipping

Actually, I did do this with my D+, (not as a clipping indicator of course, but I did mount it in a bezel so I could see it) and it sounds pretty good, plus it looks neat.

I've thought about it but I don't need such a big output signal and it doesn't help me that far in the circuit.  I'm trying to normalize the input gain to get the best possible range out of the distortion control (if that makes any sense).

Thanks for the tip though!

[wavley]

Ahh, just replace one of your clipping diodes with an led, when it lights up... you're clipping

Actually, I did do this with my D+, (not as a clipping indicator of course, but I did mount it in a bezel so I could see it) and it sounds pretty good, plus it looks neat.

I've thought about it but I don't need such a big output signal and it doesn't help me that far in the circuit.  I'm trying to normalize the input gain to get the best possible range out of the distortion control (if that makes any sense).

Thanks for the tip though!

Oh, I was just joking anyway, your idea makes perfect sense, although the D+ does have enough output signal to light up an led pretty well.

[Bill Mountain]

Ahh, just replace one of your clipping diodes with an led, when it lights up... you're clipping

Actually, I did do this with my D+, (not as a clipping indicator of course, but I did mount it in a bezel so I could see it) and it sounds pretty good, plus it looks neat.

I've thought about it but I don't need such a big output signal and it doesn't help me that far in the circuit.  I'm trying to normalize the input gain to get the best possible range out of the distortion control (if that makes any sense).

Thanks for the tip though!

Oh, I was just joking anyway, your idea makes perfect sense, although the D+ does have enough output signal to light up an led pretty well.

I figured.  I just wanted a chanced to further explain my idea.