AD73311L: Cascade running at different sampling speeds

Q We want to use two 73311L's in cascade operation but with different sampling rates.(to be specific 8 kHz and 48 kHz) Could you please inform us whether this operation possible? A The short answer is No, here’s why. Each device in the cascade must know how many devices are in the total cascade, in this case 2. In each sample interval each device will accept a total number of data words equal to the number of devices in cascade. When a device receives the last word (i.e. the number of words equal to the cascade length) it ignores all other words. In normal operation both devices in the cascade have a sample ready at the same time. The first device in the cascade passes it ADC word to the second device, and the second device passes it ADC word to the DSP. At this time Device 1 has received no words and Device 1 has received 1 word. Because the total length of the cascade is 2, Device 2 knows that it should receive more than 1 word, so it passes this word (the ADC word from Device 1) to the DSP. The DSP may also write DAC words to the cascade. The first DAC word is written to Device 1. At this time Device 1 has received 1 word, but like Device 2 it knows that the total cascade length is 2, so it passes this word to Device 2. Now Device 2 has received 2 words and keeps this word as a DAC word. The DSP then transmits a second word to Device 1. Device 1 has now received 2 words, so it keeps this word as a DAC word. If the DSP sends any more words to Device 1, they will be ignored because Device 1 has received all the words it wants in that sample interval. In the suggestion above, if Device 1 is running at a higher sample rate than Device 2, in one sample period of Device 2, Device 1 will transmit 2 ADC words to Device 2. Device 2 will think that the second ADC word from Device 1 is actually it's DAC word, and will not pass this word on to the DSP. Similarly, if it is Device 2 that is running faster, the DSP will not be able to write multiple ADC words to Device 2 in one sample interval of Device 1. Also, you can't achieve 8k and 48k sampling from the same DMCLK. All devices in cascade have to have the same DMCLK. Sample rate options are DMCLK/2048, DMCLK/1024, DMCLK/512 and DMCLK/256.

AD73311: Power supply sequencing problem, using DVDD = 3.3V and AVDD = 5.2V

Q We use two different voltages (digital 3.3V and analog 5.2V) to power the codec. With our onboard power supply we have no problems but during the production test we noticed a problem with the power-on-sequence. We have established that the codec doesn't work if between power-on 5.2V and power-on 3.3V elapsed a few ms. The problem only occured if the 5.2V comes first. Now our questions: What is the reason for this effect and how many time is allowed between power-on 5.2V and 3.3V. A There are two possible causes of the failures that I can see. It could be power supply sequencing. We actually recommend that AVDD and DVDD are powered up together. On power up AVDD should follow DVDD until DVDD reaches it's final value of 3.3V and then AVDD should then continue to it's final value of 5.2V. The second possibility is that something else happens in the intervening few milliseconds between powering 5.2V and powering 3.3V. Is it possible that the DSP or micro attempts to communicate with the AD73311 before DVDD has been established? It's critical that the absolute maximum ratings are respected for the digital inputs and remain between -0.3V and DVDD +0.3V at all times. Check that there is no high level on SDI, SDIFS and EN pins before DVDD is established. The final chapter in all our seminar books is dedicated to hardware design techniques and deals with such issues as grounding, decoupling, parasitic thermocouples and good PCB design. http://www.analog.com/support/standard_linear/seminar_material/index.html

AD73311: Power supply

Q May I use AD73311 with different DVDD and AVDD (DVDD = 3.3V and AVDD = 5V) ? A Yes you can use DVDD=3.3V and AVDD=5V. We recommend that AVDD and DVDD are powered up together. On power up AVDD should follow DVDD until DVDD reaches it's final value of 3.3V and then AVDD should then continue to it's final value of 5V. As always the absolute maximum ratings provided in the datasheet should be respected.