ΣSigmaStudio Ask AI Русский Log in

Is it impossible to set Fs=192kHz for EVAL-ADAU1452MINIZ?

SigmaDSP & SigmaStudio kegasa2014 Apr 21, 2014 00:00 12 replies Original on forum ↗
Subscribe
kegasa2014 Question
Apr 21, 2014 00:00

In EVAL-ADAU1452MINIZ User Guide UG-636, I found the line below.

" The AD1938 is configured to run at a sample rate of 44.1 kHz

or 48 kHz. It is not possible to change this setting. Even though

the ADAU1452 is very flexible and can run at any sample rate

up to 192 kHz, the analog audio inputs and outputs on the

EVAL-ADAU1452MINIZ may be distorted or silent if a sample

rate other than 44.1 kHz or 48 kHz is used for the ADAU1452

serial ports."

Is it impossible to set Fs to 192kHz when using EVAL-ADAU1452MINIZ with SigmaStudio 10.3.10 ?

Is there any easy way to wrok around this limitation?

I want to use EVAL-ADAU1452MINIZ to analyze 40kHz signal.

http://www.analog.com/static/imported-files/user_guides/EVAL-ADAU1452MINIZ_User_Guide.pdf

Audio Signal Processors

adau1452

ad1938

Integrated/Special Purpose A/D Converters

SigmaDSP Audio Processors

Audio CODECs

Show More

DaveThib Reply
Apr 21, 2014 00:00

Hello kegasa2014,

Is it impossible? Well,... no, it is not impossible. Is it easy? No, it is not easy. The AD1938 is setup to run in standalone mode. In this mode the only sample rates possible are in the 32kHz/44.1kHz/48kHz range. To run the part at an fs of 192kHz you would have to run it in "normal" mode with control of the part via the SPI port. To do this you would have to cut the ground traces going to pins 23,24,26 and 27 and run these to the USBI SPI ports and to its own SPI latch signal. There are some details to deal with that both in hardware and in SigmaStudio but it can be done. You would also have to ensure that there is a pull-up resistor on the CLATCH line.

So it is possible but it is not easy and requires some very fine soldering detail work to be able to cut those traces and solder wires to the pins of the part. The wires will have to be 30ga or less. The traces from the pins to the ground plane vias are very short so cutting the traces will be difficult.

If you choose to go this route then we can help you further with some more details.

Yogami might chime in with a detail about the ADAU1452 setup on this board, but I am pretty sure it will not pose a problem.

Thanks,

Dave T.

kegasa2014 Reply
Apr 22, 2014 00:00

Would you please tell me, which register I should write to, and what I should write, to change the sampling frequency from the 48kHz to the 192ksps? I checked the file - "AD1938.pdf" but I couldn't find the default setting after power ON. Or should I set all the 17 registers after power ON? It will be very kind of you, if you could supply me the 17*8bit data I should set in the registers!

kegasa2014 Reply
Apr 22, 2014 00:00

Thank you very much for the answer, Mr. Dave Thib!

I think I can handle the hardware all right.

As a matter of fact, I will make my own board with several amplifiers, if my software ( or block diagram ) works fine on EVAL-ADAU1452MINIZ. I will be able to use microscope when I will cut the traces and soldering the wires. I will use my favorite one chip microcontroller to send the data.

Would you please tell me, which register I should write to, and what I should write, to change the sampling frequency from the 48kHz to the 192ksps? I checked the file - "AD1938.pdf" but I couldn't find the default setting after power ON. Or should I set all the 17 registers after power ON? It will be very kind of you, if you could supply me the 17*8bit data I should set in the registers!

DaveThib Reply
Apr 22, 2014 00:00

Hello kegasa2014,

The note regarding the default register values is cleverly hiding in plain view. (I know, it is too obvious.) Right in the beginning of the control registers section of the datasheet. Page 24. I will paste it in here:

"The global address for the AD1938 is 0x04, shifted left one bit due to the R/W bit. All registers are reset to 0, except for the DAC volume registers that are set to full volume.

Note that the first setting in each control register parameter is the default setting."

Now for the register settings, it will depend on what mode you want to run in? Will you only need two channels or four? So will you use I2S or TDM?

Will the AD1938 be a slave?

So with those assumption, I2S and slave with MCLK being supplied at a frequency of 12.288 MHz.

