Created on 2010-12-15 00:35:00
An Interview with AP's VP of Engineering Tom Kite on Multitone Analysis with APx500
Basically, what’s a multitone and how does it work?
“A multitone is just an audio signal with many tones all combined together. We provide prebuilt multitones with APx500 that have 3, 5, 7, and 32 tones. During analysis, an FFT (Fast Fourier Transform) is performed and the audio spectrum is divided into a number of “bins.” The bins that contain tones are used for level and phase results. The leftover bins are used to get the distortion and noise results.”
What are the advantages of multitone testing?
“A big one is getting 21 results in 1 second. That makes it ideal for production test. But it’s also a huge advantage in the broadcast, cable, and Internet fields, because it lets you do comprehensive audio testing over very long distances. In fact, you can embed a couple of seconds of multitone in your program stream, and an APx500 instrument at the receiving end will automatically trigger on it and take a measurement. A multitone doesn’t need any coordination between the generator and analyzer, so it doesn’t really care if the signal is coming from half-way around the world, or from an MP3 player.”
The APx500 software comes with prebuilt 3, 5, 7, and 32 tone multitones. Why would you choose one over another?
“In some applications, like production testing, a small number of tones is sufficient to catch manufacturing defects, and an excessive amount of data collection and detail isn’t desired. Additionally, the more tones you add together, the lower each one needs to be to keep the overall level the same, and this can lead to less accuracy in the distortion and noise measurements. But, 32 tones is great for a detailed frequency response graph. So, it’s a choice of what’s most important to you.”
Why would you want to make a customized multitone?
“Standard multitones suffice for the vast majority of testing. However, there are a number of special cases where it is useful to alter some of the parameters. Equalizing a multitone is useful for correcting speaker irregularities when testing microphones, or when testing circuits or devices that have an emphasis or de-emphasis curve. You can add crosstalk tones and vary their number to give you a detailed look at crosstalk across the frequency range. Concentrating the measurement points in a specific frequency range can be useful when studying the cut-off area of a filter, or the resonant point of a speaker. And by making the FFT longer, you get more noise bins to average, so your noise measurements become more accurate and repeatable—at the expense of increasing test time. Oh, and of course you can alter the start and stop frequencies to match the bandwidth of your device. The APx500 software lets you create customized multitones, or alter existing ones, and store them with a project.”
What’s TD+N? How do I interpret the results?
“A multitone excites both harmonic and intermodulation distortion mechanisms in a device. When we subtract all the fundamental tones from a multitone, what we have left is TD+N (Total Distortion plus Noise). While the numbers can’t directly be compared to THD+N (Total Harmonic Distortion plus Noise), it’s very useful in that it gives you a single figure representing all the impurities in the audio signal.”
Does the new multitone analyzer measurement add any new results?
“Yes. One of the new results is noise density spectrum. While this isn’t a measurement seen on audio equipment spec sheets, it’s a routine specification for IC’s, like microphone preamps and op amps. It shows how noise is distributed across the audio spectrum, and it’s very useful for circuit designers when optimizing the noise performance of a circuit.”
