Audio Test Set (eg Lindos) alternatives

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Limitedheadroom

New member
Joined
Feb 24, 2019
Messages
2
Hi,

I'm considering a build that has quite a bit of calibration required at the end and the guide says audio test set (eg Neutrik / Audio Precision / Lindos) is required.

This is obviously a very expensive item to purchase if I may not need it again, and I may as well just buy the item rather than try to build it.  Is there a cheeper way to achieve the kind of thing I need to do?  It looks like it is applying signal at a known dBu value, and reading levels back at the output.  Also measuring TH+D and power supply ripple.

If I don't already have this equipment, am I just fooling myself with considering a build like this?  I'm confident I can construct the board, but obviously that's not the whole task!

Advice please.
Thanks
 
The noise floor of a conventional USB audio interface will be below the theoretical noise of a small resistor. So there really is almost no reason to use fancy hardware anymore. Maybe for calibration or if it has good attenuation / amplification for optimizing levels around the DUT. But attenuation is usually easy and frequently you have to make cables for the DUT in which case you can integrate attenuation into the cable (such as for a mic pre for example you can put a 150K:150R directly into the XLR).

The important component now is software. I actually don't love any of the audio analyzer software packages out there. But maybe that is because I refuse to spend money on it. Personally I just use GNU Octave which is the open source corollary of Matlab which is a declarative language interpreter for performing various, potentially very sophisticated, mathematical transformations on arbitrary data. For audio, I just create a recording and then use Octave to generate hi-res spectra. Most people just use something like Room Eq Wizard. At one point I was using something called DSSF3 from ymec.com (although I think that actually costs a little money).

But your question is a little vague. If you can share more about what you ultimately want to test, the answers to your question will be better.
 
Yes there are cheaper ways of doing it.  An audio interface with some shareware will do it.

The only snag is you have to know what you are doing to set it up properly and be able to trust the results. The thing about a dedicated test instrument is that it’s calibrated and you can trust the results.
 
I would recommend looking ar RMAA (Right Mark Audio Analyser); it's free, it's simple to use. Now, as others have mentioned, ther's the issue of calibration. You need a good meter, with a frequency response that allows proper checking of 1kHz signals. Almost any serious multimeter can do that now.
I would also recommend you install a virtual oscilloscope; that one is pretty good: http://www.sillanumsoft.org/download.htm
 
Thanks for the tip about  Sillanumsoft Scope/fft Abbey ,
,at 96khz it worked quite nicely , it seemed to give credible distortion readings . ,not quite  as versatile as REW , bears a striking resemblance to Ymec ,which I noticed Squarewave also used . It was always a bit of a head scratch why REW didnt have the scope , but I guess its on its way into a full release of the program soon .

One thing I did notice about REW , when you click on or off  the sig gen , it quickly fades in or out the  tone , and avoids any nasty clicks or burps . its thoughtfull features like that make all the difference .
 
The OP wrote he needed to calibrate level, measure PSU ripple and THD. A good multimeter will cover everything except THD.  A test signal source would also be needed. I’d say a multimeter is a minimum requirement for building anything.

What’s a percent or two of  extra THD anyway? You could tune by ear for least funk.  Somewhere in the middle is probably close enough.
 
I use REW most of the time now but like all software and sound card based systems you do need to be able to calibrate it. At the minimum you need a reasonably accurate audio voltmeter. You only need to calibrate the sytem initially at 1Khz. You can then do a loop back test to calibrate it over the entire audio spectrum. I like REW because it is free, easy to use and works on Mac, PC and Linux.

Cheers

Ian
 
The answer is it all depends....I'm a Lindos user (a LA1 analogue unit rescued from a skip, a very secondhand LA100 pair and a MS20 bought new)....the "instant access" and portability were key features for me as was the ability to function without a pc but in many ways it's about personal preference...what works for you.  My choices are also driven by personal history, I started with an old Ferrograph test set for Distortion measurements and a MJS401 for noise and level, then a Wayne Kerr AMS1(?)....whilst dreaming of having access to AT or HP kit..The LA100 made life so much simpler...automated frequency respose measurement and the ability to create custom test sequences was stunning....and thus my future was set.

Re second hand kit I have over the years bought two LA102 measuring units (both at very good prices) and the calibration of both was almost spot on (within 0.2dB)..The LA1 is wonderful, a piece of pure nostalgia...fully analogue (save the frequency meter) and consequently subject to the joys of old pots, dreadful interboard connectors and the 'Studio Interface' which can be entertaining to repair).
Good luck with your choices
tc


 
I own an Audio Precision Sys-1, not only because its calibrated but due to its very low residual distortion and ultra-low distortion signal generator. I dont know the distortion spec of a digitally generated sine wave played through a sound interface, but I dont think it is as low as the one on an AP-1.

