Input overload distortion

[Dave2002]

I realise that I mentioned input overload distortion in a recent post, and I've also had recent experience of what I think was just that. My experience was hearing a friend try to play dvds from a computer through his amp, and there was far more distortion than I would have expected. This was much reduced when I suggested turning the volume down on the computer, and up on the amp.

What got me thinking is that I don't really understand the mechanisms for this form of distortion. I can understand clipping distortion at the output of power amps - even pre amps, as this is most probably simply due to the inability of the power supply to handle voltage and power extremes.

I must confess to being very ignorant about what happens at the input end of an amplifier which could cause distortion. A straightforward search on t'web hasn't shed any more light on this.

[MrGongGong]

I would imagine that he is using the headphone output of the computer , these are mostly horribly noisy even on a £1,000 macbook
Of course overdriven preamps can be put to great use as Hendrix shows ..... (but probably not useful for a Feldman concert )

[Dave2002]

On reflection I think I'm getting there. In a gated circuit (either valve or some form of transistor), if a small positive or negative going signal causes maximum current to flow, then the output of that gated device would be.a heavily clipped version of the analogue input signal - typically either on or off, or maybe 3 level 0, -ve,+ve depending on the input signal. The actual details will depend on the circuit design used, but this shows how input overload distortion can occur. It could be symmetrical, or asymmetrical if the max current only flows for one polarity of input. In any case, once the distortion has been introduced to the reproduction chain it is virtually impossible to remove, and can sound extremely unpleasant.

[Serial_Apologist]

Of course overdriven preamps can be put to great use as Hendrix shows ..... (but probably not useful for a Feldman concert )

Daryl Runswick told me a lovely story concerning unintended feedback at a live performance of a Cage piece he was giving with the group Electric Phoenix. At one point the sound engineer's sleeve accidentaly caught in one of the faders. Cage had specially written the piece for the group and was in the audience. Afterwards, with much trepidation, he was asked for his opinion on its performance, and he replied that the best bit had been the feedback!

[Gordon]

I realise that I mentioned input overload distortion in a recent post, and I've also had recent experience of what I think was just that. My experience was hearing a friend try to play dvds from a computer through his amp, and there was far more distortion than I would have expected. This was much reduced when I suggested turning the volume down on the computer, and up on the amp.

That's the right solution. Keep the INPUT level lowest and set listening volume by the latest possible control. The amp will expect typical input around a couple of hundred millivolts to produce its maximum output power. You have got too much going in from the DVD player. The early stages of the amp have an overload margin but that is limited.

What got me thinking is that I don't really understand the mechanisms for this form of distortion. I can understand clipping distortion at the output of power amps - even pre amps, as this is most probably simply due to the inability of the power supply to handle voltage and power extremes.

I must confess to being very ignorant about what happens at the input end of an amplifier which could cause distortion. A straightforward search on t'web hasn't shed any more light on this.

Clipping happens when the amplifying device [valve or transistor] requires a voltage swing at its output that approaches its power supply voltage. This a matter of design. At the rated output power the voltage swing at the final stage is defined and so the overall gain or amplification needed will be the ratio of the design output swing to the design input. Exceed these design limits and ask the amp to produce any more it can't, so if you drive it too hard at its input it will clip the output.

You MUST always have less input than the max design input given in the spec. That spec usually requires the volume control to be set at maximum for maximum gain. The gain control usually reduces the drive level to the output stage so if that is near its minimum and the output is still clipped then there is far more than the maximum rated input being applied and there is danger that the early stages of the amp are also in overload. The only solution to avoid clipping is: KEEP THE INPUT DOWN and the final volume control low or near mid range!!