Hello,
I'm looking into possibly using a compressor/limiter between my souncraft signature12 mixer an my F5m amp. I want to primarily use it as a "Brickwall' limiter to ensure the amp never sees more than 3V.
I want to be able to set a threshold voltage, in my case 3V as opposed to setting a ratio like 10:1 or 20:1 since that is less accurate. I'm also looking a solution which has short attack and release times. I rather not design my own hardware for this, but would be willing to build a DIY.
I've looked at Rane, DBX, Empirical Labs Distressor, Waves C4 and a few others but they are packed full of features I don't need or want. I just want the Brickwall limiter and not color the signal.
Does anyone use such a device or have suggestions where I can find a simple device with not so many additional features? ( I plan on getting a 19" rack to mount my amps and other related gear )
I'm looking into possibly using a compressor/limiter between my souncraft signature12 mixer an my F5m amp. I want to primarily use it as a "Brickwall' limiter to ensure the amp never sees more than 3V.
I want to be able to set a threshold voltage, in my case 3V as opposed to setting a ratio like 10:1 or 20:1 since that is less accurate. I'm also looking a solution which has short attack and release times. I rather not design my own hardware for this, but would be willing to build a DIY.
I've looked at Rane, DBX, Empirical Labs Distressor, Waves C4 and a few others but they are packed full of features I don't need or want. I just want the Brickwall limiter and not color the signal.
Does anyone use such a device or have suggestions where I can find a simple device with not so many additional features? ( I plan on getting a 19" rack to mount my amps and other related gear )
Hello!
Do you play with electronic circuits?
Are you willing to build a circuit or looking for an off the shelf device?
If want to build a circuit, we can develop it together.
It will use components such as Quad Opamp (TL074), resistors, capacitors, LDR, LED and small signal BJT (BC547) and a simple +/-15V power supply (100mA).
Do you play with electronic circuits?
Are you willing to build a circuit or looking for an off the shelf device?
If want to build a circuit, we can develop it together.
It will use components such as Quad Opamp (TL074), resistors, capacitors, LDR, LED and small signal BJT (BC547) and a simple +/-15V power supply (100mA).
I'm still researching the best way to solve this. I'm in the process of testing the output voltage of my mixer to get an idea of min/max peak V ranges for 20kHz and 30Hz sine waves. I may end up just ensuring the voltage never goes over 3V by using the vue meter built into the mixer and master fader. I suspect the mixer will be accurate enough to allow me to manually keep the voltage within safe levels. If this case, the limiter/compressor would never 'kick-in' and may end up leaving it out considering I want to preserve the analog signal ( synths ) by keeping few devices in the path.
@ron68 yes, I do dabble in electronics so I would be interested in a DIY project if nothing is already available. I have made a short list of clip detection and peak detection circuits I can use, with a couple of them 'rounding' the signal instead of cutting it off square when it goes over set voltage max. The problem is the signal is slightly distorted at higher levels since the circuit needs to start 'rounding' the signal before it hits the brickwall setpoint.
@ron68 yes, I do dabble in electronics so I would be interested in a DIY project if nothing is already available. I have made a short list of clip detection and peak detection circuits I can use, with a couple of them 'rounding' the signal instead of cutting it off square when it goes over set voltage max. The problem is the signal is slightly distorted at higher levels since the circuit needs to start 'rounding' the signal before it hits the brickwall setpoint.
Ok! Let me think about a circuit for this purpose,
I built a simple compressor sometime ago for listening podcast at night and works very well.
I'll redesign it to be a limiter. It's something very simple and will not distort.
I like to play with these things 🙂
I built a simple compressor sometime ago for listening podcast at night and works very well.
I'll redesign it to be a limiter. It's something very simple and will not distort.
I like to play with these things 🙂
^^ Y-E-S.. YES!!
I had posted another related questions and the Klever Klipper circuit may be interesting to look at but I have not purchase the book yet.
I had posted another related questions and the Klever Klipper circuit may be interesting to look at but I have not purchase the book yet.
Would I be correct in assuming that the actual purpose of this limiter is to prevent physical damage to speakers? It's placed just before the power amp...so....
