Hello there,
We need an ultra-precise counting mechanism for the N2HET timer machine in order to reduce as much as possible the induced error from the OSCIN. Right now we want to operate the N2HET at 100 MHz to have each HR of 10 ns and LR of 40 ns (4 instructions). The resulting counter for each sampling period is between around 1.5 to 2 million counts. We will use a crystal with a tolerance of 1 ppm or maximum 2 ppm.
We understand that the crystal used in the TMS570LS HDK is a 16 MHz so that we can program the PLL in order to obtain the desired CPU/VCLK operating frequency. We don't use FlexRay in our design so the following requirement in the TRM doesn't apply to us:
The OSCIN frequency is limited to 10MHz, 16MHz or 20MHz to reach the 80MHz required by the FlexRay sample clock. The crystal must be selected accordingly.
Here comes my questions:
- To obtain the most precise possible CPU operating frequency of 100 MHz, what OSCIN frequency is recommended?
- Is a 20 MHz OSCIN better suited than a 16 MHz since the desired 100 MHz is a multiple of it, or it doesn't make any difference for the PLL?
- What is the maximum input frequency to the OSCIN?
Thank you very much.
