Linearity is very strange. :-SAdditional measurementsMeasurements & Review E1DA Cosmos ADC - L7Audiolab
台式设备 前几天NGA著名科Hi大佬“黑人哥”给我发了一个设备,说是他关注了很久的 E1DA Cosmos ADC ADC? 本站之前还没测过这样的设备……但是测量之后,我决定给它加急!这可是难得一见的廉价的同时有着极高性能的AD转换设备啊…… 拥有无限可能……www.l7audiolab.com
I am also interested in I2S output, not just USB from the Cosmos unit. That would really of interest to those of us that want to use the output digitally, either as I2s or convert to S/PDIF, AES/EBU, TOSLINK, etc...Hey IVX,
I've tried sending an email to order as per your instructions but haven't received a reply. Is there another way?
(Have checked spam etc).
Do the AD you are selling now have the I2S output?
Cheers,
Ruairi
Ivan measured linearity as well and got this:Linearity is very strange. :-S
Low frequency distortion as well (I wonder how many harmonics @WolfX-700 is allowing in his THD vs frequency measurement)
Yes, I also think there is probably a mistake in @WolfX-700 measurements here.Ivan measured linearity as well and got this:
View attachment 157748
Perhaps WolfX-700's measurement was limited to 16bit.
As for the distortion peak at 40Hz, WolfX-700 did another 40Hz FFT and got this (not published for whatever reason):
View attachment 157749
114dB here vs 94 in the sweep that's part of the review.
Well, even for measurements it's interesting.I am also interested in I2S output, not just USB from the Cosmos unit. That would really of interest to those of us that want to use the output digitally, either as I2s or convert to S/PDIF, AES/EBU, TOSLINK, etc...
Yes, I also think there is probably a mistake in @WolfX-700 measurements here.
This kind of linearity for a device that is said to have 128dBA SNR makes little sense.
I think @amirm is using a very narrow band around 200Hz.I will update the narrowband linearity measurement (800-1200HZ)
I don't do any standard linearity measurement. I use a quarter sample rate tone (11,025 hz for 44.1). Any harmonics are out of band. I use a signal that only has one bit level turned on at a given time. Pretty much any sigma delta DAC has clean linearity to 20 or more bits. Those with low enough noise I can confirm as having the proper level down to 23 bits and pretty much 24 bits. I believe even the noisy ones are getting linearity correct all the way down, they just get swamped by noise at the lowest 3 or 4 bits.I think @amirm is using a very narrow band around 200Hz.
I wonder if that's not just one FFT window bin.
@WolfX-700 And how many harmonics do you allow in THD vs Frequency ?
You can chose in the AP settings for THD measurement.
I also wonder if and where @amirm sets a limit for that test.
My chart is THD+N frequency sweep and the bandwidth range is 20-20K. Not THD frequency sweepI think @amirm is using a very narrow band around 200Hz.
I wonder if that's not just one FFT window bin.
@WolfX-700 And how many harmonics do you allow in THD vs Frequency ?
You can chose in the AP settings for THD measurement.
I also wonder if and where @amirm sets a limit for that test.
Well, if you add some compression, I'm not sure you would see it in THD.that is simply impossible even theoretically if the linearity is 16bit but the distortions harmonics level -140db.
The linearity test to me is redundant because it duplicates THD and is less precise. Any nonlinearities will be detected by the THD 100.0% unavoidable.
If the sine remains the same but only its level changes "dynamically" (with input level), a static sine signal will still be undistorted.Rja4000, THD is the sum of harmonics powers. Any nonlinearity will add harmonics to the sin(x) function stimulus. Why do we need to test the linearity one more time after the THD test?
That's not what linearity test is.Rja4000, THD is the sum of harmonics powers. Any nonlinearity will add harmonics to the sin(x) function stimulus. Why do we need to test the linearity one more time after the THD test?