PCM1794 oversampling rate question

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
You can just turn off the digital filter as well


You can not turn off the oversampling / interpolation

Is FIR filtering called interpolation as well? A very high cross-talk in terms

http://www.diyaudio.com/forums/digi...-192-24-dac-pcm1794-waveio-usb-input-259.html
That is some ironic Kastor L

The WM8741 when is modified in the 8FS Mode (MODE8X) allows the use of custom digital filters by bypassing the internal digital filters. When MODE8X is set, the PCM data input to the WM8741 is applied only to the digital volume control and then the analogue section of the DAC system, bypassing the digital filters.

In PCM1794 only the front end 8X Oversampling digital filter is bypassed in DFTH = Digital Filter Bypass (or Through Mode).

Do you think that WM8741 is more appropriate for use in a NOS DAC than PCM1794?
 
Do you think that WM8741 is more appropriate for use in a NOS DAC than PCM1794?

Or - using more acceptable terminology from ... - WM8741 is more appropriate for use in a NDF DAC (No Digital Filter DAC) than PCM1794?

As is stated in WM8741 datasheet, when the chip is configured in MODE8X not only the PCM digital filter section is bypassed, it is also bypassed the section of the two Σ-Δ Modulators, i.e. clean sweep.
 
Last edited:
I need to study more, but I think 2x PCM1792 in dual-mono might be most appropriate for a true Nos DAC, not NDF.
Allow me to disagree. I came across the datasheets of both PCM1792a and WM8741 again and again. There is no case to bypass the Σ-Δ modulator of PCM1792a and thus the digital signal will always be under oversampling process.
I saw the scope screenshots of dddac NOS DAC, indeed the ringing of square wave in higher frequencies is negligible but this does not say anything.
Square wave is a non continual function while a Σ-Δ modulator uses sinusoidal functions (see Fourier equations) to approach its reconstruction in analog form. As the terms of the Fourier equation are increased - by using higher oversampling rates - the ringing are accumulated in the corner of the rising edge of square wave (that is the named Gibbs phenomenon). I think that using a well calculated compensation capacitor across the feedback loop of the operational amplifier used either in the I/V conversion section or in the low pass filtering section you could completely eliminate the ringing residues without reducing the rise time. I tried in simulator the suggested by TI I/V circuit (around the NE5534 which off pins 8 and 5 are exactly offered for compensation) and i got similar waveforms like those of dddac. By any way, these are my personal thoughts and estimations and maybe are wrong.
Going further, you talk about dual PCM1792. I can interpret it only as that you are looking for the higher possibly dynamic range. Actually the referred DNR in PCM1792 - 1794 datasheet can be interpreted in a "misleading" way. Maybe you should take into account that the PCM1792 DNR of 127dB (2Vrms out / 44.1 KHz Fs / A-weighted) is measured before the external I/V conversion stage, whilst the WM8741 DNR of 125dB (2Vrms out / 48 KHz Fs / A-weighted) is measured after its internal I/V converter. I do believe that if the DNR of PCM1792 will be measured after the external I/V stage will be the same like WM8741. A very good example is the sound card of ASUS Xonar Essence STX, which uses the PCM1794 and the measured DNR in its output is 123 - 124dB.
 
How about daisy-chain four PCM1794 in true Nos mode? To achieve 24-bit performance, 6-bit thermometer x4 = 24.

Haha, just an idea, perhaps not realistic.
You probably mean that you could find a way to bypass the Σ-Δ modulator and to use only the ICOB decoder output?
Unfortunatelly there are not resources (still in TI website) to explain in detail the architecture of this kind (Advanced Segment) of DAC to i help you.
 
That's okay, wish you luck in your project.
Thank you so much for the wish, you are a true gentleman.
You know that i use WM8805 + WM8741. So far, i used only WM8805 (SPDIF receiver) in software mode, i.e. managed by microcontroller as i have 7 SPDIF inputs in my project. I had configured WM8741 DAC to work in hardware mode for simplicity, but after reading all those about PCM filters offered i decided to also use it in software mode. For all sampling rates, are offered either "Minimum Phase apodising" and "Linear Phase apodising" filters. I will arange them in two main categories selectable by the user, to give him a chance to try different audio tastes : Minimum and Linear phase. For the rest i.e. the correct OSR for each sampling rate will take care the microcontroller through the program code.
I hope you will like this modification.
Regarding a true NOS DAC, when i wll finish my current project, i promise you that will try an implementation in colaboration with you offering my knowledge about microcontrollers.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.