due to a problem within the power supply my ADAU1701 was exposed to a voltage of about 6 volts for about 1 minute.
The IC was pretty warm but still eays touchable. Is there a chance the IC has survived it?
There is a good chance that it did survive and will work properly but the part was obviously stressed and there may be some small damage that may cause an issue in the future. I am always surprised how much punishment some parts will take and still work. Yes, sometimes the slightest thing and the part goes bad.
So if you are using this as a test bed or a prototype that will stay in your office then you are probably fine and if any future abnormalities crop up then replace the part. If this is a customer unit or something that will be put into long term use then I would replace the part so I could trust its long term reliability.
Hello Dave T,
thank you for your quick reply.
I am using this IC for the purpose of prototyping. It's not for a customer. I am just practicing as a student.
So I think, I will give it a try!
I powered up my ADAU 1701. I can't communicate with it using SigmaStudio. On pins VOUT0, VOUT1, VOUT2 and VOUT3 are constant 1.5V. Is this ordinary?
I have the suspicion, the chip is dead...
It is normal. The DAC outputs are biased up at a common mode voltage of AVDD/2.
What is your hardware setup? Are you using the evaluation board or your own hardware?
I use my own hardware. As a template I used the FreeDSP design. I can program the EEPROM. The ADAU1701 doesn't start the self-boot sequence (After Reset, no communication on SCL and SDA lines).
So I assume you can read back the EEPROM and see that it contains the correct data?
If the EEPROM is not correct it will read in the first byte and then not do anything else. So since you say you see no activity at all on the SCL and SDA that it is not even trying. It is difficult to catch that first transmission, you have to set the scope on single trace to capture it.
So if there is no attempt then we have to look to make sure the DSP has what it needs to boot up a part.
Power and ground... of course but do not make assumptions that it is getting all the way to the part. If you can carefully place probes then you can even check right on the pins but you have to be super careful. Usually the bypass caps are good enough of a test.
Master Clock. Mare sure the MCLK jumper is correctly configured and check with a scope that there is clock on the MCLKIN pin. See that it is high enough in level and that there are no horrible reflections.
Make sure the reset pin is sitting high.
I see that the Selfboot pin is tied high, make sure it really is high.
Is there voltage on the CM pin? That would mean that there is some power on the part.
Are you able to load the program into the DSP directly from SigmaStudio and not use the Selfboot? You have to power it up and then wait a short time for the selfboot to complete ( or the attempt) then you should be able to program the part directly.
If you do not see any obvious problems to track down from all these tests I detailed, then I guess the part is bad.
I can read back the EEPROM. It contains:
01 00 05 00 08 1C 00 58 03 03 03 03 03 03 03 03 03 03 03 03 03 03 03 03 03 03 01 00 23 00 08 00
It looks like the EEPROM Data Example of the data sheet.
I checked power and ground directly on the pins... all normal.
The reset pin is high.
CM has a voltage of 1.52V.
I can't load the programm directly to the DSP. A USB communication failure accures...
I think the DSP is bad...
Thank you Dave!
It is looking like the part was damaged.
I measured the Masterclock at MCLKIN pin. There is no oscillation at all, just a steady dc at about 0.47V. and a bit noise.
Well I would chase that down! With no MCLK input the DSP will just sit there and do nothing.
I designed a new board with better decoupling. Now it seems to work fine.
Dave, thank you for your support!
A small question about FreeDSP is a design that is used as a prototype of that project. The BRD_RST signal is connected through a 470 ohm resistor. This signal can be a high level, even when power to the ADAU1701 is turned off. The voltage could be above the supply voltage of ADAU1701. But the current is limited by the resistor. Is this permissible? Can it be better to make BRD_RST signal open drain, as it is done for I2C signals?
Retrieving data ...