• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Subscribers 82 subscribers
  • Views 1485 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

TMS320F28335: Maximum LSPCLK frequency

Dainius Stankevicius46
Prodigy 60 points
Part Number: TMS320F28335


Hi,

I have found a note regarding maximum LSPCLK frequency in TMS320F28335 datasheet saying:

"Although LSPCLK is capable of reaching 100 MHz, it is specified at 75 MHz because the smallest valid "Low-speed peripheral clock
prescaler register" value is "2" for 150-MHz devices." (doc. SPRS439N, pg. 43, Table 5-4).

This sentence is a little confusing, because the actual prescaler for LSPCLK is ([the value of  "Low-speed peripheral clock prescaler register"] x 2). Thus, if the minimum value is "2", the prescaler would be "4", and the LSPCLK frequency would be 37.5 MHz. But I guess this is just a small mistake and it was supposed to mean the actual prescaler of "2", not the value in the register.

More important question: what happens if I set the prescaler to "1"? Can the LSPCLK get into some undefined state? If so, why wasn't it mentioned anywhere else in the documentation, for example at the description of the prescaler register (doc. SPRUFB0D, pg. 39, Table 19)?

Regards,

Dainius

  • 0
    TI__Guru* 95315 points
    Dainius,
    Your observation is correct. Note (4) in the datasheet should call out a value of "1". It should read "Although LSPCLK is capable of reaching 100 MHz, it is specified at 75 MHz because the smallest valid "Low-speed peripheral clock prescaler register" value is "1" for 150-MHz devices." For a register value of 1, the effective prescaler value is 2 , as you correctly infer.

    An effective prescaler of 1 could indeed result in unpredictable operation of the peripherals clocked by the LSPCLK. The expectation is that designs would always conform to datasheet limits. That is why there is no explicit mention of what would happen otherwise.
  • 0 in reply to Hareesh Janakiraman
    Prodigy 60 points
    Thanks Hareesh.

    Of course, I agree that designs must always conform to the datasheet limits.
    However, it would be nice if such important things were easier to notice. There could be a short note at the description of the register. Something like "Check the exact device datasheet for the valid range of the value of this register".
  • 0 in reply to Dainius Stankevicius46
    TI__Guru* 95315 points
    Danius,
    I have already filed a correction for the datasheet to change that "2" to a "1". I will file an enhancement request for SPRUFB0D to include a comment similar to the one you mentioned.