Why is THD higher than SFDR in specification? Why is the SFDR Test done with a -60dB Signal?

[Hinrichs]

NI Hardware: Multifunction DAQ device USB-9211

The following questions arise in the decision making process regarding a 24bit-  or 16bit - ADC for acceleration measurements on a gearbox

Why is the  THD value higher than the SFDR in the specification? The harmonics are excluded for the SFDR measurement?

Why is the SFDR Test done with a -60dB Signal?

Whats about a 0dB or -3db sine tone? I expect the SFDR will be much lower.

[UM]

Hello Mr Hinrichs,

Could you please give me more detailed information’s, about where you found this data’s about THD and SFDR. I had a closer look to the specifications of the USB-9211, but I couldn’t find any information in the user manual  USB-9211/9211A User Guide and Specifications Are you sure, you named the right devices in your post?

I’m waiting for your response, so we can go further with this issue

Kind regards

Ulrich Möschel

Application engineer

[Hinrichs]

sorry, must be usb-9233

[edjones93]

Total Harmonic Distortion is the sum of distortions over a frequency range.

Spurious Free Dynamic Range is the maximum spur across a frequency range and is not related to harmonics as harmonics are a multiple of the frequency.

More can be found here:

https://www.ni.com/docs/en-US/bundle/ni-scope/page/spurious-free-dynamic-range.html

THD Should always be greater than SFDR as THD is a sum of harmonics whereas SFDR relates to one specifically which is the maximum across a spectrum.

[dsb@NI]

Yes, harmonics are often excluded from SFDR measurement, and that appears to be the case in the published specifications for the 9234.

With a 9234 input, at a test level of 3.5 V pk (-3.1 dB ref 5 V), I consistently measure SFDR to be 112 dB or better. 

[dsb@NI]

---

---

THD Should always be greater than SFDR as THD is a sum of harmonics whereas SFDR relates to one specifically which is the maximum across a spectrum. 

Generally, there is no set relationship between THD and SFDR. If you have a case where you are including harmonics in SFDR and the max spur is a harmonic of the test frequency, then |THD| <= SFDR. 

[dsb@NI]

---

---

sorry, must be usb-9233

I refer to the 9234 in my responses because that is NI's current recommended 4 Ch, 24-bit, c-Series device.