DAC, ADC self calibration is a big subject if you read the right journals, has been implemented on commercial chips for decades, production laser trimming for even longer
external SW calibration has literature too
an example: http://class.ece.iastate.edu/vlsi2/docs/papers done/2005-11-itc-kp.pdf
external SW calibration has literature too
an example: http://class.ece.iastate.edu/vlsi2/docs/papers done/2005-11-itc-kp.pdf
Last edited:
Interesting. This is the core ADC: http://www.analog.com/static/imported-files/data_sheets/AD7767.pdf They use an HP 3458 to verify the accuracy of the digitizer. The other interesting note was the " Ultra‐pure sinewave gen.(Spurs <‐180 dB)locked to
a quartztime base was developed to investigate ADCs.". That looks like the holy grail. Sine its a government sponsored effort perhaps they published the details?
Another ADC I find very interesting: LTC2378-20 - 20-Bit, 1Msps, Low Power SAR ADC with 0.5ppm INL - Linear Technology . While ONLY 20 bits it has a claimed -120 dB THD and samples at 1 MHz. I will contact Linear Tech to confirm some details on the demo board ($300 w/ USB adapter) before I spend some money on it but it looks really interesting. SAR is an ADC architecture I have not seen in audio. Its common in precision DVM's. Some smart guy may be able to figure out how to turn it into a 250 KHz 24 bit ADC. Meanwhile I think it could make a great analyzer.
a quartztime base was developed to investigate ADCs.". That looks like the holy grail. Sine its a government sponsored effort perhaps they published the details?
Another ADC I find very interesting: LTC2378-20 - 20-Bit, 1Msps, Low Power SAR ADC with 0.5ppm INL - Linear Technology . While ONLY 20 bits it has a claimed -120 dB THD and samples at 1 MHz. I will contact Linear Tech to confirm some details on the demo board ($300 w/ USB adapter) before I spend some money on it but it looks really interesting. SAR is an ADC architecture I have not seen in audio. Its common in precision DVM's. Some smart guy may be able to figure out how to turn it into a 250 KHz 24 bit ADC. Meanwhile I think it could make a great analyzer.
spurs are counted separately from harmonics - so the THD may be lots worse
the LTC2378-20 THD falls apart at only slightly higher than the 2 kHz test frequency, would need to be highly paralleled to cut noise to commensurate levels unless you go for long averaging
in the best of all possible worlds paralleling could reduce distortion by similar factor as the noise
the LTC2378-20 THD falls apart at only slightly higher than the 2 kHz test frequency, would need to be highly paralleled to cut noise to commensurate levels unless you go for long averaging
in the best of all possible worlds paralleling could reduce distortion by similar factor as the noise
Last edited:
That must be why they only show testing at 2 KHz. You have tested the device? Just saved me some time and money.
SAR's need de-glitching or really good sample and holds, I know Samuel might have something share in the future. SAR DAC's failed miserably for audio in the past.
what causes the deviation to rise with increasing frequency? what is the main mechenism here? Can it be easily corrected?
Does it and if so-how does it contribute to THD?
[this is from the same Euramet paper... an ADC running at 24bit/128khz)
Thx-RNMarsh
Does it and if so-how does it contribute to THD?
[this is from the same Euramet paper... an ADC running at 24bit/128khz)
Thx-RNMarsh
Last edited:
What do you mean by SAR DAC? Do you mean R2R DACs or the internal conversion capacitive DAC of a SAR.
Sorry I mis-spoke we had a DAC that looked like a Kelvin-Varley divider with CMOS switches where the transitions needed de-glitching.
1-4 MB/s sampling rate for ADC --
Looks like I should buy the ADS1675 REF board ($500) and compare its results on harmonics with the others.... using 1+ MHz rate.
... and see where it deviates from the other analyzers here on harmonic levels .... the board includes FFT.
Is there another as good or better before taking the plunge?
Thx-RNMarsh
Looks like I should buy the ADS1675 REF board ($500) and compare its results on harmonics with the others.... using 1+ MHz rate.
... and see where it deviates from the other analyzers here on harmonic levels .... the board includes FFT.
Is there another as good or better before taking the plunge?
Thx-RNMarsh
Last edited: