Overload/ Clipping Indicator

Main Index Projects Index

The ESP website does have a couple of circuits for high performance overload detectors, but one is buried within the Project 30 mixer pages and is easily missed, and the other is in the Project 152 bass amp project.  Project 146 is another, and it was considered for a PCB, but I decided against that because it needs a dual supply.  Since clipping detection is something that people seem to need (especially with mic preamps and the like), the circuit has been modified, physically tested, and is presented here.  The modification is primarily to allow the circuit to operate from a single supply.  The circuit described is an improved version of the one shown in Project 146.  The improvement comes about by using a small-signal MOSFET (2N7000 or equivalent) rather than a BJT for switching the LED.  This change makes the LED 'on' time more predictable and easier to control.

The circuit is shown below - it's very simple, but works well, even with the very basic LM358 opamps.  While it could have been made faster by limiting the opamp output swing with more diodes, but doing so would be rather pointless.  The circuit is designed to detect both positive and negative peaks - a great many peak/ overload detectors only work with one polarity.  This is not really a good idea, because many (most) audio signals are asymmetrical, and detecting only one polarity could mean that some signals could be clipping without you realising that it's happening.

Another requirement is that the circuit can be connected to high or low impedance circuits without creating a non-linear load that causes distortion.  This is especially true with high impedance circuits, because any non-linearity in the detector is directly reflected back to the source.  An overload indicator that creates distortion in the source circuit is hardly useful.

Although shown here using a single 12V supply, the circuit will work fine with other supply voltages.  The maximum supply voltage is limited by the maximum gate-source voltage of the MOSFET.  With the 2N7000, namely 20V.  The window detection thresholds are set from the supply rails, and the ratios remain the same regardless of supply voltage (V_window = VCC / 3).  Only the LED series limiting resistor will need to be changed in order to maintain a useable current at reduced voltage.  For example, with a +5V supply, you might reduce the LED series resistor to around 820Ω.  At that voltage, the detection window is only &plusmh;830mV, which is a bit too limiting.

If you want to use the P246 to turn off an existing 'power on' LED, it's very simple to do.  The two LED terminals are joined, and wired in parallel with the existing Power-On LED.  Note that it must be connected to ground with a limiting resistor from the positive supply.  If the existing LED is not wired this way it must be changed so it matches the arrangement described.  Failure to ensure that the external wiring is correct may result in instant destruction of the MOSFETs and possible PCB damage!

The advantage of using the circuit this was is that no additional holes are needed in the front panel.  The existing 'power on' LED is simply re-purposed, so it works normally except if the amp is driven to the onset of clipping.  If the power LED starts flickering, it means that you're reached the clipping threshold as set using the trimpots.  The compete details will be in the construction guide when the PCBs are available.

The circuit diagram is shown in Figure 1, and it's shown with an LM358 opamp.  While you can also use expensive high-performance (rail-to-rail) opamps, there is no reason to do so - the circuit only lights an LED.  The biggest advantage of the LM358 is that the output can swing to the negative supply rail, so there is no chance of the LED being on all the time.  For other opamps, it may be necessary to use two diodes in series for D1 and D2.  This isn't recommended.

Despite the simplicity, the circuit works very well.  If used with a mic preamp or similar, VR1 (trimpot) will allow you to set the peak voltage where the LED will come on.  With VR1 at maximum, the detection voltage is ±2V, so there is more than enough headroom before the signal clips.  Normally, I'd expect VR1 to be set to roughly 1/2 resistance, which provides a detection threshold of ±4V.  This is about the maximum you'd normally use for a circuit operating with ±15V supplies.  Setting VR1 to lower resistance increases the detection threshold voltage.  At a 10% resistance setting (10k for a 100k trimpot) the detection threshold is ±44V.  If desired, a fixed resistor can be used instead of the trimpot.

