Head-Fi.org › Forums › Equipment Forums › Sound Science › Amplifier analyzer
New Posts  All Forums:Forum Nav:

Amplifier analyzer

post #1 of 17
Thread Starter 

we have an open project which is about building a device with logical kernel which is able to measure an amplifier (SS or tube or Hybrid) characteristics like frequency response with given tolerance, THD, 2nd, 3rd harmonics, rise time and settle time, output impedance,... all with high accuracy. the device is some kind of oscilloscope or analyzer which will do the job of a very expensive chain of tools. a display in the front panel will indicate the functions. using this tools, every amateur or professional individual will be able to measure the amplifier

the project is on the drawing board yet, any idea is much appreciated.


Edited by LogicAudio - 10/31/13 at 10:12am
post #2 of 17
Quote:
Originally Posted by LogicAudio View Post
 

we have an open project which is about building a device with logical kernel which is able to measure an amplifier (SS or tube or Hybrid) characteristics like frequency response with given tolerance, THD, 2nd, 3rd harmonics, rise time and settle time, output impedance,... all with high accuracy. the device is some kind of oscilloscope or analyzer which will do the job of a very expensive chain of tools. a display in the front panel will indicate the functions. using this tools, every amateur or professional individual will be able to measure the amplifier

the project is on the drawing board yet, any idea is much appreciated.

 

Doesn't RMAA already do everything you listed, except for Zo?

 

For Zo you can figure that out with even simpler tools than RMAA. 

post #3 of 17
If I were to do this DIY, I'd measure the amp driving a real speaker load. You can put the speaker in a closet so it doesn't drive you nuts while testing. Then I'd make a 40 dB voltage divider from resistors and take the signal from the speaker terminals. I'd feed that into the line input of a high quality modern sound card, such as my Focusrite Scarlett 8i6 which is very flat and has very low distortion. I'd play sine waves at various frequencies into the amplifier while recording from the sound card. Then I'd analyze the recording using an FFT tool in Sound Forge, or a freeware equivalent. You'll also need to create two-tone tests for IMD. Check out the freeware program Visual Analyser from Sillanum Software:

http://www.sillanumsoft.org/

--Ethan
post #4 of 17

Although these are rather user-unfriendly, and therefore probably not very useful for most people, I have some free (with source code) utilities that can be used for generating and analyzing various test signals (including some tests not performed by RMAA), and creating graphs. They can be downloaded from the third link in my signature.

post #5 of 17

I suggest adding the ability to add DSP manipulation tools to the test signals (tone sweeps, m-sequences, square waves and so forth). By this I mean quality equalizers (convolution, parametric, and so forth), reverb, compression/decompression, and so forth. The ability to load/generate your own test signals would also be a nice feature.

 

Linearity tests would be welcomed, and arbitrary multi-tone test signals would be useful. Support for THD20, jitter tests, and THD vs power/Vrms, and slew rate would be fantastic.

 

Perhaps adding a dynamic range stress signal such as a 0 to 20-96 kHz "wide-band" "impulse" signal.

 

Not sure how useful, but perhaps tools to load/record/compare music files (different common formats) in loopback under different loading conditions...

 

Good support for ASIO, WASAPI and so forth across different rates would be great, as well as a well implemented set of calibration routines.

 

Most importantly, stability, accuracy and ease of use.


Edited by ultrabike - 10/31/13 at 1:50pm
post #6 of 17
Thread Starter 
Quote:
Originally Posted by nikongod View Post
 

 

Doesn't RMAA already do everything you listed, except for Zo?

 

For Zo you can figure that out with even simpler tools than RMAA. 

