• State TI Thinks Resolved
  • Locked Locked
  • Replies 6 replies
  • Answers 1 answer
  • Subscribers 85 subscribers
  • Views 1633 views
  • Users 0 members are here
Support feedback
Options
Options

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.

CCS/TMS320F28035: internal crystal, system clock

Houjian Xu
Intellectual 280 points
Part Number: TMS320F28035

Tool/software: Code Composer Studio

Hi all,

The customer has very strict requirements on system clock. They tested the 28035 using internal crystal， there are few micro seconds error in one minute. There is a 10MHz crystal in 28035? I want to know the certain accuracy of the internal crystal?

  • 0
    TI__Guru* 95315 points

    Houjian,

    I presume you are referring to the Zero-pin on-chip internal oscillator, not a crystal, which is external to the device. The information you are looking for is in Table 5-7 in page 37 of SPRS584K.

  • 0
    TI__Guru 53195 points
    For people who may stumble across this post in the future, this question was answered in this thread:

    e2e.ti.com/.../618733
  • 0 in reply to Whitney Dewey
    Intellectual 280 points

    It give the information about the frequency varied vs temperature. However, it didn't tell me the max frequency and the min frequency of the internal clock.

  • 0 in reply to Houjian Xu
    Intellectual 445 points
    Hi Whitney,

    If this is similar to the 28377 delfino parts, the internal oscillator is fixed frequency. However, because of manufacturing it is difficult to get these internal oscillators trimmed to precision. With the 28377 we tried to use the internal oscillator (against TRM recommendations) as the source for the CAN clock. As predicted, we failed miserably.

    Fortunately, we designed in a more accurate external oscillator because we thought ignoring the data sheet warnings on this internal oscillator may prove fruitless.

    Bottom line is that you will likely need to use a more precise external oscillator to get the precision you need. The data sheet will tell you the error you can expect from this internal oscillator (and you can use that to calculate the min/max), but it is unlikely there will be registers that will allow you to tune/trim it....

    Good luck
    Kevin
  • 0 in reply to Kevin Stallard
    Intellectual 445 points
    I just took a look at your data sheet, and you are right, there doesn't seem to be any numbers that would allow you to judge its accuracy except for the temperature vs frequency chart. However, judging from this temperature vs frequency chart, I gather that this isn't a precision trimmed internal oscillator (as it has a max and typical lines drawn on it).
  • 0 in reply to Kevin Stallard
    TI__Guru 62975 points

    To elaborate a little more on this, the zero-pin oscillator is very temperature sensitive.

    For the F2802x, F2803x, F2805x, F2806x series of devices, there is a temperature trim software sequence that can be executed to nudge the frequency back to about +/- 1.5% accuracy for the temperature at which the trim is executed. The trim sequence needs to be executed periodically (as frequently as the temperature is expected to change) in order to keep it in the +/- 1.5% range.

    The F2807x and F2837x devices do not have a software trim sequence. Instead, the oscillators are continually kept within a +/- 3% range using analog circuitry. This eliminates the overhead of software trimming, but also widens the trimmed error range.