By "noise" I guess you mean the that the D.C. output of the converter is stochastic in nature. For a periodic signal, the "ripple" on the output is a function of the "averaging" capacitor, and with this there is a trade-off settling time.
The Linear Tech development boards come with the post filtering option -- will take the two dev-boards I have and look at the output on the AP SYS2722 and get you a nV/Rt Hz figure but can't do it until Tuesday as I am babysitting my grand-daughter. I only have the less expensive ADI parts, not the high end AD636, 637.
Are the classic RMS converter chips AD637, AD736 and similar, good enough for a DIY distortion and noise audio analyzer which is in the THD+N class of a modern (analog only) Audio Precision System?
Do you think that it is better to go the digital way and use AD converters to do the RMS calculation in software?
E.g. the AD7982 and AD7960 (18 bit, 1/5 MSPS) have impressive specs...
Anyway the RMS values should be digitalized and go to the Laptop... so a dedicted AD converter and a precise voltage reference is needed anyway.
I don't have mutch trust in the integrated 12-14 bit AD converters of microcontrollers.
What are specific requiremets for precise RMS measurements of noise AND audio signals (bandwidth, crest factors, ease of implementation, any pitfalls....) and which approach is the best for a DIY system?
Thx - Udo
Do you think that it is better to go the digital way and use AD converters to do the RMS calculation in software?
E.g. the AD7982 and AD7960 (18 bit, 1/5 MSPS) have impressive specs...
Anyway the RMS values should be digitalized and go to the Laptop... so a dedicted AD converter and a precise voltage reference is needed anyway.
I don't have mutch trust in the integrated 12-14 bit AD converters of microcontrollers.
What are specific requiremets for precise RMS measurements of noise AND audio signals (bandwidth, crest factors, ease of implementation, any pitfalls....) and which approach is the best for a DIY system?
Thx - Udo
Original post only specified 100kHz bandwidth, so difficulties at high frequencies don't really come into play. Even an off-the-shelf soundcard with 192kHz sample rate would *almost* work and here's a 384kHz one for about $150: <http://www.amazon.com/dp/B00A2QL1CI/?tag=fox003-20> and <https://hifiduino.wordpress.com/2011/08/01/musiland-monitor-03-us/>. Would it meet the 2uv noise requirement? Be the first on your block to find out!
And who tested it ? And who is capable to perform any calibration verification ?
Sounds cards is very generic part, and even electrolytic capacitors over them eventually will suffer from aging and over all performance will degrade.
The world of measurements does not rely on soundcards.
Au contraire, DIY-ers rely on soundcards quite a lot! As for calibration, let necessity be the mother of invention.100 kHz really isn't very high. I imagine you could generate a waveform at increasing frequencies in that general range, then use diodes to clip to well-defined levels. The humble 1N914 is rated 4ns; very short in relation to 1/2 Hz. Observe the signal on a 'scope for additional confidence. Now try digitizing that clipped signal and observe what happens as frequency increases?
ASUS Essence STX soundcard test report via AP2722:
http://images10.newegg.com/UploadFi...ssence_STX_Audio_Test_Report1403939991275.pdf
Ain't that bad ...
http://images10.newegg.com/UploadFi...ssence_STX_Audio_Test_Report1403939991275.pdf
Ain't that bad ...
Neither my Creative X-Fi music is that bad, but eventually many people got the soldering irons at hand, so to replace chips and capacitors, with the false hope to make them to sound better.
If you need to verify op-amp true performance and capacitors performance, you need production level test and measurement tools.
If you need to verify op-amp true performance and capacitors performance, you need production level test and measurement tools.
Among other things, a UT-61 DVM. It easily has 20 KHz bandwidth, reads true RMS volts, and its open circuit noise is about 1 mv. The last number is either a powerful dissuader or persuader depending on your needs. It is about $50 with a USB interface and other goodies. eBay or Amazon.
You better stay away of UNI-T until 2025, until then there is a possibility for them to learn of how to make measuring instruments.
At that price you get 3 kilos of peanuts, which is much better to prefer those as eatable goodies.
Not a fan of theirs, I take it? Get burned by them? BTW as you no doubt know with meters, getting burned by cheap products can be more than just a metaphor if you work on some serious voltages like > 1 Kv.
Even just > 250v. Of course that is pretty rare in audio unless you work with electrostatic speakers or tubes and the like. Been there, done that but not for several decades.Most of their line is pretty average for low cost pacific rim product, but the UT-61 is kind of special, almost like they blundered into it. But its only safe up to 250 v so I can use it on my house wiring, but it would be a problem if I was still working on big tube amps. If I was in the Eu some of the house wiring could be a problem as they have 400+ roaming around a lot of houses. Not here, 230 is about it.
At that price you get 3 kilos of peanuts, which is much better to prefer those as eatable goodies.
Those seem like some expensive darn peanuts! Around here 7 pounds of really pretty good shelled roasted peanuts are like about $25, and probably less if bought in bulk instead of 12 oz jars.
;-)
Last edited:
Right now in our times there are about seven multimeter chip developers, and another few R&D teams which combining ready chips and additional electronics, so a multimeter to become able to measure other than Law of ohm, and half measurements than an Oscilloscope can do.
For complex measurements there is need for higher CPU power.
UNI-T is in denial to use first class technology, today only 20% of users need basic Law of Ohm measurements.
Multimeter safety and compliance to latest standards is another topic, which UNI-T also shown denial to comply.
In conclusion they are dishonest to consumers, and they do that on purpose.
What seems as justice is consumers to turn their back and to them and never look back.
For complex measurements there is need for higher CPU power.
UNI-T is in denial to use first class technology, today only 20% of users need basic Law of Ohm measurements.
Multimeter safety and compliance to latest standards is another topic, which UNI-T also shown denial to comply.
In conclusion they are dishonest to consumers, and they do that on purpose.
What seems as justice is consumers to turn their back and to them and never look back.
It seems that you have a vision for a relatively inexpensive handheld device that replaces the DVM as we know it today, but includes more of the functions of more advanced instruments such as an oscilloscope and a FFT analyzer.
You seem to know something about Uni-T's management and you think they don't share your vision?
Agreed.
I agree that the market will benefit from these more advanced test tools that you envision, and that companies that don't remain competitive should loose their market share.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.