as we don't know exactly the max input voltage of F5 i decided to guess it:
From Firstwatt F5 manual:
Max uncliped out is +/- 20 V. Which means 40V peak/peak(*). To define Rms value then we need to multiply this value by (1/2,828)= 0,353.
It is equal to 14,12V.
So F5 is able to ouput max 14,12V Rms voltage.
We know that voltage gain the circuit apply to input signal is 15,3db.
From there: 14,12V rms= approx +25dbu. Minus 15db= circa +10dbu which equal 2,5V rms.
It's in the ballpark of previously given figure of 3V rms.
It's 2 db less than the 10db pad i assumed earlier. So as a to be on the 'safe side' i would implement a 12db pad rather than a 10db.
If implementing it , you'll adapt your desk output level to your amp input and as such no need to have an automatic fader ( which a limiter is) at play.
*: here we don't consider load impedance ( loudspeaker impedance) as it doesn't mater: we are only interested into max voltage of output stage (and amplifier voltage gain) before clipping.
From Firstwatt F5 manual:
Max uncliped out is +/- 20 V. Which means 40V peak/peak(*). To define Rms value then we need to multiply this value by (1/2,828)= 0,353.
It is equal to 14,12V.
So F5 is able to ouput max 14,12V Rms voltage.
We know that voltage gain the circuit apply to input signal is 15,3db.
From there: 14,12V rms= approx +25dbu. Minus 15db= circa +10dbu which equal 2,5V rms.
It's in the ballpark of previously given figure of 3V rms.
It's 2 db less than the 10db pad i assumed earlier. So as a to be on the 'safe side' i would implement a 12db pad rather than a 10db.
If implementing it , you'll adapt your desk output level to your amp input and as such no need to have an automatic fader ( which a limiter is) at play.
*: here we don't consider load impedance ( loudspeaker impedance) as it doesn't mater: we are only interested into max voltage of output stage (and amplifier voltage gain) before clipping.
Last edited:
Here's a neat and simple limiter circuit based on the one I just designed for my BEANS amplifier. This circuit connects across the input, (represented here by V1 and R1) and does not affect the signal at all until clipping starts. Assuming the input has a drive impedance of about 600 ohms, this will effectively clip it. It uses very little power. R5 is a 10k pot to set the limit level from about 1V up to 4.5V. D3 and D4 are LEDs which will show if positive or negative clipping is occurring.
The circuit works by setting a bias on the base of the transistors, if the input voltage exceeds this, then current from the input is shunted thru the LEDs to ground. 7805 and 7905 are standard voltage regulators to give +- 5V. This should be enough I think, but could easily be increased if necessary. You could power this from a small centre-tapped transformer, with a couple of rectifier diodes like 4001s and the regulators.
The circuit isn't actually in the signal path at all until clipping starts. But it may be a little soft for your tastes. As others have pointed out, an active limiter circuit with an op-amp or two would give cleaner clipping.
Here's the output with a 5V sinewave input:
The circuit works by setting a bias on the base of the transistors, if the input voltage exceeds this, then current from the input is shunted thru the LEDs to ground. 7805 and 7905 are standard voltage regulators to give +- 5V. This should be enough I think, but could easily be increased if necessary. You could power this from a small centre-tapped transformer, with a couple of rectifier diodes like 4001s and the regulators.
The circuit isn't actually in the signal path at all until clipping starts. But it may be a little soft for your tastes. As others have pointed out, an active limiter circuit with an op-amp or two would give cleaner clipping.
Here's the output with a 5V sinewave input:
Last edited:
I was referring to a peak monitor set to indicate a specific threshold: clipping. You can set it for anything. It's a single LED that detects when a peak has crossed a threshold. Clipping is the important one, really nothing else matters. All levels below clipping are fine.
You don't need to know that. There's no reason you would. You need to know if high levels above a certain point will cause a problem. That problem is clipping.
Correct, your hearing doesn't measure voltage, but also, the loudness you hear is based on many factors outside of your mixer. Hearing is 100% subjective. Measuring a voltage or power level is 100% objective. f
The key question to answer here is: Why do you need to know peak level?
