PCM5142: Device is not responding to I2C

Robert Decarreau

Part Number: PCM5142

I have a PCM5142 connected to a stm32 via I2C, I have the MODE1 tied low and MODE2 tied high. When I attempt to read a register by first writing the 7-bit address, MSB first, followed by a 0 for write. I get nothing in response. I have a logic analyzer on the way to further ensure my data communication, but I have used this stm32 to send other I2C data successfully.

I am concerned that the pcm5142 might be in standby or powerdown and wont respond to I2C messages. Can you elaborate on the Power Down and Standby modes' response to I2C messages?
Do I need a BCK or SCK in order to keep the device awake, as hinted in section 10.5.2?
If I tie BCK or SCK high in lieu of having a clock should I at least get something?

over 8 years ago

0 Diego Melendez over 8 years ago

TI__Guru 57800 points

Hi, Robert,

Welcome to E2E, Thanks for your interest in our products!.

The PCM5142 requires the clocks to be operational and valid to start normal operation. The device has a power on-reset function where the part will remain in RESET state until power supplies are stable and valid clocks are provided to the device.

Regards,

-Diego Meléndez López
Audio Applications Engineer

0 Robert Decarreau over 8 years ago in reply to Diego Melendez

Prodigy 110 points

Thanks Diego. Is that to say that it needs both SCK and BCK? I am also guessing that the device will not respond to I2C when in RESET mode?

0 Diego Melendez over 8 years ago in reply to Robert Decarreau

TI__Guru 57800 points

Hi, Robert,

That is correct, clocks are required by the device to enter normal operation mode (and establish I²C communication). Even if SCK is not provided, the part can work in 3-wire mode as the device will start the power-up sequence with valid BCK and LRCK.

Best Regards,

-Diego Meléndez López
Audio Applications Engineer

0 Robert Decarreau over 8 years ago in reply to Diego Melendez

Prodigy 110 points

I applied SCK and power while monitoring VNEG. When I apply power, VNEG never turns on, in fact it jumps to about 930mV when power is applied.

It seems odd that LRCK and BCK are required to communicate via I2C.

I hate to ask so many clock questions, maybe it is my misinterpretation of the datasheet. Is there any way to communicate with the device while applying only SCK and power?

0 Robert Decarreau over 8 years ago in reply to Robert Decarreau

Prodigy 110 points

I connected another cmos clock at 24.576MHz and it seems to be working fine. I had some weirdness with the I2C address. Now though, the device is responding twice to the register read. I have a logic analyzer trace of the I2C traffic, I will try to load it on here.

0 Robert Decarreau over 8 years ago in reply to Robert Decarreau

Prodigy 110 points

(I have taken this posting one step further, should I make another one?) This is the logic trace. You can see the setup write to the device at address 0x98, then the register request 0x3E, a read request to the same device, a proper response 0x30+ACK and then another 0x30 response... This is NAK'd. I think that is because the Master doesn't know what to do about that data.

Any thoughts?

0 Robert Decarreau over 8 years ago in reply to Robert Decarreau

Prodigy 110 points

I continued this in another thread.
e2e.ti.com/.../581299

Accepted the above answer. Thanks for the help!

Audio

Audio forum

PCM5142: Device is not responding to I2C