DC voltage

The direct current (non-changing) component of a voltage. In practice, the DC voltage should not change over the period of observation, that is, the measurement time.

DC measurements are used for a broad range of applications. These include the measurement of

• Static Load (pressure)
• Temperature
• Atmospheric pressure 
• Power supply voltages.

Reducing Noise in DC Voltage Measurements

Noise is often the largest source of error in DC voltage measurements, therefore the measurement must be set up properly to reduce undesired noise.

DC voltage often has noise (AC voltage) on top of it. Thus a certainly knowledge of noise and AC voltage measurements may actually be desirable to make accurate DC measurements.

The noise riding on top of DC signals can be broken into several categories:

• Normal mode noise. This is noise which is only present on one wire refereced to the other wire. This is in constrast to common mode noise, where the same signal is present on both wires. To get rid of normal mode noise, you must average the signal over a long enough time to reduce the noise as necessary. In many of the circuits found in traditional voltmeters, this "averaging" is performed by a low pass filter, which has a cut-off frequency below the normal hum frequencies of 50 or 60 Hz, so that these frequency components are rejected. In addition, the voltmeter may have an analog or digital averaging specifically designed the match the signal period of the disturbing hum frequency. This is illustrate in the demonstration below.

• • Periodic normal mode noise (such as sine waves, especially 50 or 60 Hz noise): When reducing the influence of 50 or 60 Hz signals, the integration time of the instrument must be a precise integral multiple of the undesired frequency. Otherwise, the noise rejection will not be as high. However, the longer the averaging time, the less important it is to use an integral number of periods in the measurement. You can try this in the interactive demonstration below.
  • Random noise has a broad range of frequencies of random amplitude, and the noise reduction is proportional to the averaging. time.
• Common Mode Noise. This is noise where the same signal is present on both leads of the signal wire leading to the instrument. The signals are in phase, and hence can be cancelled out if they are "seen" by a balanced (also known as a differential) input. The degree of balance of the input circuitry of the instrument determines the Common Mode Rejection Ratio (CMRR). Common Mode noise often arises due to cross talk from adjacent circuits or electromagnetic fields. However, since the same voltage is normally induced in both wires, it is possible to cancel it out.

It can be a significant advantage to use a digitizer coupled to a PC to measure DC voltage because it is thus possible to also view the spectrum of the noise signals sitting on top of the DC voltage. By examining the noise spectrum it is possible to make more qualified decisions about how much averaging/filtering is necessary to get rid of the noise. When using a traditional DC voltmeter, you are measuring "in the dark" because you can only see the end result, but not the actual waveform being measured.

Demonstration


The following demonstration lets you interactively see

• How to get rid of normal mode noise (50 or 60 Hz)
• How the averaging time for a DC measurement affects its accuracy.
• How to get rid of noise that is not precisely at 50 or 60 Hz.

Click Demo to illustrate the above concepts. (The demo requires that you have downloaded the LabVIEW Player)

Digital signal processing techniques using the Fourier transform (in the form of FFT) can be useful in separating the DC component from the undesired AC components. A new set of sophisticated measurement algorithms in LabVIEW make significant improvements in both the speed and accuracy with which the measurement of AC and DC signals can be made, also in relatively noisy environments.

Demo of DC Voltage measurement.

Reader Comments | Submit a comment »