Neutrik Minilyzer ML1HANDHELD AUDIO ANALYZER Have you been avoiding the plunge into the world of geekdom? Neutrik has solved that problem by packing an arsenal of the most 1/01/2001 7:00 AM Eastern
HANDHELD AUDIO ANALYZERHave you been avoiding the plunge into the world of geekdom? Neutrik has solved that problem by packing an arsenal of the most essential and powerful measurement tools into an affordable, easy-to-use package called the Minilyzer ML1. This is a pocket-sized audio analyzer designed for the many times you've needed to investigate an audio mystery but were afraid to call a technician.
Beyond the primary capabilities of this powerful little box - detailed in the seven major headings below - the ML1 includes a handful of little touches that make it as audio-friendly as a piece of test equipment can be. Test equipment is often unbalanced, a serious issue when trying to measure a balanced device under real-world conditions. The ML1 includes both balanced XLR and unbalanced RCA inputs. Three AA batteries supply power, automatic turn-off time is adjustable, and four user presets allow the ML1 to boot into your favorite mode. There is even a headphone jack.
Of the seven main functions, many are intermixed within the large, easy-to-read, backlit LCD screen. For example, the LCD is large enough to display "level" in large bold characters, while simultaneously including an analog-style linear meter at the bottom, a frequency counter in the upper left corner and an inspired "input balance" indicator in the upper right corner (see Function 5).
Note: Despite its piccolo footprint, the ML1 is quite capable of getting you into trouble. Translation: It is just as easy to take a bad measurement as a good one. Read the sidebar, "The Ultimate Test," to learn how to compile real data.
FUNCTION 1: LEVELThe ML1 displays the RMS value of a sine wave test signal in millivolts (mV), as well as the relative level in dBu and dBV, referenced to 0.775 volts and 1 volt RMS, respectively. Additionally, the ML1 has a "relative" mode for precise comparisons of two or more signals - left and right, for example - with a signal-to-noise ratio as wide as 119 dB!
For years, technicians purchased the Fluke 8060A ($479 at www.imcm.com) for its ability to measure RMS volts and dBu. The Fluke doesn't do dBV; its noise floor bottoms out at -74 dBu (compared to -99 dBu for the ML1), and it's too old to have the DSP do all the cool things that the ML1 can do. Because the Fluke is a volt-ohm-milliamp meter, it has no upper level limit - compared to the ML1's 7.75-volt/+20dBu ceiling. I would like to see this ceiling raised 10 dB.
FUNCTION 2: THD+NWhen measuring Total Harmonic Distortion + Noise (THD+N), the "+N" implies that noise is included as part of the measurement process. How can it not be? Unless pushing a device purposefully into the red, you hope that both distortion and any noises - hiss and hum - are way down in the sub-basement of random electron movement. The only way to separate distortion from noise easily is to use the built-in Third Octave analyzer, which will be discussed under the heading "Function 7."
Inside the ML1 is an A/D converter capable of 119 dB of dynamic resolution. The sample rate limits bandwidth (frequency response) to 20 kHz, as pointed out by Neutrik in the manual. This can potentially yield different (most likely lower) readings relative to more sophisticated gear, but only when in the Linear (flat response) mode. Analog test equipment is not bound to the sample rate. The Hewlett Packard Model 8903B Audio Analyzer, for example, has a 30kHz "window."
The ML1 also includes built-in filters so that the measurement window can be "weighted" (bandwidth restricted) to include what is relevant to the test. For example, the popular A-weighted filter (as per IEC 651) is often used, because it reflects the ear's sensitivity to noise. The other filter options are C-Message (IEC 468-4), HP22 (highpass at 22 Hz), HP60 (highpass at 60 Hz), HP400 (highpass at 400 Hz) and a Voice-band filter. These filters are available in the Level, THD+N and Third Octave modes.
Note: Searching the IEC on the Internet for specifications yielded nothing - their search and destroy mechanism is about as useful as Microsoft Help, and they want money for documentation I couldn't find. If you want an IEC clue, go to the heading "Function 6."
I used the ML1 to measure the THD of two oscillators - the GTC Tone Plug and the Neutrik MR1 Minirator - and two mic preamps - a Great River transformerless prototype and an Altec 1566 tube preamp. The Tone Plug is a handy "generator in an XLR plug" commendable for its size, not cleanliness. (Check out Table 1 for the results.) While the Minirator is respectable for its price range, to truly measure the Great River's performance, a better oscillator would be required. As you can see, there is almost no difference in the performance of the MR1 alone compared to its use with the Great River preamp.
FUNCTION 3: VU+PPMThe ML1 emulates three metering standards: mechanical VU meters (referenced to +4 dBu), Type I and "Nordic" Peak program meters (PPM, +6dBu ref), and Type IIA PPM (+8 dBu). The user can reconfigure all references. Both VU and PPM are simultaneously displayed. Each includes a numeric peak hold indicator, plus there are two integration time options: normal (Type I and Nordic: 5 ms. Type IIa: 10 ms). In Fast mode, the integration time is 1 ms for all standards.
While observing the output of a Panasonic SV-3700 DAT, I immediately realized that the ML1 could use one additional metering standard capable of being calibrated to digital audio's 0dBfs maximum. For example, the SV-3700 has a -18dBfs nominal reference; the range should accommodate a "low" of -20 dBfs and a high of -10 dBfs. Neutrik could probably turn the VU+PPM function into a stand-alone stereo product with both analog and digital inputs (and a larger LCD screen). It would be a helpful mastering tool for viewing accurate peak information, while maintaining some consciousness of "volume" per the VU meter. The VU meter should not be solidly in the red while the PPM would be kissing 0 dBfs.