 

Uh actually it's first time I heard about RMAA. I think it's a software which needs some other tools like sound card, you name it. but we want built something all-in-all and ready to go

 

Quote:
Originally Posted by EthanWiner View Post

If I were to do this DIY, I'd measure the amp driving a real speaker load. You can put the speaker in a closet so it doesn't drive you nuts while testing. Then I'd make a 40 dB voltage divider from resistors and take the signal from the speaker terminals. I'd feed that into the line input of a high quality modern sound card, such as my Focusrite Scarlett 8i6 which is very flat and has very low distortion. I'd play sine waves at various frequencies into the amplifier while recording from the sound card. Then I'd analyze the recording using an FFT tool in Sound Forge, or a freeware equivalent. You'll also need to create two-tone tests for IMD. Check out the freeware program Visual Analyser from Sillanum Software:

http://www.sillanumsoft.org/

--Ethan

rightnow we don't think about speakers. that's something different. we want to built something that can measure the amplifiers parameters. if one would like to measure a speaker, then a calibrated mic + phantom power sound card and a software like ARTA is the most common way to do the job.

 

Quote:
Originally Posted by stv014 View Post
 

Although these are rather user-unfriendly, and therefore probably not very useful for most people, I have some free (with source code) utilities that can be used for generating and analyzing various test signals (including some tests not performed by RMAA), and creating graphs. They can be downloaded from the third link in my signature.

we want to build a device which is extremely user friendly. no need to things like signal generators and so on. connect output RCA of the analyzer of amplifier input. connect speaker output of amplifier to the input of analyzer. switch on both devices and here we go. chose the required function for example sine wave, and you'll see everything on the monitor like two sine waves, one before amplification and other after that. there will be also numbers for comparison. chose frequency response function and press start, accurate frequency response of amplifier will we shown with a graph and also number within a minute. you wanna check a brand honesty about their product specification? you doubt your amplifier THD meets 0.000000000001% which is stated in the specs?! give this device a try and it will tell you whole the truth within a minute

 

Quote:
Originally Posted by ultrabike View Post
 

I suggest adding the ability to add DSP manipulation tools to the test signals (tone sweeps, m-sequences, square waves and so forth). By this I mean quality equalizers (convolution, parametric, and so forth), reverb, compression/decompression, and so forth. The ability to load/generate your own test signals would also be a nice feature.

 

Linearity tests would be welcomed, and arbitrary multi-tone test signals would be useful. Support for THD20, jitter tests, and THD vs power/Vrms, and slew rate would be fantastic.

 

Perhaps adding a dynamic range stress signal such as a 0 to 20-96 kHz "wide-band" "impulse" signal.

 

Not sure how useful, but perhaps tools to load/record/compare music files (different common formats) in loopback under different loading conditions...

 

Good support for ASIO, WASAPI and so forth across different rates would be great, as well as a well implemented set of calibration routines.

 

Most importantly, stability, accuracy and ease of use.

 

Uh you post is lot of many good suggestions. I printed it right away and will mention in our designing think tank. I appreciate it

post #7 of 17

there's Visual Analyzer and other free soundcard software too - some serious amateurs even pay for Liberty Instruments, SpectraPlus, other analysis SW

 

there don't seem to have been any new high quality, cheap soundcards to compete with ESI Juli@, EMU or RME "prosummer" offerings of  >5 years ago but you can still buy those

 

as for standalone hardware you're intending to compete with Picoscope, National Instruments or AudioPrecision? Beat out Russ at Twisted Pear to releasing a USB Audio Class 2.0 ESS ADC?

 

any proposing "open" design should really look at, understand the commercial offerings in the field - and have a good story for why they can do it for so much less than companies that invest man-years of skilled engineering time on both hardware and software, then make, sell thousands of units in open markets with competitors


Edited by jcx - 11/1/13 at 7:28am
post #8 of 17
Quote:
Originally Posted by LogicAudio View Post
 

Uh actually it's first time I heard about RMAA. I think it's a software which needs some other tools like sound card, you name it. but we want built something all-in-all and ready to go

 

 

:facepalm:

Way to do your market research. 

:facepalm:

post #9 of 17
Quote:
Originally Posted by LogicAudio View Post

rightnow we don't think about speakers. that's something different. we want to built something that can measure the amplifiers parameters.

You missed my point. An amplifier behaves differently when driving a loudspeaker, versus a resistor, versus nothing. If you want to assess how an amplifier behaves when it's working as intended, it should be connected to a loudspeaker. You don't use a microphone to measure the speaker! Read my advice again, which was very specific.