@jaddie I actually want to ensure the voltage is optimal. It's possible I'm currently running less than 1V signal into my amp from the mixer. Clipping aside, I do want to see what voltage is coming out, just like a digital meter but with sample and hold, and more permanent like a peak monitor. I realize the PAD circuit could protect against overvoltage, but I would still need to ensure I'm feeding an optimal signal to the amp. Maybe this calls for both a VU meter AND a peak meter. Still learning about this🙂
VU Meter vs Peak Meter
VU Meter vs Peak Meter
The fellow has done quite a bit of work on a DIY Vu Meter:
https://s-o.webnode.cz/vu-metr/vu-metr-opa/
https://s-o.webnode.cz/vu-metr/vu-metr-opa/
You now have to define what optimal is to you. 🙂
Very probably. But we would need more info about your instalation.
Why double what you already have?
In case you didn't get it's your console vu meter purpose... and how you set up your input gain on your console which define what you call the 'optimal signal'.
Peak meter if of no interest if you know how to use a vu meter.
Iow your issue is relative to gain staging (aka gain structure) and how to manage it.
This will ask to learn about what reference level is and some bit of knowledge about dynamic range of signal in audio.
Both are relatively easy to get.
And this have to be supported.
And don't take me wrong, i don't have issue with limiters or compressors, i use them constantly but in that case they are not needed.
May i suggest you read about gain staging first. Once the concept is understood other things can be incorporated.
Hello @roboDNA
See the quick circuit I've created to act as a limiter.
This doens't clip the signal, but just reduces its volume in order to not go beyond a certain limit.
I've tested in a protoboard. See the results in the attached videos (sine wave and music).
Blue signal is input, yellow is output.
I built a VCA based on a LDR+LED. This set must be protected from ambient light.
It's a L-PAD based on a fixed resistor (100k) and a LDR to ground.
As the signal increases, the integrator feeds a LED.
The LED optically feeds back the LDR, attenuating the signal and limiting the output.
I put 2 LED's: one for illuminating the LDR and the other as an indicator for the limiter action.
This way, while the limiter doesn't kick in, there is not influence in the sound, since it passes only by two buffers and the passive attenuator (L-PAD).
By modifying resistors and capacitors, you can adjust the threshold through the GAIN, adjust the attack and the decay.
If you have access to buy integrated VCA's, you can simplify this circuit, but use the same concept.
See the quick circuit I've created to act as a limiter.
This doens't clip the signal, but just reduces its volume in order to not go beyond a certain limit.
I've tested in a protoboard. See the results in the attached videos (sine wave and music).
Blue signal is input, yellow is output.
I built a VCA based on a LDR+LED. This set must be protected from ambient light.
It's a L-PAD based on a fixed resistor (100k) and a LDR to ground.
As the signal increases, the integrator feeds a LED.
The LED optically feeds back the LDR, attenuating the signal and limiting the output.
I put 2 LED's: one for illuminating the LDR and the other as an indicator for the limiter action.
This way, while the limiter doesn't kick in, there is not influence in the sound, since it passes only by two buffers and the passive attenuator (L-PAD).
By modifying resistors and capacitors, you can adjust the threshold through the GAIN, adjust the attack and the decay.
If you have access to buy integrated VCA's, you can simplify this circuit, but use the same concept.
Something like this would work: https://dbxpro.com/en-US/products/166xs . Use the compressor section and set the ratio to infinity for limiting. You would need to set the threshold for the point where the limiting starts. It also has a "Peak Stop" one knob limiter that may be all that you need or you can use it in combination with the compressor section. You would need to keep the "stereo Couple" button in for stereo use. Use fast attack and use your ears for the release time set point. You can lower the threshold to get the feel of the controls and the sound of the limiting and then raise the threshold to where you need it for use. There are loads of expensive limiters that recording studios use to give more density to the sound. The DBX stuff is low end but may be just fine for what you need. There are better devices at a good price offered in the 500 series powered rack format.
If you're getting the volume you want without clipping, you have accomplished your goal.
Is your volume too low?
