Setting up a Test Bench Part 5, PC-based measuring sets

Status
Not open for further replies.

SergeAuckland

Certified Measurist
Wammer
May 6, 2008
18,607
1,915
173
HiFi Trade?
  1. No
PC-based measuring sets

Using a PC and a Sound Card can be a very cost-effective way of having an ‘all-in-one’ measuring set, albeit with limitations. An old laptop and a good quality USB Sound Card is really all one needs to achieve some quite detailed measurements.

The quality and facilities of the Sound Card are key, however, as the quality of the generator is dependent on the DAC in the Sound Card and its subsequent analogue output, and the quality of the measurement depends on the analogue input and ADC. 

Nevertheless, with a USB soundcard capable of 24 bit operation at a sample rate of 98kHz, one can get distortion levels down to 0.005% and a measurement bandwidth of around 48kHz, sufficient for checking distortion at up to 10kHz or thereabouts. With a better soundcard offering 192kHz sampling, then a wider bandwidth is possible, but going any further in bit depth is unlikely to bring any improvement given that the limiting factor will be analogue noise. Even the difference between 16 bit and 24 bit only brings a small reduction in the minimum THD+N measurement due to analogue noise.

One virtually standard facility in all PC measuring software is the FFT, Spectrum Analyser. This allows the waveform to be analysed according to its frequency spectrum and the amplitude of each component. Very useful for seeing what the levels of each distortion component is, 2nd, third, fourth etc harmonics, similarly with intermodulation distortion and the ability to see any spurious frequencies standing out in a noise spectrum, which might indicate instability in an amplifier.

The real downside of using a PC and Sound Card, is however that as these Sound Cards are made for recording use, the inputs and outputs are uncalibrated as far as voltage is concerned. They are therefore not a substitute for an audio millivoltmeter or calibrated signal generator, but a very useful addition.

Measuring software is plentiful, and quite a bit of it either Freeware, or sold at very small charge.

Which Software you choose will depend on what it is you’re trying to measure. They fall broadly into two main types, Test Bench software and Acoustic Measurement Software, although there can be a lot of overlap.  I don’t have huge experience with lots of different packages, but the ones I have tried are as follows:-

Test Bench Software:-

RightMark Audio Analyzer (RMAA)      (audio.rightmark.com)

This is a very common software, which does either manual or a whole suite of automated measurements. One sees the results tables often reproduced in on-line reviews.

Visual Analyser     (sillanumsoft.org)

This is my main measurement software package, I just like the way things are presented. The Oscilloscope section is particularly good, albeit uncalibrated as are all Sound-Card apps unless individually done for each user.

PassMark Soundcheck    (Passmark.com)  and TrueRTA (Trueaudio.com)

Little to say about these, I have them and they work, but don’t use them much.

Acoustic Measurements:-

Room EQ Wizard (REW)  (www.roomeqwizard.com) This is probably the best known of the acoustic measuring softwares for home use.  As the name suggests, its main application is in measuring the loudspeakers in the listening room, and creating an EQ curve, which can be directly imported into certain DSP-based equalisers. It can also measure loudspeakers pseudo-anechoically, which is the main use to which I’ve put it, doing frequency sweeps and adjusting the equaliser so that the loudspeakers measure flat anechoically, then checking that the room doesn’t mess that up too badly!

ARTA (ARTAlabs.hr) Also a great piece of acoustic measuring software. It does pretty much what REW does but, in some respects is easier to use. It also is useful for deriving the Theil-Small parameters for loudspeaker drivers.

AudioTester (audiotester.de)  This one crosses over both acoustic and bench measurements, andprovides all sorts of 2D and 3-D display modes. It’s also able to measure a driver’s T-S parameters. I find this one a bit cluttered and the documentation isn’t very clear, at least not to me, but it does seem to be able to do a lot.

For all these acoustic measurements, a calibrated Measurement Microphone is essential. The microphone must be physically small, (or at least, thin) and omnidirectional. The calibration is the most important part, and indeed accounts for most of the purchase price. For example, the Behringer ECM8000 microphone can be bought for under £50 uncalibrated, and indeed, one can download ‘generic’ calibration files for free but if one looks at the possible spread of frequency responses from that microphone, the results could be several dBs out at higher frequencies. Nevertheless, if the aim is just to EQ the room at low frequencies, then the ECM8000 with a generic calibration file is OK. For ‘proper’ measurements, I would recommend either having an ECM8000 calibrated properly, or more easily, just buying a calibrated EMM-6 microphone from Dayton Audio. These are available from several sources, and are only a little more expensive than an uncalibrated ECM8000. Because I fancied doing it, I bought an uncalibrated ECM8000 and calibrated it myself against a calibrated Earthworks microphone I borrowed, but I couldn’t be bothered to do it again. Probably wouldn't have done it even then if the EMM-6 had been available.

External USB sound-cards like the Lexicon I use, or Focusrite or M-Audio, Roland or many others all have microphone inputs for acoustic measurements and line inputs for bench measurements. These will have 48v phantom power to power condenser microphones, so they can be used for measurement as well as their primary purpose, that of live recording. When using a USB sound card for bench measurements, make sure that the maximum input is sufficient for your needs. Almost certainly, you will need some sort of variable attenuator if you’re going to be doing amplifier output power measurements, as no USB sound card I know of can accept an input of 20+ volts without clipping. The best you’ll get is +22dBu (10v) and some are lower than that.

If you are only interested in doing acoustic measurements, not bench measurements at all, then a calibrated USB microphone is a sensible alternative. These have the ADC built-in to the microphone and plug directly into a USB input, and get their power from that. You can use the headphone or audio line output on the PC to get the signal out and select the audio inputs and outputs accordingly.

I think the best way of selecting a package is to download the software, install it and play with it. See whether it makes intuitive sense to you, or do you have to keep referring to the ‘Help’ screens to find out how to do something? Do they even have decent documentation or ‘Help’?

I have also found that which ones work properly depends very much on the audio drivers you’re using, and which version of Windows you use. For example, my RMAA works perfectly on an old Windows XP computer I keep for live recording. It uses an even older PCMCIA Digigram VXPocket card which is great for 16 bit 44.1 recording. (It doesn’t go on a network, so no issues with security.)  However, I can’t get RMAA to work at all with my Windows 10 laptop using a Lexicon USB sound card, although Visual Analyser works fine with the Lexicon.  

S.

 
  • Upvote
Reactions: insider9
Status
Not open for further replies.

Forum statistics

Threads
113,444
Messages
2,451,263
Members
70,783
Latest member
reg66

Today's Birthdays

Latest Articles