CDCM6208: Problems with I2C communication

Joshua Metz

Part Number: CDCM6208

Hello TI-support team,

we are using the CDCM620V2 and have a strange behavior with I2C communication.

Means that the CDCM6208 doesn't send an ackknowledge with the first try of read_id after power up in one of our products.

If we send a second read_id directly after the first one than everything is fine with the communication and we can write all registers.

In the past we had kind of the same problems in other products with your CDCM6208 means that some times the ackknowledge is missing while writing the registers.

Lower pullup resistance made the problem better but we are thinking that the CDCM6208 is a little tricky with I2C.

Thank you very much for your help.

Best regards

J. Metz

7 months ago

0 Kadeem Samuel 7 months ago

TI__Genius 15150 points

Joshua,

Are the I2C timing requirements being met? I see a similar thread where this may be an issue: https://e2e.ti.com/support/clock-timing-group/clock-timing-internal/f/clock-timing---internal-forum/1161038/cdcm6208-i2c-ack-issue

See the timing requirements here:

Any noise on the VDD pullup voltage may cause an issue as well: https://e2e.ti.com/support/clock-timing-group/clock-and-timing/f/clock-timing-forum/571482/cdcm6208-i2c-communication-problem?tisearch=e2e-sitesearch&keymatch=cdcm6208%25252525252525252520i2c#

Issues with setup and hold timing can cause the device to see a Start or Stop condition: e2e.ti.com/.../cdcm6208-i2c-latchup-aviat

Thanks,

Kadeem

0 Joshua Metz 7 months ago in reply to Kadeem Samuel

Prodigy 40 points

Hello Kadeem,

the timing requirements where the first thing we measured and we have a rise time of aorund 100 ns from 10% to 90%, so this couldn't be the issue.
We also measured all the supply voltages while i2c communication and everything looks fine.

Interesting is that if we do any i2c communication, for example reading an other i2c address, and then send the read_id command for the CDCM6208 than it's working without any problems. Or if send the read_id for the CDCM6208 twice, after the second i2c command the clock chip is working fine.

I hope that TI could know anything about the internal chip design which cause this strange behaviour.

Best regards
Joshua

0 Kadeem Samuel 7 months ago in reply to Joshua Metz

TI__Genius 15150 points

Joshua,

The design repository that contains the models for this part is out of service for the next two weeks. When it is online again, we can look into the design of the I2C block.
Thanks,

Kadeem

Clock & timing

Clock & timing forum

CDCM6208: Problems with I2C communication