You should also define what causes speaker damage, and there are several things actually.
You'd need to also define what is meant by a "brick wall" limiter. A limiter and a clipper are two different things, both modifiy the input waveform, and any signal modification could loosely be defined as "color", though it is a question of degree. Both will color the signal. And both can provide a fixed threshold above which no signal would exceed.
A peak limiter is a gain modification device that adjusts its gain based on the peak value of the signal. The speed at which that gain adjustment occurs is defined as the attack time, and the speead at which the gain is returned to normal is the release time. Faster attacks result in fewer excursions above threshold, and slower releases result in lower distortion.
A clipper modifies the signal differently. It stops peaks from exceeding a threshold by clamping it to a fixed voltage level. There is no attack or release time, but clippers always generate distortion, the degree depending on how much signal is above the clipping threshold and for how long. A brief peak can be severely clipped inaudibly, and a longer peak can be barely clipped without audibility.
Speaker damage is fundamentally based on RMS power value, or heating value. When there's too much heating, damage occurs to the voicecoil. Woofers can take far more power than tweeters, often 10X or more, and that's ok because the spectral distribution of audio puts most of the energy in the woofer's range. But heating also takes time, so damage can occur at a lower RMS level if sustained, and at a much higher RMS level if the duration is very short.
Now here's the problem. A clipper, or very fast attack/release peak limiter, does NOT limit the RMS level. In fact, a exceeding a clipping threshold barely changes the resulting RMS level, but does change the peak level. The ratio between RMS and peak value is no longer constant, reduces, and the difference is translated into a wider spectral energy distribution, which can actually make damage more likely. To limit, and hold the RMS to peak ratio constant, requires both a slower release peak limiter, and an RMS-based limiter, but not a clipper. And the amount they have to work varies continually with spectral content. Otherwise, they'll do far more audible "damage" to the signal without actually fully preventing damage to the speakers.
Active speaker protection is complex. Has to be. I would seriously look at commercial products with the specific feature on the used market rather than trying to reinvent the wheel.
^^ @jaddie thanks for the great info. I realize this is a rabbit hole and that is why I had decided to just manually 'limit' the output using the vue meter and fadder on my mixer after I understand the output voltage better using an oscope. I'll just try to keep the output under 3V using the db vue meter once I see how 'accurate' it is relative to the voltage output.
What commercial product would you suggest if any? I looked around a bit but only found expensive limiter/compressor units which had more features than I required. It would be good to have a basic device but that is easier said than done when considering all of the pitfalls your mention, as well as many designs.
I though of just having an accurate peak monitor/detector with sample-hold between the pre-amp and amp so I can visualize when the voltage is nearing 2.7V, well bellow the 3V max. I think it would need a uC which would use FFT to caclulate RMS voltage from audio signal. ( not a clean sine ) It would not do clip detection or protection.
What commercial product would you suggest if any? I looked around a bit but only found expensive limiter/compressor units which had more features than I required. It would be good to have a basic device but that is easier said than done when considering all of the pitfalls your mention, as well as many designs.
I though of just having an accurate peak monitor/detector with sample-hold between the pre-amp and amp so I can visualize when the voltage is nearing 2.7V, well bellow the 3V max. I think it would need a uC which would use FFT to caclulate RMS voltage from audio signal. ( not a clean sine ) It would not do clip detection or protection.
A VU meter is a "volume indicator". It's not an RMS meter, it's not a peak meter. The purpose of a VU meter is to show how audio will look on another VU meter. It's a standard meter, but not suited to speaker protection at all.
An oscilloscope is a graphic display of a waveform - voltage vs time. Unless the 'scope has a "hold" function it's pretty difficult to ascertain much useful level information in actual system operation. It's a very useful test tool.
Why are you choosing 3V/ 2.7V?
Why go through all that effort to calculate RMS? It's not something you reallly need to know, you just need to set things up right.
Remember, the maximum RMS level a speaker driver can handle without damage is represented by a curve, it's not a single voltage/power.
