Forum Discussion

Altera_Forum's avatar
Altera_Forum
Icon for Honored Contributor rankHonored Contributor
12 years ago

MAX V Oscilator Trimming

Hi

I'm going to use the internal oscilator on a MAX-V device to produce some (roughly) acurate timings. (usually in the 100s of mS or seconds)

The datasheet for the MAX-V says the OSC frequency is between 3.9HMz and 5.3MHz. Dividing down to give a 1mS clock from the mid point frequency (4.6MHz) would give a range of 0.87mS to 1.18mS. In some systems this might not be noticed but a customer might notice that one of his systems starts in say 8 seconds but another takes 11 seconds.

What i need to find out (but can't see anywhere on any data sheets) is what affects the frequency. Is the frequency related to temperature and supply voltage or is it reasonably stable over temperature and will simply vary from device to device.

If the frequency can drift all over the range on a sigle device affected by temperature (and possibly other things) then I just have to put up with the potential variation in frequency.

BUT if the frequency is reasonably stable then i have an opportunity to make a frequency divider that uses a value that could in theory be stored in the flash memory of the device. Thus allowing me to adjust the timings to improve accuracy (via memory interface on mys system)

Any comments would be appreciated.

2 Replies

  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    12 years ago

    I agree - in that I cannot find anything definitive in Altera's documentation.

    So, to speculate - any single internal oscillator is not going to drift wildly across the specified frequency range with voltage or temperature. I'd suggest this is true of any type of oscillator circuit used internally, be it crystal, ceramic or capacitive based. The specified frequency range is far more likely to be as a result of process variation.

    If you intend to characterise each device as you suggest, I expect you're likely to end up with a solution far more accurate than the specification suggests.

    I look forward to a definitive answer - should there be one:)
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    12 years ago

    --- Quote Start ---

    I agree - in that I cannot find anything definitive in Altera's documentation.

    So, to speculate - any single internal oscillator is not going to drift wildly across the specified frequency range with voltage or temperature. I'd suggest this is true of any type of oscillator circuit used internally, be it crystal, ceramic or capacitive based. The specified frequency range is far more likely to be as a result of process variation.

    If you intend to characterise each device as you suggest, I expect you're likely to end up with a solution far more accurate than the specification suggests.

    I look forward to a definitive answer - should there be one:)

    --- Quote End ---

    I logged a support request with Altera Support. They could not help either. All they would say is that the frequency will stay within the specification. When I get some boards working I'll run some tests and measure the frequencies (with some supply / temperature variation). I suspect that the wide frequency range implies the oscillator is capacitor based.