You will have to make the following writes:

PLL Clock Control Register 0, write 0x80, to enable the ADC and DAC

DAC Control Register 0, write 0x04, DAC to 192kHz fs

ADC Control Register 0, write 0x80, ADC to 192kHz fs

That should do it.

When programming your microcontroller to talk to this part. Make sure the CLATCH line remains high until you need to send a message. If possible, you should also leave the CCLK line high until just before you send a message. Then do not interrupt the communications for too long. You should have a pullup on the CLATCH line.

Let me know how the soldering went. You will need a microscope so it is good that you have one. Be careful not to short the wires to the ground via that is very close to the pins.

kegasa2014 Reply
Apr 24, 2014 00:00

Thank you very much for the answer, Mr. Dave Thib!

So, Global Address 0x04 shift 1 bit left + write '0' or read '1'.

For write, 0x08 is the global address.

So, 24bits serial output data for each Registers are,

0x080080 = address 00 PLL Clock Control Register 0, write 0x80, to enable the ADC and DAC

0x080204 = address 02 DAC Control Register 0, write 0x04, DAC to 192kHz fs

0x080E80 = address 0E ADC Control Register 0, write 0x80, ADC to 192kHz fs

Am I right?

I'll write the program to send these data (from MSB to LSB) to the AD1938 tomorrow.

Thank you very much again for your help!

DaveThib Reply
Apr 24, 2014 00:00

That looks good.

kegasa2014 Reply
Apr 25, 2014 00:00

Thank you very much for checking my data, Mr. Dave Thib!

I made the program to send the data on the one-chip-microcontroller all right.

I'll connect the programmed microcontroller to the AD1938 in the evening.

I hope it will work all right!

Thank you very much again for your help!

kegasa2014 Reply
May 8, 2014 00:00

It is possible to set 4*Fs=192kHz for EVAL-ADAU1452MINIZ !

Thank you very much for the answers, Mr. Dave Thib!

Now I can output 40kHz signal from EVAL-ADAU1452MINIZ all right.

I connected one chip microcontroller to AD1938's CDATA(CIN), CCLK, CLATCH(with 30k ohm pullup).

I set the 3 registers(each 24bit).

I copy-pasted some part of my microcontroller C source code below.

CLK0 = 0x80;

DAC0 = 0x04;

ADC0 = 0x80;

latch_low();

send8bit( 0x08 ); // global address 0x08 = 0x04 shift 1 bit left + write mode 0

send8bit( 0x00 ); // address 00 = PLL Clock Control Register 0

send8bit( CLK0 ); // data 0x80 = enable the ADC and the DAC

latch_high();

latch_low();

send8bit( 0x08 ); // global address 0x08 = 0x04 shift 1 bit left + write mode 0

send8bit( 0x02 ); // address 0x02 = DAC Control Register 0

send8bit( DAC0 ); // data 0x04 = DAC sampling frequency = 192kHz

latch_high();

latch_low();

send8bit( 0x08 ); // global address 0x08 = 0x04 shift 1 bit left + write mode 0

send8bit( 0x0E ); // address 0x0E = ADC Control Register 0

send8bit( ADC0 ); // data 0x80 = ADC sampling frequency = 192kHz

latch_high();

I changed the SigmaStudio's Hardware Configuration of IC1-ADAU1452 on tag CORE_CONTROL and SERIAL_PORT.

I changed CORE_CONTROL tag's Start Pulse Selection to Base_Fs*4(192kHz for 48kHz base sample rate).

I also changed SERIAL_PORT tag's SDATA_OUT0 Sampling Rate to 4*Fs.

Then I set Tone Generator's frequency to 40kHz and output to 0.

I monitor the signal using Oscilloscope and found I got 40kHz 260mVp-p sine wave all right.

Thank you very much again for your help.

kegasa2014 Reply
May 9, 2014 00:00

My EVAL-ADAU1452MINZ with one chip microcontroller and the waveform (40kHz output).

DaveThib Reply
May 9, 2014 00:00

Looks great. I like the very creative modification! Gets the job done.

francescodenetto Reply
Dec 25, 2015 00:00

So it's possible to change sampling rate frequency from 48khz to 192khz?

DaveThib Reply
Jan 1, 2016 00:00

Hello Francesco,

This is a good thread to study. You will see that it can be done but it is not easy. Mr Koyo, had a microscope to use to do the trace cuts and soldering. Without one it is risky. It is easy to end up with a short.

Dave T