Take a look at the Behringer DCX2496LE, for a start. Yes, it's complex and more than you think you need, but they did actually put speaker protection limiting in there, which you could just use alone. New, it's $270 US. Used, less. dbx also makes a speaker management product.
I would first ask why you think you need any of this, though. And specifically what speakers, power amp, and application. Many, even most, applications run far less power to speakers than people think, and protection is unnecessary. Can't say in your case becasue I don't know enough about it.
An oscilloscope is a graphic display of a waveform - voltage vs time. Unless the 'scope has a "hold" function it's pretty difficult to ascertain much useful level information in actual system operation. It's a very useful test tool.
Why are you choosing 3V/ 2.7V?
Why go through all that effort to calculate RMS? It's not something you reallly need to know, you just need to set things up right.
Remember, the maximum RMS level a speaker driver can handle without damage is represented by a curve, it's not a single voltage/power.
Take a look at the Behringer DCX2496LE, for a start. Yes, it's complex and more than you think you need, but they did actually put speaker protection limiting in there, which you could just use alone. New, it's $270 US. Used, less. dbx also makes a speaker management product.
I would first ask why you think you need any of this, though. And specifically what speakers, power amp, and application. Many, even most, applications run far less power to speakers than people think, and protection is unnecessary. Can't say in your case becasue I don't know enough about it.
Keep in mind I don't know much about audio amps/pre-amps and have been learning about it recently.
I built the F5m amp and have new wharfdale super linton speakers.
The F5m handles up to 3Vrms input.
For the pre-amp: I have a Soundraft Signature12 analog mixer which outputs a maximum of just over 7.75Vrms ( aux and group outputs of the mixer ). I'm still learning about how the mixer's faders and gain knobs will affect the voltage level, but I'm fairly certain 7.75Vrm is the max it will output with faders at max and gain knobs at full ) To keep it simple, I'm accounting for 7.75Vrms as the potential voltage my amp could see.
I connect analog synths into the mixer, as well as a USB from my mac so I can play youtube/spotify through the mixer. I'm getting a turntable which would also connect to the mixer.
I want to try to preserve the analog signal and not have any digital in the signal path at all. ( analog synths )
I realize this is not a typical 'hifi' setup.
I want to ensure I'm feeding an optimal voltage level to my F5m amp, and realize the voltage levels will change depending on audio types. I need to visualize the voltage using some type of permanent solution instead of a meter or oscope. Currently, I think I'm only sending 0.5Vrms out of my mixer with the gain all the way down and fader at 0db but not sure. It all sounds great but I don't know if I'm running my whole setup at 20% capacity 🙂 I want to get the right voltage into the amp before trying to build a more powerfull amp. I"m sure my F5m can get as loud as I need once I have a max 2.7Vrms voltage feeding into it.
I think if I keep the vue meter on the mixer under +6db ( +10dBu ) it will be under 3.16V. That might be the best solution since I can avoid adding more circuits in the signal path. I still need to connect an oScope to get the output voltage with a 20kHz and 30Hz sine.
I built the F5m amp and have new wharfdale super linton speakers.
The F5m handles up to 3Vrms input.
For the pre-amp: I have a Soundraft Signature12 analog mixer which outputs a maximum of just over 7.75Vrms ( aux and group outputs of the mixer ). I'm still learning about how the mixer's faders and gain knobs will affect the voltage level, but I'm fairly certain 7.75Vrm is the max it will output with faders at max and gain knobs at full ) To keep it simple, I'm accounting for 7.75Vrms as the potential voltage my amp could see.
I connect analog synths into the mixer, as well as a USB from my mac so I can play youtube/spotify through the mixer. I'm getting a turntable which would also connect to the mixer.
I want to try to preserve the analog signal and not have any digital in the signal path at all. ( analog synths )
I realize this is not a typical 'hifi' setup.
I want to ensure I'm feeding an optimal voltage level to my F5m amp, and realize the voltage levels will change depending on audio types. I need to visualize the voltage using some type of permanent solution instead of a meter or oscope. Currently, I think I'm only sending 0.5Vrms out of my mixer with the gain all the way down and fader at 0db but not sure. It all sounds great but I don't know if I'm running my whole setup at 20% capacity 🙂 I want to get the right voltage into the amp before trying to build a more powerfull amp. I"m sure my F5m can get as loud as I need once I have a max 2.7Vrms voltage feeding into it.