U1A and U1B form what's known as a 'window comparator'.  Provided the signal voltage remains within the boundary reference voltages at pins 2 and 5, the outputs remain at close to zero.  Should pins 3 and pin 6 (which are joined) go above or below the reference voltage (+4V and +8V), the output of the corresponding opamp will swing high (about +10V).  C1 charges immediately via the diode, and the LED is turned on by Q1, a small-signal MOSFET.  After the transient has gone away, it takes time for C1 to discharge, so the LED remains on for long enough for you to see it.  C1 cannot discharge back through the opamp outputs because of the diodes (1N4148 or equivalent).  The LED can be any colour you like, and the LED current is about 4.5mA.  This can be reduced by increasing the value of R7 and vice versa.  Note that the circuit is mono - if you need to monitor a stereo signal you'll need two of them (the PCB will be stereo).

The nominal signal level (with VR1 at maximum) is ±2V (1.414V RMS).  It can be increased by making R6 a higher value, but I suggest that you keep to the default.  VR1 provides plenty of adjustment, so the detector can be used with power amplifiers (including BTL, dual or single supply).  The adjustment of VR1 will become fiddly if the detector is used with a high-power amplifier unless R1 is selected appropriately (see below).

Figure 1 - Overload Indicator Schematic

There's not a lot to the circuit, and it is economical to build.  The input has an earth/ ground reference set by VR1.  A degree of ground isolation is achieved by using R_g.  This prevents supply noise for the supply used for the detector from being coupled into the audio ground.

Note:   The LM358 opamp should not be substituted.  It's used because its output can get to the negative supply voltage (within a few millivolts).  Some CMOS opamps can do the same, but few can handle the required supply voltage (12V) and most are only available in SMD packages.  These can still be used (with an adapter board), but there's no point.

There is one very important point that you must be aware of.  Because the opamp comparators are fairly fast and there is a LED being switched on and off, the circuit can introduce noise via the 12V supply lines (in particular the ground return).  The LED and switching connects to the earth/ ground bus of the 12V supply, and should be isolated from the signal ground to minimise ground noise.  For this reason, it is very important that all power wiring is returned directly to the power supply, and not daisy-chained from the supplies used for preamps.  The preferred option is to use a 12V SMPS or a separate zener regulated supply - it won't be perfect, but the circuit is only an indicator, and extreme accuracy isn't necessary.

If a large number of these circuits is used (in a multi-channel mixer for example), there's a lot to be said for including a secondary power supply to power all 'noisy' electronics.  These include overload detectors (like this one) and metering amplifiers.  If this is done, the power rail decoupling becomes less of an issue as long as all noisy supply busses are kept separated from other circuitry.

Note that the inputs have no protection from voltages outside the opamp supply rails.  Adding diodes will provide good protection, but they occupy a fair bit of PCB real-estate and weren't included.  Provided you start with VR1 set for minimum output and advance the trimpot until you see 'action' (the LED flashing) it will be perfectly safe to use with any circuitry.  Where the input level will normally be (perhaps significantly) higher than the supply rail, the input pot allows a wide adjustment range.

To use the detector as a power amp clipping detector is simple enough, but be aware that unlike Project 23, it cannot compensate for supply rails that collapse with sustained high power.  Therefore, it would normally be adjusted so that any signal greater than ~80% of the nominal supply voltage will cause the LED to come on.  This is pessimistic, and in normal use it will be ok for the LED to flash occasionally.  For an amplifier using ±35V supplies, you might want the detector to operate with any transient signal above 28V peak (about 50W into 8Ω).

The input attenuator (VR1) must include the 10k resistor from the wiper (R3), as that limits the maximum fault current.  The value used for VR1 depends on the levels you need to detect.  100k is convenient for preamps because it presents a high impedance that won't load most circuits down.  It will also be alright with power amps using up to ±40V supplies, but an input attenuator is preferred, using R1 to reduce the level to the trimpot.

For amplifiers with different supply rails (and therefore different power ratings), VR1 can be adjusted to suit.  Other than for 'line level', never assume that VR1 can withstand the full amplifier voltage!  A 100k trimpot can handle up to 50V and remain below its voltage and dissipation limits, but adding the series resistor (R1) and using a lower value pot is a better idea for power amplifiers.  Ideally, the pot should be more-or-less at the centre of its travel to allow accurate level setting.  Assuming use with a power amp and a 10k trimpot, the following will be useful ...

