AU1450 Crashing with SPI??

billlance Question

Jul 28, 2016 00:00

Hello,

My AU1450 is working great EXCEPT that it crashes once in a while (every 5 to 10 minutes, maybe) when I'm writing control coefficients via the SPI slave port. The SPI control works; I can load the code, and control the variables, it's just that the thing crashes sometimes. When it crashes, the CLK OUT goes from 12.288 MHz to half that (6.1MHz), and all audio I/O stops. If I don't write any SPI stuff (other than loading the code and starting it), it will run forever. I've looked very carefully at the SPI lines for noise, timing, etc. and everything looks fine.

Thanks in advance for any help!

Report

ADIApproved Reply

Jul 28, 2016 00:00

Hi billance:

Can you share the code that you are trying to execute, perhaps form a SigmaStudio project?

Also, can you better describe what you mean by "Crash"

Thanks a lot

Miguel

Report

billlance Reply

Jul 28, 2016 00:00

Hi Miguel,

Thanks for responding. The dsp code that's executing was created in SigmaStudio, but incorporated into assembly-language code that's running on a Freescale micro. I don't think there's any problem with that code, as it all works fine and does what it's supposed to do.

The micro code that writes the SPI control stuff also seems fine, and does what it's supposed to do, so I don't think there's a code problem, I think there might be a hardware problem of some sort. When I let the micro write the SPI control (a block of about 40 coefficients (about 175 bytes or so) every 30 milliseconds or so, it causes the 1450 to intermittently crash... not right away, but maybe after 5 or 10 minutes of operation - kind of randomly.

By crash, I mean that somehow I think registers in the 1450 are getting corrupted... the clock out goes from 12.288 mhz to 6.1 mhz, and the audio outputs quit.

This morning I tried connecting the micro's spi to the 1452 on an AD eval board, and it hasn't crashed yet. This has me thinking it's some kind of hardware problem on my board, I just can't figure out what it is. The chip is well-bypassed, the pcb is a four-layer board with good grounds and power plane. I've looked at all the SPI lines, and they look clean, and the timing is good.

It's got me puzzled, but I think the software part is fine.

Thanks.

Report