--Ethan
post #10 of 17
Quote:
Originally Posted by EthanWiner View Post


You missed my point. An amplifier behaves differently when driving a loudspeaker, versus a resistor, versus nothing. If you want to assess how an amplifier behaves when it's working as intended, it should be connected to a loudspeaker. You don't use a microphone to measure the speaker! Read my advice again, which was very specific.

--Ethan

 

Yup.

 

LogicAudio, I would also suggest to evaluate the free version of ARTA which I think is pretty solid in frequency domain, and the free version of TrueRTA which I like for some time domain stuff.


Edited by ultrabike - 11/1/13 at 10:13am
post #11 of 17
Thread Starter 
Quote:
Originally Posted by jcx View Post
 

there's Visual Analyzer and other free soundcard software too - some serious amateurs even pay for Liberty Instruments, SpectraPlus, other analysis SW

 

there don't seem to have been any new high quality, cheap soundcards to compete with ESI Juli@, EMU or RME "prosummer" offerings of  >5 years ago but you can still buy those

 

as for standalone hardware you're intending to compete with Picoscope, National Instruments or AudioPrecision? Beat out Russ at Twisted Pear to releasing a USB Audio Class 2.0 ESS ADC?

 

any proposing "open" design should really look at, understand the commercial offerings in the field - and have a good story for why they can do it for so much less than companies that invest man-years of skilled engineering time on both hardware and software, then make, sell thousands of units in open markets with competitors

yes it's exactly what we're gonna do

post #12 of 17
Thread Starter 
Quote:
Originally Posted by nikongod View Post
 

 

:facepalm:

Way to do your market research. 

:facepalm:


not much wonder in a 3rd world :(

post #13 of 17
Thread Starter 
Quote:
Originally Posted by EthanWiner View Post


You missed my point. An amplifier behaves differently when driving a loudspeaker, versus a resistor, versus nothing. If you want to assess how an amplifier behaves when it's working as intended, it should be connected to a loudspeaker. You don't use a microphone to measure the speaker! Read my advice again, which was very specific.

--Ethan


I think it was a common language barrier. not I got you. yes amplifiers and speakers behave differently when connected together. each of them affect the other. damping factor and impedance behavior,...

but I don't think a commercial amplifier with a stated 0.03% of THD in the specification, is measured under the working condition! I think all the product are measured in a sound laboratory not in a listening room. take me for example as a producer. I might have Wilson Audio XLF for evaluating the final result of an amp I'm building. that speaker has it's own impedance curve, phase shift, etc...

it's crossover have their own manners, the speaker drivers have specific QTs QMs,... (in the relation with damping factor) so measuring my product when connected to a speaker which is not common and everybody just doesn't have, it not a professional work.

post #14 of 17
Thread Starter 
Quote:
Originally Posted by ultrabike View Post
 

 

Yup.

 

LogicAudio, I would also suggest to evaluate the free version of ARTA which I think is pretty solid in frequency domain, and the free version of TrueRTA which I like for some time domain stuff.


yes we use ARTA for measuring speakers.

post #15 of 17
Quote:
Originally Posted by LogicAudio View Post
 


yes we use ARTA for measuring speakers.

 

Yes, but it's got some unique features that may be applicable to amplifier measurements. For example, it can do IMD for multiple and arbitrary tones. RMAA is a bit more stringent in that. The interface is a bit more complex though. I think RMAA has an upper hand in simplicity, but it has some issues with calibration and stability. For example, the latest version of RMAA 6.3.0 supports Focusrite 2i2 ASIO at 44.1 and 48 kHz, but I'm having problems with other sampling rates. Without support for at least 96 kHz, THD20, which is not currently supported by RMAA, is difficult... should they want to support it. So good soundcard support is IMO something that should not be overlooked.


Edited by ultrabike - 11/2/13 at 11:05pm
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Sound Science
Head-Fi.org › Forums › Equipment Forums › Sound Science › Amplifier analyzer