Audio amp distortion indicator: feeding a distorted signal into it.

[Ed.Kloonk]

Hi.

I've been looking at an old Altronics ~175W mono amplifier kit I bought years ago and need to dust it off.. I intend to give it to someone who wants to play music from his computer loudly when he has a party.

The circuit board contains a ideal distortion sensor circuit that lights up a LED when the amp is over-driven so to protect tweeters from DC. All it does is compare the input and the output and light-up the LED if there is a difference such as if voltage is clipped on the output. It's only an indicator, not -really- a protection circuit. The vintage of the amp is mid 90's and at the time matched a carousel multi-CD player nicely.

If the amp was fed an already distorted signal from a computer with a cheapy sound card that seems to be able to clip the output audio without to much difficulty, do you reckon this would likely defeat the warning indicator on the amp and blow the damn speakers anyway without the drunken user knowing?

Loaded question I know, but I'd be interested to hear your thoughts/experiences.

[ejeffrey]

First off, the problem with clipping is not DC to the tweeters: that is impossible in most setups as they are capacitively coupled.  The problem is that clipping is a non-linear process that creates harmonics: feed a low frequency sine wave into a clipping circuit and you get a square wave out that has not only the fundamental frequency but all odd-numbered harmonics.  Music has relatively little acoustical power in the high frequency range, and tweeters have much lower power handling ratings than the system power.  When you overdrive an amp, you convert a lot of the bass/midrange signal into high frequency components and put too much power (at high frequency, not DC) into the tweeters.

Anyway, yes: if you feed a clipped signal into a power amplifier, you will get a clipped signal out even if the amplifier itself is not saturating.  The amplifier clipping indicator will not trip, and whatever level of warning it provided will be defeated.

[Ed.Kloonk]

Bugger. Back to the drawing board. 

[jahonen]

It is not very easy to do good overload indicator, one would need accurate model of the speaker thermal model to see when the actual thermal power limit is reached. Although tweeter do not handle long term overload well, the protection should not be activated by short transients.

I believe the active speaker manufacturers have devoted a quite bit of research to develop good protection circuits. Although situation is much easier if the crossover is at the low-level side, because one can then drive the bass/midrange amplifier(s) into clipping without any distortion in the tweeter channel.

Regards,
Janne

[ciccio]

I do not know the actual schematic, but usually the  clipping indicators in audio amplifiers operate on one of two principles:
1)  compare output voltage to DC supply voltage: when the output peak voltage reaches a value near power supply, the indicator flashes.
This is the simpler, and most times meaningless, version: it will not indicate a current limiting or slew limiting condition.
Sometimes it is simplified by referring to a fixed voltage level, and not to supply voltage: it will indicate when the output level reaches a fixed value that is near the maximum .

2) look for actual distortion: this may be done in two ways:
2a) feed input and output to a comparator, via correct resistive dividers. Not easy to do so reliably (in my experience)
2b) look for voltage present at some points of the amplifier (mostly at the output of the input stage, inside the feedback loop).
If the amplifier is not clipping (or into any distortion condition) there will be no actual signal, but a spike will be present when any type of distortion is present. This spike will trigger a comparator that will light the LED. This circuit was first implemented by CROWN (Amcron) in the 70's. They called it IOC (input-output comparator)

Both solutions will work only if the amplifier is distorting the input signal: if the input is already distorted, the amp will try to reproduce it faithfully, including all the distortion.

I do not see any danger for the tweeters from such condition.