Clarification on NI 6353 Clock Accuracy and Stability

[Sepehr.S]

Hello,

I am currently working with the NI 6353 X Series Data Acquisition (DAQ) device, and I have a few questions regarding its internal clocking system and stability:

1. Definition and Use of ppm:

   • The specifications mention 50 ppm accuracy for timing. Could you clarify how ppm is defined in this context?
   • Which specific clocks are affected by this ppm rating? Does it apply to the 100 MHz base clock, the sample rate clock, or other derived clocks?

2. Clock Stability Over Time:

   • Is there any data available on the stability of the clock over time (e.g., short-term vs. long-term drift, temperature dependence, or aging effects)?
   • Does NI provide any recommended calibration intervals specifically to compensate for clock drift?

3. Clock Sources and Oscillator Count:

   • How many oscillators does the NI 6353 contain, and are they all equally stable?
   • When using the default internal clock, which oscillator is driving the sample timing?
   • If an external clock is provided via PFI/RTSI, does the internal PLL fully synchronize to it, or does it only compensate for drift?

I would greatly appreciate any insights or reference materials on these topics.

Thank you for your time and support.

[ZYOng]

1. Definition and Use of ppm:

   • The specifications mention 50 ppm accuracy for timing. Could you clarify how ppm is defined in this context? Timing accuracy is typically measured in parts per million (ppm). To convert this accuracy value to Hz, multiply by the accuracy value divided by 1 million. See Timing Accuracy in Specifications Explained: NI Multifunction I/O (MIO) DAQ
   • Which specific clocks are affected by this ppm rating? Does it apply to the 100 MHz base clock, the sample rate clock, or other derived clocks?

2. Clock Stability Over Time: 

   • Is there any data available on the stability of the clock over time (e.g., short-term vs. long-term drift, temperature dependence, or aging effects)? I have no idea. I will leave it to someone else to answer.
   • Does NI provide any recommended calibration intervals specifically to compensate for clock drift? B/E/M/S/X Series Calibration Procedure provides step to verify timebase at page 18, as part of recommended calibration every 2 years. However, It is not possible to adjust this timebase, so only verification can
     be performed.

3. Clock Sources and Oscillator Count:

   • How many oscillators does the NI 6353 contain, and are they all equally stable? Only one, the remaining clock signals derive from it. See Clocks - NI DAQmx
   • When using the default internal clock, which oscillator is driving the sample timing? Same as above.
   • If an external clock is provided via PFI/RTSI, does the internal PLL fully synchronize to it, or does it only compensate for drift? Yes, as shown in the clock diagram.

[Sepehr.S]

Thanks. This is helpful. Now all that remains the stability of the clock over time.

[mcduff]

---

@Sepehr.S wrote:

Thanks. This is helpful. Now all that remains the stability of the clock over time.

---

Here's one data point for you. Some colleagues had three identical systems located at different areas in a facility. They collected data for 1 week continuously. There was a drift of 1-2s between devices over that time. The time here is calculated using the sample rate not the internal clock of the computer. Temperature fluctuations will cause drift in your clock.