Vpeak	R1
25V	39k Ω
35V	56k Ω
50V	82k Ω
100V	120k Ω

The above gives you a rough idea, and there's plenty of adjustment available to cover intermediate voltages.  V_peak is the supply voltage for the amp.  Of course you can use a lower value pot, but the idea is to minimise dissipation.  Frequency response isn't an issue (there are no high-level, high-frequency transients in music).  That means that the high impedance won't create any problems for detection.

Overload detectors such as the one shown here can be a blessing or a curse.  If you often use your system turned up pretty loud, then you'll likely be horrified to see that the clipping indicator LED is on much of the time.  It's not at all uncommon for amplifiers to be clipping on transients, and most of the time the clipping is entirely inaudible.  An overload indicator makes it very easy to see that's what is happening, and you could easily discover that when operated below clipping at all times, the amp isn't loud enough.

Figure 2 - Typical Operation With Noise Input

Figure 2 shows the simulated output (LED current in red), the noise signal in green, along with the upper and lower threshold voltages (8V and 4V respectively).  Any time the input is greater than 8V or below 4V, the LED is turned on.  This example is deliberately set so that there is plenty of activity.  Although the diagram is a simulation, the waveforms are no different on an oscilloscope.

If you look very carefully, you will see that there are some excursions just on the thresholds that don't cause the LED to light.  This is normal - the signal voltage needs to be at least a few millivolts greater than the threshold.  While we might assume that 'fast' musical transients have a large high frequency component, this is usually not the case at all.  The most common cause of amp overload is bass and midrange, especially when there is additional transient information 'riding' the bass or midrange waveform.  The energy in music rolls off naturally above ~1.5-2kHz, and a super-fast detector serves no useful purpose.

All circuitry shown in this project is operated with an unbalanced input.  Since it's intended to be used within a preamp or mixer that's not a problem.  It can also be used as an external unit, and will work fine even with balanced circuits.  Because the input impedance is very high (when VR1 is 100k), the circuit can monitor one of the two signal lines of a balanced interconnect, and because both usually have exactly the same voltage (just the polarity is reversed) if one line is close to clipping, then so is the other.  To maintain acceptable balance, the un-monitored signal line should have a 100k resistor to ground.  Alternatively, you can use the Project 176 differential amplifier to convert from balanced to unbalanced.

Please Note:  If the circuit is used with a single-supply BTL amplifier, there will be a DC voltage present at the output.  To minimise needless dissipation in the attenuator, you must use a capacitor from the amp's output to the 'Input' terminal.  1μF will generally be enough, even at the lowest voltage range (25V, indicating a single 50V supply).  The cap's positive terminal must go to the amp's output (assuming a normal +ve supply).

A PCB for this project will be available based on demand.  It will have two channels, and each can be adjusted independently for use with preamps, power amps and other high voltage sources.  The inputs are AC coupled, and VR101/202 can be replaced with a fixed resistor if that works for you.  Normally the pots will be used so the detection thresholds can be adjusted to suit the application.  If used with a power amplifier, protection diodes are employed to ensure that the input stage isn't damaged.  For high-power systems, R101/201 should be increased in value.

A single channel is shown below, and R1, R2, R3, C1 and R_g are common to both channels.  The second channel uses R201, C201, etc.  While R1-3 are shown as 2.2k, the value can be changed to suit what you have available.  Don't go above 10k though, as the reference voltages may become unstable.  R105/205 sets the LED current, so use the same value for R1, 2 & 3.

Figure 3 - Dual Clipping Detector (PCB Version)

With the values shown, the detection thresholds are ±2V (0k input resistor, 100k or 10k trimpot to ground), but this is easily changed.  Mostly, it's not necessary, because ±2V is a sensible limit for 'low-level' circuitry.  For power amplifiers, VR101/201 should be 10k, and R101/201 are as determined from the table shown above.  This allows peak voltages up to ±70V (over 300W into 8Ω) to be handled with ease, using 0.25W resistors.  Higher voltages are accommodated by increasing R101/201 further.  The trimpot is essential of course, otherwise you can't set the voltage.  You can calculate the respective values of R101/201 and fixed resistors in place of VR101/201, but inconvenient values are more likely than not.