I think if I keep the vue meter on the mixer under +6db ( +10dBu ) it will be under 3.16V. That might be the best solution since I can avoid adding more circuits in the signal path. I still need to connect an oScope to get the output voltage with a 20kHz and 30Hz sine.
Last edited:
You've focussed on the wrong thing here. The F5m looks like it's a 10 watt amplifier. Your speakers will handle up to 200 watts. So this isn't a matter of speaker protection. In fact, I'm not sure what you're trying to do...keep the amp from distorting?
None of this is necessary. If you're running into distortion issues, you need more power, not a limiter.
I'd suggest if you really want to build something, build up a peak flasher that lights an LED (and holds it on a bit) when 3V is crossed, positive or negative. That'll let you know when the amp is about to distort. And let you know when you need a more powerful amp.
@jaddie it's 25watt per channel and my speakers sound loud enough. I'm not trying to make it louder, but just trying to ensure the input voltage is optimal, without exceeding 3V. I do agree a peak monitor would be good. It's also a rabbit hole and I put that on hold until I understand my mixer output better. The mixer is a peak monitor but with no sample/hold so not perfect. You bring up a good point that my speakers are 200watts so not as likely to get damaged with my F5m as long as I keep the mixer output below 3V using its VU meter.
@ejp thanks, I'm starting to realize that may be the best solution. I just need to check the output on my oscope and that should confirm where the master fader should be.
The main problem I'm facing is the mixer only has a VU meter for master out, and not Group A and B outputs. Master has a higher output voltage so I was trying to avoid using it and use the Group A an B outputs instead. An external vu meter with sample/hold would be good to have...
@ejp thanks, I'm starting to realize that may be the best solution. I just need to check the output on my oscope and that should confirm where the master fader should be.
The main problem I'm facing is the mixer only has a VU meter for master out, and not Group A and B outputs. Master has a higher output voltage so I was trying to avoid using it and use the Group A an B outputs instead. An external vu meter with sample/hold would be good to have...
Actual level is signal level + gain setting (fader position plus internal gain). You can’t really see easily get an idea of gain, but you’ll have no idea at all about signal level without some sort of signal indicator. The type of indicator will determine the type of information you get.
This is not about level matching because there's a volume control before the power amp. The information needed is what level is too much, and that will change all the time.
I couldn't find specs, what happens at 3V (I assume distortion)?
Assuming that's the input voltage at which the amp clips, but couldn't verify. Again, why the concern?
All peak monitors or indicators have a hold function or the peak indiciation would just function as a variable brightness light. That's not what they typically do, so all have some form of pulse stretcher so that once triggered the light/lights stay on long enough to be visible as a solid light. True of bargraph meters too, though done in a different way.
Again...not a VU meter. Too slow. Realize that a real VU meter takes 300ms to move from the left resting pin to "O". And another 300ms to fall back down.
Again..no. That won't tell you what's going on unless you put the scope at the amp output (careful!) and find where it clips, then back into the input voltage.
Fader position is meaningless because it's only an indication of static system gain, not level. Audio level changes with signal, the fader doesn't change with signal. And all you have to do is apply a different signal (music, recording, playback device) and the fader position will be wrong.
Why don't you just pad the master output?
Are you sure the mixer is a VU meter? What's the mixer?
@jaddie you bring up very good points, thanks. The mixer is a Soundcraft Signature 12.
I now see what you mean about the VU Meter not responding fast enough.
I have a few peak detector circuits which use a capacitor to hold the value and a reset switch to drain it. I started a Kicad project a few days ago to use an STM32's ADC ( built-in op-amps ) and a mosfet to drain the cap using an I/O pin instead of a manual switch. It's more complex than I thought so put it on hold. I know there are a couple projects but with no microcontroller. I would need to investigate further to determine which circuit is best.
Since I know see my soundcraft has limitations and its vu meter won't show me what I need 🙂
I do after all need a peak monitor...