Why? Just curious? What action would you take if your output voltage "looked wrong"? Turn it up even though the system was already playing at an appropriate level?
The optimal signal to the amp is whatever results in the sound volume, from your speakers, that you desire. There are no specifications for this, it's a matter of personal preference. You can calculate the expected SPL based on distance to speakers, number of speakers, efficency of speakers, and gain of the power amp, assuming a well defined test signal, but you can't do this with music.
If it's loud enough, and your amp isn't clipping, you're done. Enjoy the music.
It requires no meter at all. If you have to turn down your master because your speakers are too loud, look at were the master ends up, note the attenuation in dB and pad the amps input the same amount. If the fader ends up at -15dB, you need a 15dB pad so you can run the fader at its normal position.
The title of the article is incorrect, a VU meter does NOT represent loudness of a signal. Loudness is a psycoacoustic property, a VU meter is a standardized, average-responding (not peak, not RMS) indicator without any psychoacoustic algorithms. It was developed in 1938 so telephone companies and radio networks could use the same indicating device.
It does show how a signal will look on another VU meter, and can show very approximate relative loudness within a defined system with clearly defined reference levels, like a broadcast network or recording device. It cannot represent loudness of speakers in a room. The article starts to define the differences between the two meters, but never addresses the actual purpose of a peak meter, the reason they're useful, nor the differences between peak, semi-peak, true peak, and PPM. A peak meter doesn't indicate loudness either. About the only thing right in that "article" is there will be an offset in reading music between a peak and VU meter, but their figures are wrong. It's 8-10dB.
That's a very old idea dating back 70 years or so. It might be though of as a limiter if you adjust the ratio to something very high but it's not a peak limiter. LED/LDRS are far too slow to limit true peaks. Not half bad as a pseudo-rms compressor though.
Nice unit, I use them often. But its an RMS compressor, even at infinite slow, and the "peak stop" function is a soft clipper, not a true peak limiter. From the manual: "PeakStop is a smooth well-controlled soft clipper..." If the goal is just to stop peaks and accept the distortion that results, it's fine. If the goal is to stop peaks and maintain low distortion, then no.
Yes but if something had to be implemented it would not be a compressor but a limiter.
Like Jadie said opto is usually to slow to act as a limiter.
Your comp circuit is classic, i'm sure as an instrument compressor it'll works well ( in fact i use something very similar as my 'main' instrument comp, with some twists like filtering of sidechain, autorelease,...).
Like Jadie said opto is usually to slow to act as a limiter.
Your comp circuit is classic, i'm sure as an instrument compressor it'll works well ( in fact i use something very similar as my 'main' instrument comp, with some twists like filtering of sidechain, autorelease,...).
Last edited:
This circuit is for limiting volume of music (original post question), so LDR is fast enough for this purpose.
TR=18ms
TF=120ms
18ms attack is way to slow for limiting.
In mine i use vtl5c9 and it's 6ms attack time is at the edge (limit too slow) for voice purpose. As a reference an la2a is in the 10ms range attack time.
A 1176 is able to attack from 20 to 800 nano second... but it's fet based.
I will look into my dsp but i think the limiters are sub 1ms attack ( those are for protection purpose only but i never used them...)
Whatever, it's not needed in there. The issue is gain staging/reference level related 😉
In mine i use vtl5c9 and it's 6ms attack time is at the edge (limit too slow) for voice purpose. As a reference an la2a is in the 10ms range attack time.
A 1176 is able to attack from 20 to 800 nano second... but it's fet based.
I will look into my dsp but i think the limiters are sub 1ms attack ( those are for protection purpose only but i never used them...)
Whatever, it's not needed in there. The issue is gain staging/reference level related 😉
Last edited:
I know you can have super faster components, but this is just a home made DIY circuit with components lying around, not a SpaceX project.
I posted for talking about concepts not to compete with anyone or to produce a professional high end product.
The post was answered for sometime so, I decided to answer it and I like to play with simple (and old) electronics.
I posted for talking about concepts not to compete with anyone or to produce a professional high end product.
The post was answered for sometime so, I decided to answer it and I like to play with simple (and old) electronics.