The basic technology behind an amplifier is really quite simple—it increases the voltage of a signal to match the power needs of a speaker. However, when you get into the details, it gets more complicated. There are a number of features available on amplifiers, and determining exactly what you need can be daunting. One feature, in particular, that has become prominent in recent years is a built-in digital signal processor (DSP). Signal processing is a feature that has become increasingly available at every point in the signal chain, so the particular advantages of having it in an amplifier can seem confusing, although the advantages are very real.
To better understand what having onboard DSP offers and how to use it, I reached out to Emilian Wojtowycz, Solutions Manager, Restaurant and Retail at HARMAN Professional Solutions. Emilian served as a front of house engineer and sound system designer before joining HARMAN in 2010.
[SKD]: Once you’ve determined the wattage and power needs for your amplifier, there are still a lot of other features to consider—particularly DSP. How do you go about determining what you need for your application?
[EW]: The most important thing to remember is that you never buy an amplifier in isolation. You’re buying an amplifier depending on the speakers you’re driving, and you select the speakers depending on your sound application. So, you have two other steps that come first before you start looking at amplifiers. It all starts with looking at the space you’re trying to fill with sound and deciding the sound pressure level (SPL) you want to get out of the system—that will help ensure the sound system is loud enough. Then, depending on that demand, you select your speaker. That, in turn, determines how powerful an amplifier you need for that individual speaker or transducer.
However, even when you get to that point, you still have a number of other decisions you need to make before you determine the specific amplifier you need. For example, if you have a three-way speaker enclosure (in other words, a speaker with low-, midrange- and high-frequency transducers), it may not have a passive crossover network, so you would need three amplifier channels, each one optimized for a particular transducer. Even if it does have a passive crossover, you may want additional control over what you’re sending to each transducer to increase efficiency, so you would still need three channels. Either way, you would need a DSP to limit the frequency range that you’re sending to each transducer.
[SKD]: When you say optimizing for each transducer, what do you mean?
[EW]: In order for speakers to generate sound waves, they need to physically move the air. Lower soundwaves are larger, and creating them requires more air movement. Because of this, speaker transducers that generate lower frequency signals use more electricity. When you’re creating high frequencies, the transducer is vibrating very quickly and is a very small speaker, like a tweeter or compression driver. Those don’t take very much energy.
Transducers only handle a certain amount of voltage, depending on their power rating, and if you send it more power than a speaker can handle, it will burn up. Likewise, if you send it frequencies that it can’t produce, you’re just wasting power, adding in distortion to the signal and could even damage the speaker. If you look on the speaker spec sheet, it will tell you what amount of power and what frequencies it’s able to handle, so you need to match the signal from the amplifier to those specifications.
For example, take a JBL Control 28-1 indoor loudspeaker. The specifications list a frequency range of 45 Hertz (Hz) to 20 Kilohertz (kHz). However, that is measured at negative 10 dB. If you look at the line below that specification, you see a listing for the frequency response. This indicates what frequencies are heard at plus or minus 3 dB. In this case, that rating is 62 Hz to 16 kHz. In other words, you won’t hear as much for frequencies outside the response range, even if they fall into the overall frequency range. You might want to augment those frequencies with a subwoofer to add more low end that the Control 28-1 can’t make. So, any sounds from 60 Hz to 80 Hz and below, you want to send to the subwoofer, because it can make those sounds more efficiently.
[SKD]: How do you ensure you’re only sending the speaker the frequencies it needs?
[EW]: To do this, you need some type of crossover or bandpass filter—essentially some way to send certain frequencies on one channel and other frequencies on another channel. You can have a crossover inside the amp or before the amp, but either way, a crossover sets the frequency point at which the sound “crosses over” and starts going to the Control 28-1 rather than the subwoofer (in this case, the crossover point would be somewhere in the 60 Hz to 80 Hz range). That way, you’re only sending frequencies that speaker can make. If you send a bunch of high frequencies to a subwoofer, it can’t make them. You’re just sending a lot of additional electricity to the subwoofer that’s unnecessary. The same thing happens if you’re sending very low frequencies to a speaker like the Control 28-1. If you send music like hip hop that has a bunch of low end, you’re not going to hear it from the Control 28-1. Even if you turn it up, you’re eventually just going to burn up the speaker. Speakers need specific tunings, because they make specific frequencies. If you send a speaker frequencies that are below its cabinet tuning, you can cause excursion damage, even if you aren’t sending it too much power.
[SKD]: What you’re talking about is a bandpass filter to limit the frequency range and a limiter to control the maximum voltage. Couldn’t you set that up in DSP earlier in the signal chain?
[EW]: You certainly could. HARMAN provides speaker tunings for its JBL speakers that list the right voltages and frequencies that a certain transducer or speaker can make. You can go to the JBL website, download them and input them into your DSP individually for every speaker type in your system. However, in amplifiers like the Crown CDi DriveCore, these tunings are already built into the unit.
One of the advantages of this is that you can actually limit the voltage output of the amplifier, so you’re not going to give more voltage than the transducer can handle. LevelMAX limiting in Crown amplifiers is actually limiting the voltage output of the amp. If you’re using DSP upstream of an amplifier (and we make some excellent signal processors), you can certainly setup speaker tunings in them. However, once you get to the amplifier, you now have an additional gain stage that is bringing it up from line level, and that needs to be set correctly to ensure you’re not sending too much voltage out of the amplifier. If your settings are mismatched, you could either be limiting too early and missing out on some efficiency and headroom available on the amplifier or limiting too late and you could still blow the speaker. That’s one of the advantages of having DSP and speaker tunings built into the amplifier.
[SKD]: I would think there would be some logistical and integration advantages to that approach as well, correct?
[EW]: Absolutely. Having the speaker tunings built into the amplifier makes the system a lot easier to configure and install, especially if you have multiple speakers or multiple speaker types. The Crown CDi DriveCore Series actually has pre-installed speaker tunings for nearly every JBL installed speaker available. So, if you have a system with a number of JBL speakers, it is a simple matter to add the amplifiers and get them setup; you can do that from the color front panel interface or remotely over the network. Simply select the speaker you have from a list, and you’re done.
I’d like to offer a big thanks to Emilian for explaining more about how speaker tunings in amplifiers work. If you have insights into signal processing in amplifiers or speaker tunings, please share them in the comments!