Ok, that's the basis.An oscilloscope and a signal source are the bare minimum i think. Oh and a multimeter naturally. More than one is handy.
But then? I would like to know how people here get to measure about let's say distorsion , THD, TIM, Pass Band(i.e Gain), Phase.
Yea I'm wondering how you guys pump out these awesome FFT graphs and get the measurements effebi mentioned.
I have a 1MHz signal generator, Fluke DMM as well as several chinese DMM's and some analog meters. I also have a 100MHz analog scope and a 50MHz DSO.
I know a few members use some pro DAC units to produce their test frequencies like the DAC1 PRE and I presume they have some equally pricy ADC equipment. Is their software that you use to measure your setup with a patch cord then measure your test subject and it will give you the response of just the unit under test?
I have a 1MHz signal generator, Fluke DMM as well as several chinese DMM's and some analog meters. I also have a 100MHz analog scope and a 50MHz DSO.
I know a few members use some pro DAC units to produce their test frequencies like the DAC1 PRE and I presume they have some equally pricy ADC equipment. Is their software that you use to measure your setup with a patch cord then measure your test subject and it will give you the response of just the unit under test?
Ok, that's the basis.
But then? I would like to know how people here get to measure about let's say distorsion , THD, TIM, Pass Band(i.e Gain), Phase.
Gain and phase measurements, pass band (frequency response) can all be done accurately with a scope and generator. Same for rise/fall times etc.
I think many quote distortion measurements and so on direct from running simulations of the design rather than actual real world measurements on the finally constructed article.
The equipment to measure distortion and so on down to 0.00xxx levels etc doesn't come cheap... and using PC USB "add ons" just doesn't cut it
So you mean for frequency response you input say, a 1 V RMS at 10 Hz, measure the output, then set it 20 Hz, measure the output, .......100Hz, 200Hz.....1000 Hz and so on. Then plot the measured points on a nice Excel logarimic graph, and you are done?Gain and phase measurements, pass band (frequency response) can all be done accurately with a scope and generator.
That is it?
better soundcards are quite capable
better soundcards with 24/192 ADC easily reach 0.001% THD and IMD and the Nyquist frequency of 96 KHz fits well with traditional "audio analyziers" 80 KHz THD measurement bandwidth
with indirect IMD techniques you can push nonlinear distortion resolution down to the spot noise floor - often as good as -130 dB with 1 second averaging
check RMAA loopback on ESI Juli@ or ASUS Xonar Essence STX
the "accuracy" of 'scope measurements can't compare - but the 'scope and signal generator are required for looking for stability, HF performance issues
better soundcards with 24/192 ADC easily reach 0.001% THD and IMD and the Nyquist frequency of 96 KHz fits well with traditional "audio analyziers" 80 KHz THD measurement bandwidth
with indirect IMD techniques you can push nonlinear distortion resolution down to the spot noise floor - often as good as -130 dB with 1 second averaging
check RMAA loopback on ESI Juli@ or ASUS Xonar Essence STX
the "accuracy" of 'scope measurements can't compare - but the 'scope and signal generator are required for looking for stability, HF performance issues
Last edited:
That's it basically. You want a constant voltage undistorted signal going into the amp. You can use 1/2-octave steps or whatever, but of course the more measurements the more accurate the plot. You don't need a scope necessarily for gain and FR; an AC voltmeter with its own proper freq response can be used.
Ok so do you mind taking us through the steps you would do with your equipment if you just built yourself a new headamp? Detailing what you are actually doing with the equipment and how you are compiling your results.
I am working on creating a DSS with a goal of at least a -100dB noise floor and <0.001 THD that is capable of 0.01Hz to 5MHz that way I have a very accurate signal and I'm not spending time weeding out noise on my project when its the source all along.
The ASUS Xonar has some impressive specs. 124dB SNR, 0.0003% THD and almost no harmonics to worry about. Is it pretty common to use something like Audacity to output a test tone to your headamp then record the output and use some of its plugins to measure responses?
I don't want to waste my time or money on equipment I think I need that's why I'm so interested in how you guys get your readings. Specifically nwavguy.blogspot.com he has ALOT of measurements I'm not sure how he gets and OPC on this forum who built "The Wire" with its crazy noise floor.
I am working on creating a DSS with a goal of at least a -100dB noise floor and <0.001 THD that is capable of 0.01Hz to 5MHz that way I have a very accurate signal and I'm not spending time weeding out noise on my project when its the source all along.
The ASUS Xonar has some impressive specs. 124dB SNR, 0.0003% THD and almost no harmonics to worry about. Is it pretty common to use something like Audacity to output a test tone to your headamp then record the output and use some of its plugins to measure responses?
I don't want to waste my time or money on equipment I think I need that's why I'm so interested in how you guys get your readings. Specifically nwavguy.blogspot.com he has ALOT of measurements I'm not sure how he gets and OPC on this forum who built "The Wire" with its crazy noise floor.
I use a function generator and also a very old low distortion wien bridge oscillator and scope for most tests, such as tweaking opamp compensation. Non of this would show on pure test tones (sine),
http://www.diyaudio.com/forums/chip...g-audio-integrated-opamps-51.html#post2012422
http://www.diyaudio.com/forums/head...le-ended-class-headphone-amp.html#post2075959
Perhaps someone who uses Audacity etc could provide some real "off screen" shots of say testing something at 10 or 50 or 100khz using a fast risetime squarewave generated by the software... because although I'm no expert whatsoever in software/applications like Audacity... I just don't think it will do it.
