ADAU1701 works with active xtal?
Hi Guys,
1) I want to know if ADAU1701/1401 works with Active Xtal ( 4 Pins ones, with GND, Vcc and Out ), if yes,
a) Out is to be connected to MCLKI?
b) OSC0 is left floating?
2) Also, i want to know if accepts dual active xtal ( 11.289 and 12.288 ) with output switched by IC like CD4066? so that way board can operate at 44.1 Khz or 48 Khz selected by user.
Thanks.
Audio Signal Processors
adau1401
adau1701 adau1401 crystal 11.289 12.28812.288
SigmaDSP Audio Processors
adau1701
Hello Rocha,
Yes, you can use an active crystal or packaged oscillator. Connect it as you have described. Where I work we’ve built a mixer with six ADAU1701s all driven from the same 4-pin oscillator— we buffered its output, which you likely won’t need when driving just one chip.
Your other question is a bit more complex. When changing between 44 and 48K sample rates, a resistor associated with the -1701’s ADCs needs to change value (see the data sheet) although ignoring this may only cause a small change in ADC gain. There may be additional issues depending upon your application. What will the selectable sample rate be used for?
Best regards,
Bob
Hi Bob, thanks for anwser! About ADC resistor, i only know a resitor that can change input gain, i think gain cannot change with frequency, can u expalin a bit more please?
I have seen dataseheet and found the foloing:
21k8=(2+18)X(48000/44100) for 2v input
20k=(2+18)X(48000/48000) for 2v input
So there is 1k8 diference for 44.100 and 48.000, so this will cause an input voltage increase on 44.1, this will only affect gain?
Hello Rocha,
The ADCs have switched-capacitor inputs which sense the input charge (current x time) at each sample interval. At the lower 44.1K sample rate, the same current over the longer time builds up a higher charge with a corresponding larger effect. Re-reading the data sheet (p. 20), It appears that both the input and reference resistors need change with sample rate. Thus if these resistors aren’t adjusted for the 44K rate, charge will increase at all three associated pins. I honestly don’t know, then, whether a gain change or some other effect would occur — only that this resistor adjustment is in the data sheet for a reason. Perhaps a member who has worked with a dual-rate scheme, or an ADI support engineer, knows how important it is to adjust the resistors for what amounts to a < 10% rate change.
Just in case your MCLK rate switch is intended to adapt the -1701 to external off-the-board I2S sources — the -1701’s digital audio I/Os are intended to connect with other on-board chips all driven by the same MCLK. All must be exactly synched together or else you’ll hear artifacts as the different MCLKs slip cycles with each other. More advanced SigmaDSPs like the ADAU1452 include ASRCs (Asynchronous Sample Rate Converters) for handing audio from arbitrary sources.
Best regards,
Bob
Hello Rocha,
KJBob is correct with all his answers. There is a post that explains this a little bit more.
The confusion is that it seems like any change in sampling frequency will need a change in the ADC resistor but that is not true. It is a change in the ADC's sampling rate which is always at a constant rate divided down from the master clock. It is an oversampling ADC. The decimation down to the fs of the core is done with filters and it will decimate more or less depending on the sampling frequency settings.
So you only need to be concerned when the MCLK frequency is changed and you have done that calculation. Then if you change to 88.2kHZ the resistor will not need to be changed. Unfortunately, the formula was simplified by using the sampling rate rather than the oversampling rate derived from the output of the PLL. So this makes it somewhat confusing.
If you do not change the resistor value it will cause a small change in gain and the noise floor will also slightly change but it might not be a significant issue with your application.
Dave T
Please log in to reply.