How necessary is an official NI Calibration or any DAQ hardware calibration?

[labview12110]

How much can a USB DAQ like a 6218 or a 6210 really drift in terms of it's analog voltage read over time? Is this even a thing? As long as I feed it 5V and read 5V with little drift over time, can I assume it is "calibrated"? What will NI do differently?

[Dennis_Knutson]

I'm guessing you are a student? If you were working, your company would have standards that, for example, would prevent a product from shipping without a valid calibration certificate.

[labview12110]

I'm confused abou that, i mean it just reads voltage, do daqs fall out of calibration ever? They aren't exactly a sensor in the traditional sense...What component would degrade in performance over time in a daq?

[Dennis_Knutson]

Any electronic device can degrade. There are numerous capacitors, resistors, etc. that can change characteristics due to voltage, current, temperature over time.

[Henrik_Volkers]

---

@labview12110 wrote:

I'm confused abou that, i mean it just reads voltage, do daqs fall out of calibration ever? They aren't exactly a sensor in the traditional sense...What component would degrade in performance over time in a daq?

---

If you do a measurement, it's for naugth unless you don't know the uncertaincy.  Almost all DAQ devices use an internal voltage reference in conjuction with a ADC and signal conditioning. If any of these blocks change due to temp, overvoltage, shock, time, dirt on the PCB, radiation, chemical&physical processes, .... you can't trust the number you read.

The manufactor of the metering device (hopefully)  has the experience how to choose, select, place and handle the components and after some testing he declares the specifications and the intervall and procedure how to check it. However statistics can beat and bite you, so the only way to be shure is regular calibration.

After some time and multiple calibrations you can start to trust your device. 

Since I do my own calibrations (the standards I can access have less uncertaincy than most other cal labs 🙂 ) I can tell you that I found NI devices out of cal spec that needed justification to meet their spec for the next periode.

In modern DAQs there are usually only two 'screws' you can access and 'adjust' : The internal reference timebase and the value of the internal reference.

While the crystal of the timebase can be trimmed (varicap and DAC) or not (in cheap devices) , the internal voltage reference value is internally measured,calculated and stored by applying a known voltage. Details are found in the calibration documentation.