We develop an active acoustics, where we use ADAU1702. When power is turned on/off, at the output circuit (DAC) is formed emissions. Tell me please, do not come across anyone with a similar problem? How can I remove these emissions?
What do you mean by the word "emissions"? Are you referring to the spike that appears on the power-up and power-down diagrams?
Yes, I was referring to these spikes
What is the method for turning power on and off? Are you disabling the power supply somehow? Is there any kind of spike on the power supply when you do this?
We feed the chip from a regulated laboratory power source through a voltage regulator for 3.3V. Power source and voltage regulator don't allow spikes. We tried to gently apply the voltage to the chip (from 0 to 3.3V), and this is what we saw:
Could you check on your board is the appearance of these spikes on the ADAU DACs?
We're going to start mass production of circuit boards, whose composition will include adau1702. But we can not start production, yet we don't find a solution for this problem.
should I wait for a response? time is running out
I was out of the office Wednesday an Thursday. Today I am working remotely. I do not have evaluation hardware here with me, so I cannot perform any tests of my own, but I have a few ideas for you.
1. are you muting the DACs via register write prior to shutting down? That might help.
2. have you tried increasing the amount of capacitance on the power supply line? That will make the power up and power down curves much slower so that the common mode voltage stays constant longer and the noise is smaller in relation to CM.
3. Do you have the possibility of muting the power amplifier prior to removing power from the SigmaDSP?
4. What kind of DAC filter topology are you using? Active? Passive?
5. Do to see the same spikes on the evaluation board? I don't have a board with me today so I can't check until Monday.
Brett, all your recommendations have been taken into account by us, but the spikes remain. We are using an active DAC filters. On the evaluation board we are seeing the same spikes. We haven't possibility of muting the power amplifier prior to removing power from the SigmaDSP.
I also confirmed these spikes on the evaluation board. When power is suddenly removed from the device, the digital engine goes into an "unknown" state because there is not enough power for them. In other words, bits may start randomly toggling for a short period of time because the power supply to them is unstable. There will also be some discharge from capacitive loads from when the system is powered down until it reaches equilibrium. So, unfortunately, I do not think that it is possible to avoid these spikes. My only suggestion, if you can't mute the power amplifier, would be to maybe create some kind of analog switch that would be opened prior to powering down the SigmaDSP, thus guaranteeing that the spikes do not reach the power amp.
Retrieving data ...