The 100k/ 10k trimpots let you set the trip voltage to anything you like.  With 10k input resistors and 10k trimpots, the minimum voltage that can be detected is ±8V (4W, 8Ω), so it's suitable for amps down to 15W.  BTL amplifiers can also be monitored, but if they operate from a single supply you'll need to reverse the polarity of the input capacitor to reject the DC component.  The circuit has been designed for maximum possible flexibility, and it's a small PCB (expected to be around 60 × 40mm).  To keep DC out of the input pots, you can add external caps for use with single-supply BTL amps.  Mostly, they are not necessary, and they are only required with single-supply BTL amps.

The reference voltages will almost always be symmetrical (R1, R2, R3 equal values).  You can have asymmetrical thresholds, but it's neither necessary nor useful.  Avoid the temptation to increase the value of R2, as that may cause slight instability in the reference voltages.  If you wish to use different opamps, ensure that their output voltage can get to the negative rail/ ground.  Many opamps cannot reduce their output voltage to the negative supply, so the switching MOSFETs may never turn off.

The LED 'on' time can be extended by increasing C102/202.  I wouldn't recommend more than 22nF as that will extend the 'on' time way too far.  As shown, a short transient will turn on the LED for around 2ms - plenty of time to see it, and a very good indication of momentary transients.  Note that you can use a single LED for both channels - simply join the drain terminals of the two MOSFETs, and use a single resistor and LED.  If either channel clips, the LED will come on.

For all of the circuits shown, be careful with supply and ground wiring.  The circuit is designed to minimise ground current so it can't inject 'hostile' waveforms into the common ground bus (a few microamps at most), but it's not so easy to keep spike currents out of the supply lines.  The current waveform is quite capable of causing audible noise, so ensure that the clipping detector has its own set of supply leads wired back to the output of the regulator board.  Keep these leads well away from signal wiring, and consider using shielded dual-core cable to keep radiated noise to a minimum.  Even though the selected opamps aren't fast, the transistor is fast, and will switch on in less than 100µs when the input threshold is exceeded.

To help minimise noise, use a high-brightness ('ultra-bright') LED (blue LEDs are generally the brightest, but are intrusive).  Aim for a brightness of 100mcd (milli-candela) or more, and increase the value of R105/205 to get a comfortable brightness.  With 2.2k as shown, LED current is about 4.5mA, and that's more than enough with a high-efficiency LED.  You can also use a separate supply to ensure that the audio and indicator supplies can't interact.

One nice thing is that the component values are pretty flexible.  R101/201 have to be correct for the incoming signal level and the trimpot value, but most of the others can be changed without affecting the performance.  R102/202/ 103/203 are shown as 220k, but you can use anything from 100k to 330k with little change in performance.  R1, 2 and 3 are shown as 2.2k, but you can use anything from 1k to 10k - they just have to be the same value.  I used 2k2 because that's what I included for the LED, reducing the number of values needed.

This is not the simplest circuit you'll find, but it is about as simple as you can get while maintaining good accuracy and speed.  It's very flexible, and will work happily with everything from mic preamps to power amps.  Using a single supply makes it easier to power, as you don't have to worry about using a dual supply.  It will also work happily with single-supply BTL amplifiers (with an added external capacitor, ~10μF, +ve to amplifier).  All parts are easy to get, and the number of different values has been minimised to reduce the chances of errors during construction.

I have the prototype (mono) inside my bench amplifier, and I know instantly if it's clipping.  Prior to this, I had to use a scope to verify that the signal was clean.  Clipping is easy to hear with a sinewave, but not with programme material (especially that from FM radio).  I powered it directly from the amp's positive supply, so it's running from +23V.  The outputs of LM358 opamps can't reach the positive supply (about 1.5V is 'lost') and coupled with the diode drops, the gate of the MOSFET is just within the maximum specified.  Being a bench amp it's easy for me to to fix it if necessary.  However, for normal use I suggest that you do as I say rather than do what I do. 

Main Index Projects Index