dboomer

All that and you can leave your sticks on the dashboard and park in handicapped zones!

98.6 - Keith I’m truly ashamed of myself for this one.

Limiters for speaker protection "limit" the amount of voltage an amplifier can deliver to a speaker. This should go a long way to protecting the speaker from thermal burnout (you still need to protect against excursion limits with HP filters). It should be said for the record that it doesn't matter what the waveform looks like ... only the amount of power contained in it ... square wave, sine wave or anything in between. Amp clipping itself does not burn speakers and underpowering does not burn speakers. Too much voltage for too long will burn the voice coil or separate it from the former itself. Protecting your amp from clipping has very little to do with anything except your amp (and most amps have clip limiters built in) Let's start with the basics and add exceptions as we go along Set the limiter for about a 20:1 ratio or higher. In the real world there is almost no practical difference between 20:1 and infinity:1 (because you have almost nothing to drive above threshold. If you set the threshold at "0" you'd have to drive 20 dB more to get the output to rise 1 dB). Anything less than 20:1 doesn't give you much protection. The attack and release will depend of the averaging style of your limiter and the frequencies involved. You probably want the attack on the order of 15ms - 20ms for woofers and maybe 5ms for tweeters (not written in stone). The release is generally set to be 10 times the attack time. Now comes the tricky part ... how do you set the threshold? Now you need to know how much voltage your speaker will handle. It will be called the "continuous", "average", or "rms" rating. If it is called "program" then cut that number in half and if "peak" divide by 4 (generally). Understanding the "real" number is a bit difficult ... so setting a limiter like this depends on how good this number really is. Adjust (down) as necessary for your own comfort. It is helpful to know by which method the manufacturer has used to make this rating. If done by the AES method then it considers the minimum impedance. Speakers have a "nominal" rated impedance e.g. 4 ohms or 8 ohms but the true impedance varies with frequency. It would be typical for an 8 ohm speaker to vary between a low of maybe 6 ohms (at low frequencies) to a high of 20 or 40 ohms at high frequencies. So your amp is delivering varying amounts of "wattage" at different frequencies. You may have to subtract a little bit of power handling ability. Convert the speaker rating to Volts; sq root (speaker rating * impedance) so a 500W 8 ohm speaker looks like this ... square root (500*8) = 63.24 volts (just type "square root (500*8) into google and it will do the math for you) At this point you can simply drive signal through your limiter to your amp and bring down the threshold until you reduce the output of the amp to 63Vac measured on an AC voltmeter (assuming you have proper gain settings)

Brand and model aside, how is it that a less-tuned system responds worse to the feedback killer than a well-tuned system? Feedback always happens for the same reason. That reason is you cross the unity threshold point (where the sound of the speaker is as loud into the mic as the sound of your voice into the mic). So maybe the question is how and where did I do that. Let's try a simplified explanation ... Let's consider one mic and one speaker. There is a distance from your mouth to the mic and a different distance from the speaker to the mic. So when the sound from your mouth and the sound from the speaker arrive at two different times it sets up a comb filter. The top of the teeth of the comb are in phase and the gaps between the teeth are out of phase. You can only get feedback on the top of a tooth in the comb and cannot get feedback in the gaps. When you change these distances you will change the number and the width of these teeth. Your sound system is not perfectly flat ... it has peaks and dips ... especially off axis from the speaker and because of the pattern of the mic. Now consider that ragged frequency response. As you turn up the system when one of the peaks in your system crosses the unity threshold right on top of one of the tops of a comb tooth ... presto ... you get feedback. A "more tuned system" simply minimizes the sound level of the speaker from reaching the mic. The first thing here is to aim the speaker away from the mic as best you can while still pointing it in the direction needed. Now there are two tunings you need to make. The first is tuning the speaker so that the sound reaching the audience sounds pleasing. The second tuning is cutting those frequencies that present themselves as peaks to the mic. These may or may not be the same thing. In fact they can be exactly opposite. Feedback (if systems had no distortion harmonic or otherwise) would be a single frequency and would be close to a sine wave. So specific that you could remove that single frequency. The human ear can't identify such a narrow bit of info so your brain just fills it in (integrates it). If fact your ear/brain combination (on average) can't hear the number of points that would be 1/3rd of an octave at frequencies below about 400Hz and about 1/6th of an octave above that. There is of course a first frequency that crosses the threshold and then a second and a third and so on.You need to grab those single frequencies only without grabbing anything else. The standard practice is to throw a 1/3rd oct graphic at this ... but a 1/3rd octave graphic has filters that are about an octave wide ... which you can hear and will miss. They are called 1/3rd oct graphics because the filters are SPACED a third of an octave apart. So what you really want to do is tune your speaker for listening with a 1/3rd oct GEQ for the listening part and use a very narrow parametric EQ for the feedback part.