FUNCTION 4: POLARITYThe polarity test requires both the MR1 and the ML1. The MR1 generates a pulse that is easily detected even after traveling through the air. Selecting Polarity on the ML1 engages the input select option - a choice of either the XLR/RCA connectors or the built-in microphone. It works!
FUNCTION 5: BALANCEQuite unexpectedly, the very first "if only" feature I thought of was an "input balance" indicator - not left and right, as this is a mono box - of the incoming signals on pin 2 and pin 3. This feature is a reality on the ML1, putting this gizmo and me on the right foot from the very beginning. A 6dB level problem in the analog world is not uncommon - active balanced outputs can become damaged or interrupted via a dirty patch cord or bad cable - and the ML1 will tell which pin isn't doing its share of the work.
Note: The ML1 "loses" what little headroom it has if the signal is not precisely balanced. I noticed this when testing the SV-3700, whose pin 2 and pin 3 outputs were particularly unmatched, reducing the max headroom in this case to +19.3 dBu.
FUNCTION 6: SWEEPSweep has two options, the traditional RMS level vs. frequency, or time vs. any of the following: level, THD+N or frequency. Getting this mode to function was most difficult, and the manual was not perfectly clear (perhaps due to translation). I got results simply by copying the example in the manual. Here, a picture was worth a thousand words. The best example would be to plot THD+N to show how distortion increases with increased levels.
FUNCTION 7: THIRD OCTAVEThe third octave analyzer can display the audible bandwidth from 20 to 20k Hz in 31 bands. As mentioned, THD+N does not separate distortion from noise. Harmonic distortion is a lack of sonic cleanliness relative to the input signal. Some vacuum tube gear is famous for its pleasing even-order (octave) harmonics. Input 1 kHz, for example, and push the device into its nonlinear region (not hard clipping) and watch the second harmonic (2 kHz) pop up. See Figs. 1 and 2.
The cursor can be moved to each of the 31 bands to confirm both frequency and amplitude, the latter in both dB and %. An op amp circuit in hard clipping will produce odd-order harmonics (square waves are made from these), and the third harmonic of 1 kHz is 3 kHz. See the Altec preamp specs in the table on page 138.
FUNCTION 8: SCOPEIf you read my article on troubleshooting bad capacitors using square waves, the "oscilloscope-like" waveform display is the icing on the cake. Of course, the perfect companion to the ML1 is the $139.95 MR1 Minirator, featuring both sine and square waves. As with most LCD scopes, the ML1 does not have amazing resolution, further hampered by the 20kHz bandwidth, which softens square waves until 10 kHz looks like a sine wave. That's okay. About the cost of a cheap-but-real `scope alone, the ML1 is much more likely to see active duty.
GOOD DADAReviewing the ML1 was a good brain exercise, emphasizing the relative ease of making a bad measurement compared to the work involved to acquire good data. That said, it would be helpful if Neutrik offered an external attenuator as an optional accessory to pad the output of the MR1 (or any oscillator) so that noise measurements would reflect the device under test and not the audio source. Otherwise, there is only one flaw. The maximum input of the ML1 is +20 dBu, which is not high enough and easily compromised if the signal balance is not perfect.
The ML1 is small enough to be kept in a control room or clipped to a belt for those technicians on the move who are also looking to start a fashion trend. (PDAs might suddenly become less cool!) Consider how many times you've returned to the scene of an audio crime only to find no suspects and no problem? Now you can whip out this nifty little "tri-corder" whenever a problem occurs. By creating the Minilyzer, Neutrik has given more people the power to troubleshoot. The more you use it, the more you'll understand that a little science never hurt anyone. If you weren't lucky enough to get one of these in your Christmas stocking last month, you can pick one up for about $399.
Neutrik USA, 195 Lehigh Ave., Lakewood, NJ 08701; 732/901-9488; fax 732/901-9608; www.neutrikusa.com.
The carpenter's adage "measure twice, cut once" couldn't be more applicable to the art of acquiring good data. Measuring nominal level from a device is child's play, but measuring signals at or into the noise floor is quite another. This is not a fault of the ML1. Several attempts were made until the results were consistent. You might not think much about "impedance," but when taking measurements, both the source and the destination impedance must be addressed.
When measuring mic preamp performance at both minimum and maximum gain, the gain structure between it and the sound source (an oscillator) becomes quite critical. You can't simply turn down the level at the oscillator, because, in some cases, a buffer amplifier follows the level control. The oscillator's output amplifier has plenty of noise when looking down the high-gain barrel of a mic preamp. The top of Table 1 proves that reducing the level of the Minirator increases noise revealed in the THD+N measurement. Fortunately, the Third Octave analyzer helps to "see" the various noises.
I fashioned a quickie attenuator between the oscillator and the preamp using a 1-kilohm pot across pin 2 and pin 3 of the female XLR. Pin 3-f was connected to pin 3-m, while the wiper from the pot was connected to pin 2-m. With the Great River preamp set to max gain, I adjusted the trim pot so that preamp output fell just under +20 dBu (+19.9 dBu), keeping in mind the ML1's inability to tolerate anything higher. I then checked the attenuator output (-45.7 dBu) and calculated the gain to be 65.4 dB.
Once the gain was determined, I removed the generator and attenuator and connected a source impedance that represented the average mic. I chose 150 ohms, because it was used to create the preamp's published specs. To calculate Equivalent Input Noise, EIN = the noise floor + the gain setting. An older style signal-to-noise ratio (S/N) measurement was also made. S/N = the max output before clipping - the noise floor.