I'd be interested to see though... some real world measurements similar to mine above.
http://www.diyaudio.com/forums/chip...g-audio-integrated-opamps-51.html#post2012422
http://www.diyaudio.com/forums/head...le-ended-class-headphone-amp.html#post2075959
Perhaps someone who uses Audacity etc could provide some real "off screen" shots of say testing something at 10 or 50 or 100khz using a fast risetime squarewave generated by the software... because although I'm no expert whatsoever in software/applications like Audacity... I just don't think it will do it.
I'd be interested to see though... some real world measurements similar to mine above.
With my Audigy II ZS about the best it can handle is 1Khz square waves, go past that and it doesn't even resemble a square wave any more (and even at 1Khz it has noticable ringing) Not sure about other cards.
It does sine waves nicely up to 96Khz (running in 192Khz mode) but square waves, forget it! Note this is generating them (and measured on a scope), I'd assume that it is just as bad at recording them if you have a separate function generator, but unfortunately I don't so can't say.
Tony.
It does sine waves nicely up to 96Khz (running in 192Khz mode) but square waves, forget it! Note this is generating them (and measured on a scope), I'd assume that it is just as bad at recording them if you have a separate function generator, but unfortunately I don't so can't say.
Tony.
With my Audigy II ZS about the best it can handle is 1Khz square waves, go past that and it doesn't even resemble a square wave any more (and even at 1Khz it has noticable ringing)
Tony.
Thats pretty much what I would expect tbh
Whay Analogue oscilloscope
Why do you prefer analogue scope, and not digital?
I evaluate to bay this: ISO-TECH | Test and Measurement | Oscilloscopes | Digital Oscilloscopes | ISO-TECH IDS8000 Series
Or are there any other scopes that you gentlemen will recommend analogue / digital?
Why do you prefer analogue scope, and not digital?
I evaluate to bay this: ISO-TECH | Test and Measurement | Oscilloscopes | Digital Oscilloscopes | ISO-TECH IDS8000 Series
Or are there any other scopes that you gentlemen will recommend analogue / digital?
Dummy loads: 4r0 50W or 100W, 4 of them.
Switched attenuator: <1% tolerance (<0.1dB).
Mains bulb tester.
Learn how to use each of your instruments.
I rarely try to measure absolute values. My instruments and ambients are not accurate enough.
I try to compare voltages or frequencies, or currents, or components, etc.
Switched attenuator: <1% tolerance (<0.1dB).
Mains bulb tester.
Learn how to use each of your instruments.
I rarely try to measure absolute values. My instruments and ambients are not accurate enough.
I try to compare voltages or frequencies, or currents, or components, etc.
OK so I wanted to know whether the Audigy II ZS (and my on-board high definition audio) did as poorly at inputting square waves as they did outputting them. A square wave generator was something I didn't have in my kit, so today I built one.
It produces a pretty clean square wave on my scope. 1st pic attached. but the sound card mungs it greatly, 2nd pic.
It was supposed to be a 10Khz square wave (approximately) but it looks like it is 9.2Khz, I guess if I want to do better than my breadboard prototype I should get some precision resistors and cap for the timing part of the circuit Not bad for about $2 worth of parts though.
Tony.
It produces a pretty clean square wave on my scope. 1st pic attached. but the sound card mungs it greatly, 2nd pic.
It was supposed to be a 10Khz square wave (approximately) but it looks like it is 9.2Khz, I guess if I want to do better than my breadboard prototype I should get some precision resistors and cap for the timing part of the circuit Not bad for about $2 worth of parts though.
Tony.
Attachments
It produces a pretty clean square wave on my scope. 1st pic attached. but the sound card mungs it greatly, 2nd pic.
It was supposed to be a 10Khz square wave (approximately) but it looks like it is 9.2Khz
You don't have enough samples to conclude anything based on the soundcard measurement. Increase the sampling rate to 192 kHz if you can. That would give you 19-ish samples per period rather than the current four (assuming you're running 44.1 kHz sampling).
So it's off by 800 Hz (8 %). Big whoop! As you say. Either get precision components or make the frequency adjustable.
Yep. Pretty nice for $2 worth of parts.
~Tom
With my Audigy II ZS about the best it can handle is 1Khz square waves, go past that and it doesn't even resemble a square wave any more (and even at 1Khz it has noticable ringing) Not sure about other cards.
A square wave can be written as a sum of sine waves:
Square(t) = sin(w*t) + 1/3*sin(3w*t) + 1/5*sin(5w*t) + ....
w = angular frequency = 2*pi*f, (f in Hz)
In other words, it's a wide-band signal. To reproduce a square wave with reasonable fidelity, you need a measurement bandwidth of at least ten times the fundamental frequency. Many sound cards have brick-wall filters that cut anything beyond 20 kHz off. Hence, the highest frequency of square wave they can be expected to sample accurately is about 1~2 kHz. And often their output bandwidth isn't much better...
~Tom
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
- Home
- Amplifiers
- Solid State
- Test Equipment