I now see what you mean about the VU Meter not responding fast enough.
I have a few peak detector circuits which use a capacitor to hold the value and a reset switch to drain it. I started a Kicad project a few days ago to use an STM32's ADC ( built-in op-amps ) and a mosfet to drain the cap using an I/O pin instead of a manual switch. It's more complex than I thought so put it on hold. I know there are a couple projects but with no microcontroller. I would need to investigate further to determine which circuit is best.
Since I know see my soundcraft has limitations and its vu meter won't show me what I need 🙂
I do after all need a peak monitor...
I'll just add that in my many decades of audio I've never had a power amp with a visible clipping indicator in my system, and never felt any need. In my pro audio work, yes, there are several amps I've used with both clip indicators and built-in peak limiting. Great for setup, then nobody ever looks again.
Good to know. Currently though, I have no peak monitor at all since the vue meter on my mixer does not show the actual Group 1/2 output levels, only the master output.
You mention you never use a clip indicator or peak limiting but I was referring to a 'peak monitor' with sample hold as you suggested. ( nothing to do with clipping or limiting )
Since this is my first go at all of this, I have no way of knowing what to expect so it very well may be that once I get going, I will not need to worry about over-voltage coming from the mixer, but at this time, I still have no way to know if the voltage is .5V or 2.5V. The faders and vue meter give no indication as I mentioned. I don't want to rely on my hearing to determine Voltage so still thinking I need a peak monitor.
You mention you never use a clip indicator or peak limiting but I was referring to a 'peak monitor' with sample hold as you suggested. ( nothing to do with clipping or limiting )
Since this is my first go at all of this, I have no way of knowing what to expect so it very well may be that once I get going, I will not need to worry about over-voltage coming from the mixer, but at this time, I still have no way to know if the voltage is .5V or 2.5V. The faders and vue meter give no indication as I mentioned. I don't want to rely on my hearing to determine Voltage so still thinking I need a peak monitor.
Imho you overcomplicate things:
Your desk as a max output level of +21,5dbu. This is equal to 8,7 V rms.
Your F5 have 3v input. Presented as is it doesn't gives enough info: is it peak, rms, peak/peak?
Let's assume it is rms. 3v rms is equal to +12dbu.
So you have circa 10db difference, your desk having potential to be 10db too loud.
Insert a 10db pad in between your desk out and your amp and it's done.
The pad: you need to know output Z of your desk. It have 150 ohm Z when balanced, 75 ohm Z unbalanced.
We then need to know the imput impedance of the F5, iirc it's in the 100k range.
From there go to here and follow recipe for an Lpad ( unbalanced drive for line input, read the full page it's on the last paragraph):
http://www.uneeda-audio.com/pads/
If you need to play with dbu and voltage:
https://www.sowter.co.uk/decibels.php
And all you need is a DMM to check things voltage related ( +4dbu= 1,23V). Check for AC voltage. There might be issue as often cheap DMM are optimised to read main so they can filter things above few hundred hz, so rather than using 1khz tone it might make sense to use 50 or 60hz one.
Your desk as a max output level of +21,5dbu. This is equal to 8,7 V rms.
Your F5 have 3v input. Presented as is it doesn't gives enough info: is it peak, rms, peak/peak?
Let's assume it is rms. 3v rms is equal to +12dbu.
So you have circa 10db difference, your desk having potential to be 10db too loud.
Insert a 10db pad in between your desk out and your amp and it's done.
The pad: you need to know output Z of your desk. It have 150 ohm Z when balanced, 75 ohm Z unbalanced.
We then need to know the imput impedance of the F5, iirc it's in the 100k range.
From there go to here and follow recipe for an Lpad ( unbalanced drive for line input, read the full page it's on the last paragraph):
http://www.uneeda-audio.com/pads/
If you need to play with dbu and voltage:
https://www.sowter.co.uk/decibels.php
And all you need is a DMM to check things voltage related ( +4dbu= 1,23V). Check for AC voltage. There might be issue as often cheap DMM are optimised to read main so they can filter things above few hundred hz, so rather than using 1khz tone it might make sense to use 50 or 60hz one.
