Powered off audio devices can cause distortion

Did you know than an audio equipment can cause “off gear distortion” on line level signals that are driving multiple pieces of audio gear? I have had situation where I had one CD player wired to inputs of two different amplifiers (Y-splitter cable). When both amplifiers were powered on, everything worked nicely. But when one of the amplifiers were powered off, the sound played through the other still powered amplifier started to get distortion added to it. At low volume from CD was not distorted, but when sound volume from CD got higher the distortion started to get very noticeable. The distortion disappeared then the second amplifier was powered on again.

What is experienced is called “off gear distortion”. That distortion is caused because the audio inputs that work nicely looking like resistive loads when equipment often become non-linear loads (load resistance changes depending on input voltage) when equipment is powered off. That non-linear load causes the signal on the cable become distorted, and the other amplifier gets that distorted sound. The amount of distortion can vary depending on how the powered “powered off” equipment is built, the audio signal level on the cable and the output impedance of the device driving the cable.

For consumer audio circuits, the audio signals (line and mic) can have typically a maximum value of around 5.5 Vp-p, and has a maximum frequency less than 30kHz. Some professional audio applications can use in some cases even higher signal levels on the audio interconnections.

The biggest electrical threat on audio interfaces is normally ESD. The IEC 61000-4-2 will be the most appropriate standard that applies for this port. It supplies a test method for verifying that the end product is not susceptible to ESD events. Audio interfaces on on commercial products have typically protection components that protect electronics against damages when ESD event happen. Depending on equipment, there can be protection built into audio ICs and external protection circuitry. The standard IC’s ESD protection includes reverse-biased ESD diodes between each I/O pin and supply pins.

Elliott Sound Products AN-015 shows some ‘traditional’ schemes for audio input protection. The idea in them is that then a high input voltage (of either polarity) be connected to the input, the appropriate diode conducts and the input is protected.

Well, not always. The standard arrangement can provided a false sense of security, and can lead to catastrophic failures in some cases. If an input limiting resistor is included the circuit will work properly, but only if the value is high enough. This kind of protection circuitry, when built properly, can protect amplifier audio input against ESD and too high input voltages without causing noticeable distortion and look like resistive load when equipment is powered on. But when the equipment is powered off, they show as a non-linear load, because the protection diodes will start to conduct when the signal level goes above the +- diode voltage drop (typically around 0.5-0.7V). Input may be using (normally reverse-biased) protection diodes going to the analog supplies. When powered down, those obviously are at 0 V, effectively turning the whole affair into a limiter (about 0.6 V peak assuming plain silicon diodes, specifics depending on source impedance). The distortion similar to diode clipping used on some electric guitar effects. Elliott Sound Products AN-015 shows also other input protection methods that can perform better both powered on and off.

There is no easy solution to the non linear impedance of ESD diodes connected to power rails when the rails are off. ICs have clamping diodes to protect them from overvoltage input. But with the power off, the input signal is now trying to power the electronics that does not get power from power supply. The load for signals under +/-600mV would be high but outside this range the impedance would be lower and equal to the value of the resistor. Unless the output impedance of the audio signal source is very low then this non linear load would distort the audio signal. Increasing the series resistor in audio input value would slowly reduce the effect but would of course add noise.

One recommendation as that you should avoid hardwired splits. Even if you avoid intentionally making them with cables, you sometimes can get unintentional hardwired signal splits. Some equipment use direct signal path split between different outputs instead of using an extra buffer or at least isolation resistor for different outputs. This makes the output circuit susceptible to rogue loads. There are many units with this same exact problem out there.

The power off distortion can also sometimes bite with professional audio equipment. Gear that is not powered up can cause “off gear distortion” on line level signals that are driving multiple pieces of audio gear. Dave Rat video Audio Gremlins – When Off is Bad shows how to hunt down an audio gremlin that can bite you even when the gear is off. Good demonstration Dave. And it only happens with some pre amp designs and not others. Very interesting (but annoying) way to create soft clipping :)

In video comments the recommendation was at “Hardwire splits should never be used to connect a monitor board because the FOH board and monitor board will interact just that way.” In real life there are many installations using hardwire or transformer splits for various reasons that work more or less well on real life situations.

The follow up video Will an ISO Transformer Stop “Off Gear Distortion”? looks for solutions for this problem. The tests showed that the issue can passe right through the transformer as expected. So even transformer splits might not get you away from this problem. Many people have experience with transformer isolated splitters is whatever is behind the transformer doesn’t affect the direct signal, but this is not always the case. Though the losses in the iso split can reduce the effect considerably, but there can still be noticeable effect on some cases.

On digital equipment it’s input protection devices on the console inputs, that are connected to the internal power supply. When the internal power supply is off, they clip at +/- 1 diode drop, rather than only when the input level exceeds the internal power supply level. The clipping at the front end prior to the ADC can be caused by input protection devices like diodes, which will be connected to the input and then one each to supply and ground. When there is no supply (unit off), you now have essentially back to back diodes to ground. That will cause clipping on both side of the waveform starting at (roughly) 1.4Vpp (basically 2 diode drops). Lower the signal well below that and you won’t see it. Because the didoes turning on at the waveform peaks drastically momentarily drop the impedance on the input, it will affect anything else directly connected to that input. It will also impact things behind a transformer, as transformers do not provide total impedance isolation, although the impact will be a bit reduced. Bottom line is, don’t ever plug any signals into active circuitry gear that is powered off – you are asking for trouble. Of course others have mentioned that hardwired splits can potentially cause issues due to parallel impedance so should never be used, but that is a little drastic. If you understand your source impedance and the console input impedance, and how multiple impedance terminations on a single line interact, then using hardwired splits is fine in many instances. If you don’t understand all that, then it’s safest not to ever use hardwired splits. But as you can see, even with transformer isolation, you still need to understand this kind of stuff. Different equipment have different types of output stages, and some of them can drive non-linear loads better than other. A lot of “better” outboard gear has bypass relays, so when the unit is off (or power fails during the show) inputs are connected directly to outputs.

Information sources:



Be the first to post a comment.

Leave a Comment

Your email address will not be published. Required fields are marked *