The other system I own is an HP-8903, you can get one on e-bay for a good price nowadays, I made a MATLAB code to control it through GPIB and then plot  the results, this way the HP closely resembles the AP-1, again the distrotion spec of the HP is not as good as the AP. I agree that the shareware PC programs are good enough for rock n'  roll, but in my opinion, if you want to measure very low THD such as the output of a well designed power amp, then you need something better than the sound card + shareware bundle.

My first THD kit was an HP334a and a Tektronix SG505, they are great for single frequency readings, you can do a multi-frequency reading by taking different readings manually but its a bit painful and time consuming setting up all the knobs on the 334a, I still own them thou, you can get the 334a for a really low price on e-bay, but the SG505 is really expensive nowadays.
 
user 37518 said:
I dont know the distortion spec of a digitally generated sine wave played through a sound interface, but I dont think it is as low as the one on an AP-1.
Obviously, it depends on the soundcard's performance, but typical high-quality interfaces have a residual distortion of significantly less than -90dB (0.003%), which is comparable to a sys1 (-92dB). The big difference is the practicality of the sys1 software and the bulletproof I/O's.

The other system I own is an HP-8903, you can get one on e-bay for a good price nowadays, I made a MATLAB code to control it through GPIB and then plot  the results, this way the HP closely resembles the AP-1,
That's interesting. What kind of set-up do you use? I would have thought GPIB interfaces were almost extinct. What version of OS do you use?
I see there are USB-to-GPIB and Lan-to-GPIB adapters, but I'm concerned with compatibility aspects. When I was the AP distributor we had a number of issues with customers that already had GPIB cards in their systems, that didn't cooperate well with SYS1.
 
abbey road d enfer said:
Obviously, it depends on the soundcard's performance, but typical high-quality interfaces have a residual distortion of significantly less than -90dB (0.003%), which is comparable to a sys1 (-92dB). The big difference is the practicality of the sys1 software and the bulletproof I/O's.

My AP is more in the order of 0.0005% distortion if I recall correctly, the residual distortion of an audio interface may be low but what about a digitally generated sine wave, is it also ultra low distortion?

abbey road d enfer said:
That's interesting. What kind of set-up do you use? I would have thought GPIB interfaces were almost extinct. What version of OS do you use?
I see there are USB-to-GPIB and Lan-to-GPIB adapters, but I'm concerned with compatibility aspects. When I was the AP distributor we had a number of issues with customers that already had GPIB cards in their systems, that didn't cooperate well with SYS1.

Believe it or not GPIB is alive and well, although its slowly being replaced by ethernet. I use a National Instruments USB to GPIB interface its expensive but it is the best GPIB I/O there is, I could use Labview if I wanted to but im fond of MATLAB, my OS is Windows 10. The HP is not the easiest beast to handle but with a little time working with the GPIB messages I ended making very efficient codes to measure THD vs Frequency, Power vs Frequency and THD vs Power. You can use the Pete Millet's programs to control the HP unit but I rather roll my own.
 
abbey road d enfer said:
Newer AP systems use digitally generated sinewaves; the best one exhibits -117dB THD - that's 0.00014% . The others are between -100 and -110dB.

I guess I'll have to generate some sine waves with my DAW and measure distortion, something tells me it wont be anywhere near those figures.
 
user 37518 said:
I guess I'll have to generate some sine waves with my DAW and measure distortion, something tells me it wont be anywhere near those figures.

You can use REW to do this in loop back mode. This will measure the total distortion of the loop which includes that in the analogue loop. If you have a 16 bit soundcard (internal or external) the best you will achieve will be around the -90dB mark which is about 0.003%. This is more than good enough for 99% of purposes.

Cheers

Ian
 
I use the "poor man's AP" - an Amber 5500 Distortion and Noise Measurement System.  I bought it because it is what my mentor used, and because I could not afford an AP.  I think I paid around $500 usd.  It certainly does what I need it to do - mainly levels, distortion, freq response - but it is all manually controlled.  Incidentally, the designer now works at AP  8)
 
mjrippe said:
I use the "poor man's AP" - an Amber 5500 Distortion and Noise Measurement System.  I bought it because it is what my mentor used, and because I could not afford an AP.  I think I paid around $500 usd.  It certainly does what I need it to do - mainly levels, distortion, freq response - but it is all manually controlled.  Incidentally, the designer now works at AP  8)

If I recall correctly, the AP engineers used to work at Tektronix on the AA501 distortion analyzer.
 
Back
Top