• State
  • Locked Locked
  • Replies 1 reply
  • Subscribers 20 subscribers
  • Views 166 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

LMK04208EVM: A bit shift in values readback from the LMK04208 after writing

Lawrence Trevor1
Prodigy 90 points
Part Number: LMK04208EVM
Other Parts Discussed in Thread: LMK04208, , USB2ANY

Hi TI.

I'm looking to confirm some behaviour I've seen when using the readback function of the LMK04208, and see if there are any more 'gotchas' for the chip.

I noticed, using two different SPI master devices, that register values that are readback from the LMK04208 are shifted by one bit after writing to that register.

For example, I took R9, which has a POR value of 0x55555540 (not including the address bits at the LSB/right end). On the LMK04208EVM, I soldered closed R59, permitting the holdover pin to act as the read back. After setting register 13 to permit readback, I then performed the necessary writes and then read back R9 as 0x55555540. I then wrote this value back to R9 (0x55555549). I repeated the same steps to read back R9 and got 0x2AAAAAA0, which is the POR value right shifted by one.

I noticed this behaviour on one other register too, but have not verified that this exists on all of them.

A further bit of confusion - If I write the value using one of my SPI masters, disconnect it and then plug into the USB2ANY device and read the register through TICS pro, then I get the correct value back, with no bit shift. This makes me think I'm either doing something really wrong, or that it is something that TICS Pro actually accommodates.

A few questions:

1. Is this a known 'quirk'?

2. Is it documented? I could not find mention of this in the programming section of the data sheet. A google search and inspection of the part documentation page did not reveal an Errata sheet or similar.

3. Is this quirk apparent on all registers?

4. Are there any other of these features that might exist and that haven't been documented?

Below is a link to the screenshot for the testing on R9 that I did.

/resized-image/__size/240x1000/__key/communityserver-discussions-components-files/48/1638.lmk.PNG

Thanks!

Lawrence

  • 0
    TI__Mastermind 28960 points

    Hi Lawrence,

    I'm not aware of any quirk like this, and as such we have no documentation or awareness that it occurs.

    Note that the LMK04208 uses a MICROWIRE interface as opposed to a SPI interface, which has several subtle differences in readback behavior; specifically I would look at the distinction between READBACK_LE=0 and READBACK_LE=1 in the timing diagram on figure 4 along with reviewing the readback discussion in section 8.5.3, and make sure that you are not accidentally writing new values to the registers on readback.

    Regards,

    Derek Payne