Last edited:
^^ thanks, I did not know about pad resistor networks. I"m not sure how that would help in my case since different sources will produce different max voltages, each channel on the mixer having its own gain. I'm not worried about accidentally moving the master fader to high, I"m looking to see what the voltage is between the mixer and the amp. I realize my original question is about brickwall limiter but the info. posted here has helped me understand things more clearly. If there was a way to see voltage levels with a peak monitor and sample/hold, I could keep it at an optimal level and deal with clip detection/protection/brickwall limiter as a separate problem. ( I don't think I'll need a clip detector )
So if I connect a roland TS-8S to 1 channel on the mixer, and a turntable on another channel, i would need to adjust the gains for each channel accordingly, and then feed those into Group 1 and Group 2 of the mixer. There is not visual meter or way to know what voltage is coming out of the mixer's Group1/2 at that point. Only by ear.
So if I connect a roland TS-8S to 1 channel on the mixer, and a turntable on another channel, i would need to adjust the gains for each channel accordingly, and then feed those into Group 1 and Group 2 of the mixer. There is not visual meter or way to know what voltage is coming out of the mixer's Group1/2 at that point. Only by ear.
Your not used to read specification or use that kind of gear on a regular basis isn't it? 😉
Read specifications of your mixer: +21dbu is the max possible output your desk is able.
As it's the limit, once the attenuation of your pad is sorted (-10db in preceding example) whatever happen before the output stage of your desk, it'll never saturate the input of your next element in the chain ( your amp).
When you operate your mixer you are suposed to set the preamp gain ( mic or line, in your case line as you use synth) to not saturate the chanel electronic.
It's done selecting solo pfl ( pre fader listening) and adjusting gain for it to stay lower than the red (last yellow ) leds on your vumeter for the highest level your source ( synth) will generate. Once done you operate on parameter area where the desk will not saturate regarding levels into the chanel's electronic circuit and with the best achievable S/N ratio ( signal/noise ratio).
If you use eq to boost signal range ( especially low and mid) then it is worth cheking with solo pfl you did not saturate the chanel electronics. In case it does, lower input line gain to be under the led red and you are done ( eg if you boost +6db @ 100Hz, then lower your input gain by same amount or a bit more) .
That said even if you saturate your desk electronics the implemented pad on ouput makes it impossible to drive into distortion your amp: saturation come from elsewhere before, so you have to check where it happen along the line.
This operation is called 'gain staging'.
Peak meters are not really useful with analog gear. It can be in front of ADC, but it's technical information you usually set and forget ( as you set your levels for not to clip next stage then don't touch it anymore no need for constant information feedback).
Read specifications of your mixer: +21dbu is the max possible output your desk is able.
As it's the limit, once the attenuation of your pad is sorted (-10db in preceding example) whatever happen before the output stage of your desk, it'll never saturate the input of your next element in the chain ( your amp).
When you operate your mixer you are suposed to set the preamp gain ( mic or line, in your case line as you use synth) to not saturate the chanel electronic.
It's done selecting solo pfl ( pre fader listening) and adjusting gain for it to stay lower than the red (last yellow ) leds on your vumeter for the highest level your source ( synth) will generate. Once done you operate on parameter area where the desk will not saturate regarding levels into the chanel's electronic circuit and with the best achievable S/N ratio ( signal/noise ratio).
If you use eq to boost signal range ( especially low and mid) then it is worth cheking with solo pfl you did not saturate the chanel electronics. In case it does, lower input line gain to be under the led red and you are done ( eg if you boost +6db @ 100Hz, then lower your input gain by same amount or a bit more) .
That said even if you saturate your desk electronics the implemented pad on ouput makes it impossible to drive into distortion your amp: saturation come from elsewhere before, so you have to check where it happen along the line.
This operation is called 'gain staging'.
Peak meters are not really useful with analog gear. It can be in front of ADC, but it's technical information you usually set and forget ( as you set your levels for not to clip next stage then don't touch it anymore no need for constant information feedback).
Last edited: