Help me understand limiting

B

Brandon White

Freshman
May 16, 2011
45
0
6
Grand Rapids, Michigan, United States
Hello fellow sound people,

I understand there are two main types of limiters. Peak & RMS, to deal with mechanical & thermal driver damage. Up until this point I have been sort of faking my way through setting my limiters.

I would really appreciate some sound mathematics and and how to find the variables necessary to execute these equations.

Thanks everyone


Sent from my iPad
 
Re: Help me understand limiting

Brandon White said:
Hello fellow sound people,

I understand there are two main types of limiters. Peak & RMS, to deal with mechanical & thermal driver damage. Up until this point I have been sort of faking my way through setting my limiters.

I would really appreciate some sound mathematics and and how to find the variables necessary to execute these equations.

Thanks everyone


Sent from my iPad
Click to expand...
The term RMS and thermal are often misused and misunderstood.

Generally the term RMS is used to mean the "continuous" rating of the loudspeaker.

The continuous means the "average" voltage-as read on an averaging meter with a 6dB crest factor signal applied at a level that does not cause damage to the driver.

The peak is 6dB higher this signal-or the peak of the noise.

Not causing damage DOES NOT mean that the loudspeaker is performing fine. Often the driver is into power compression at this point.

So the "thermal limit" is "generally" 1/2 to 1/4 the power (0.7 to 0.5 of the voltage) of the continuous rating.

The peak limiter is set to react as fast as possible, while the thermal limiter is often set to work after a couple of seconds after the threshold has been exceeded.
 
Help me understand limiting

Could you provide an example? I'm pretty sure I understand each limiters place in the system. I feel I'm just falling short of a technical / mathematical approach. Are you saying a driver rated at "600w" continuous should be "thermally limited" to 300 -420w? And if so how do I establish these settings in my DSP?


Sent from my iPad
 
Re: Help me understand limiting

Brandon White said:
Could you provide an example? I'm pretty sure I understand each limiters place in the system. I feel I'm just falling short of a technical / mathematical approach. Are you saying a driver rated at "600w" continuous should be "thermally limited" to 300 -420w? And if so how do I establish these settings in my DSP?


Sent from my iPad
Click to expand...
The 0.7 was the VOLTAGE-NOT the wattage. 0.5 (half) voltage is actually 1/4 the power, not half.

Double or half the power is a 3dB change. P1/P2 10 log=dB difference. Double or half the voltage is a 6dB change. V1/V2 20 log=dB difference

For a 600 watt driver, half power would be 300 watts. If the impedance is 8 ohms, the voltage would be 49 volts.

The simplest way to set a DSP for that level would be DISCONNECT THE LOUDSPEAKER!!!!.

Now apply a sine wave within the intended pass band of the speaker. Hook a voltmeter (preferably one that is accurate over a wide range of freq) to the output of the amp. Drive the sinewave till the meter reads more than 49V. 60V would be a good number-but as long as it is more than the "target" voltage.

Now lower the threshold of the amp until the volt meter goes down to 49V. and you are done.

Hopefully the limiter in the DSP is THE LAST thing in line before the output-including after ALL level controls.
 
Re: Help me understand limiting

Brandon White said:
I understand there are two main types of limiters. Peak & RMS, to deal with mechanical & thermal driver damage. Up until this point I have been sort of faking my way through setting my limiters.

I would really appreciate some sound mathematics and and how to find the variables necessary to execute these equations.
Click to expand...

Hey Brandon,

I Am Not A Mathematician, so I will leave that to someone else, but I wrote an article on this very topic a while ago. I may be sticking my neck out here, but I will say that I do not believe that peak limiters as implemented in any current DSP are capable of controlling excursion. This is because the signal sent to the loudspeaker does not translate into position, but acceleration. The signal has no idea where the cone is, to determine that you would need to do a little modeling of the enclosure resonance, driver resonance, and a double integral of the signal. By the time you finished that math the excursion peak would probably have passed.

IMHO long term thermal limiters work very well, I like to set them at 1/2 the AES rating of the transducer. To control excursion, a 2nd order HPF at the tuning frequency of your enclosure is your best bet. It's OK to add an EQ bump to bring the frequencies at F3 back up after applying the HPF if you have enough amp swing to support that.
 
Re: Help me understand limiting

Bennett Prescott said:
Hey Brandon,

I Am Not A Mathematician, so I will leave that to someone else, but I wrote an article on this very topic a while ago. I may be sticking my neck out here, but I will say that I do not believe that peak limiters as implemented in any current DSP are capable of controlling excursion. This is because the signal sent to the loudspeaker does not translate into position, but acceleration. The signal has no idea where the cone is, to determine that you would need to do a little modeling of the enclosure resonance, driver resonance, and a double integral of the signal. By the time you finished that math the excursion peak would probably have passed.

IMHO long term thermal limiters work very well, I like to set them at 1/2 the AES rating of the transducer. To control excursion, a 2nd order HPF at the tuning frequency of your enclosure is your best bet. It's OK to add an EQ bump to bring the frequencies at F3 back up after applying the HPF if you have enough amp swing to support that.
Click to expand...

+1 while I don't know what is done inside modern speaker processors, DSP is certainly capable of making the calculations to predict driver excursion. This is not simple and ultimately involves knowing specific parameters of box and drivers. As Bennet shared, applying a voltage to a driver voice coil creates a motive force, not a position, so you need to measure not only how much voltage is applied, but for how long that voltage is applied, and how much travel the driver can tolerate, surround compliance, etc. To simplify somewhat, the LF threshold voltage would be different for different frequencies. This is something that digital processors could actually manage, while you can probably protect a driver adequately while forfeiting some modest amount of performance with a single excursion threshold, or modest shaping of the side chain.

IMO this is too complicated for most end users, and another reason I advocate using powered cabinets.

JR

PS: I recall a discussion of the physics involved, with graphs, and charts, and equations, maybe 10 years ago over at that other SR forum. I just tried a quick search and didn't find it.
 
Re: Help me understand limiting

Limiting is an important tool, for sure. How much effort you put in will affect how close to the edge you can run, but IMO pushing too hard means you are already on borrowed time. Remember that doubling the amplifier power only gives an additional 3dB of output, so worrying about the last volt or two is probably not worth it.

+1 to outsourcing this to the manufacturer either in the form of self-powered speakers or a manufacturer-integrated system.
 
Re: Help me understand limiting

Ivan,
Thank you for the practical application. It has defiantly helped me wrap my mind around the concept.

Bennett,
If you are saying peak limiters are inadequate in protecting driver over excursion, and I should use a 2nd order HPF instead. Where should I set it? The speakers peak voltage rating?


Sent from my iPad
 
Re: Help me understand limiting

I just thought I would add, that RMS limiters are called RMS, because they typically calculate the RMS voltage (a statistical quantity) over some aperture size.

On some DSPs you can control this aperture size and it needs to be set so that it is big enough to respond to the lowest frequency which will pass through the limiter. On many DSPs this RMS statistical block is then coupled to a compressor which responds at some rate to the current RMS voltage (might instantaneous, a moving average, or some other further statistics) calculation. As the RMS voltage approaches the limit, it will engage the compressor and try and hold that level. Of course, fast transients will pass right though such a system. It's like a PID loop (temperature controller).

Like everything in DSP, almost every implementation is different, so... you'll have to talk to the manufacturer of the DSP you are actually using for specifics.
 
Last edited:
Re: Help me understand limiting

Brandon White said:
Ivan,


Bennett,
If you are saying peak limiters are inadequate in protecting driver over excursion, and I should use a 2nd order HPF instead. Where should I set it? The speakers peak voltage rating?


Sent from my iPad
Click to expand...
A high pass filter sets the freq. It has nothing to do with the speakers voltage rating.

HOWEVER-IF you can find out what the excursion limits of the speaker are (something nobody publishes and it depends on what is considered "excursion limit"), then a limiter could be used with a freq sidechain (most limiters don't have this) to help control the excursion-so the limiter would be associated with the freq response of the speaker.

This is a very ADVANCED limiter and setup.
 
Re: Help me understand limiting

Brandon White said:
Bennett,
If you are saying peak limiters are inadequate in protecting driver over excursion, and I should use a 2nd order HPF instead. Where should I set it? The speakers peak voltage rating?
Click to expand...

Brandon,

I would not set a peak limiter at all. I would set a thermal limiter, for sure, and some implementations are better than others (e.g. some actually reflect the load impedance). If you want to set a peak limiter do it by ear, turn the box up until it sounds funny and then adjust the limiter down until it sounds better. Then listen to make sure you haven't made the whole box sound worse.
 
Re: Help me understand limiting

I generally go with Ivan's method (4th post in this thread). Here's a helpful online (and downloadable) calculator for determining max speaker voltages: http://www.poulpetersen.dk/Appn/gblimthc.html . One slight tweak I would suggest is sweep the sine wave frequency around within the intended passband to get the highest voltage reading on the meter, then proceed with increasing the level past the target and lowering the threshold to the target voltage.

-Mark
 
Top Bottom