Why so many Partial Discharge measurement units?

Partial Discharge is a poorly understood phenomenon, and it is even harder to understand because many different units are used to measure it.  Making matters worse, sometimes you have to calibrate a test set up, and other times you don’t.  This article will explain the various units, when and how they are used and why we calibrate (sometimes).

The scale of the discharge is the number of electrons, measured in Coulombs.  One coulomb is 6,241,509,750,000,000,000 electrons, which explains why we don’t work in coulombs.  One picocoulomb (pC) is = 6,241,509 electrons. Discharge of electrons can cause a current but only for as big & long as those electrons can produce. One coulomb can produce 1 amp of current for 1 second (1 amp-second). Similarly, a picocoulomb can produce 1 picoamp-second.  This could be 1 amp for 1 picosecond, or 1 picoamp for 1 second or anything in between.  Typically, you see pulses in the range of microamps for microseconds.

If you can see the shape of the current pulse, the area under the pulse shape is the charge in pC.  This is what high speed, high quality HFCT/RFCT based measuring devices do.  By measuring and calculating the area under the current /time pulse shape, you are directly measuring the number of electrons and therefore can read out the direct value in pC.  If your HFCT and measuring circuit are known to be accurate, no further calibration is needed.

IEC 60270 (High-voltage test techniques - Partial discharge measurements), the standard for offline PD testing brings in the concept of apparent charge.  This is a charge, when applied to the measuring circuit gives the same readings as the same charge in the test subject, which is not measured directly.  Because the relationship between the reading and the actual PD is variable and dependent on the item under test, we must use a calibrator to establish a known relationship and get usable readings. The calibrator injects a known charge, and the resultant reading is used to scale the measurement values.

So, what’s the deal with dB, dBmW, dBmV, dBµV?      And how many dB in a PicoColoumb?

Many manufacturers and even the NETA MTS standard use various units.  Ultrasonic and TEV readings are often given in dB, dBmV, or dBµV.  We get asked all the time what reading corresponds to what pC discharge.  These totally different quantities are measured with different modalities and can only loosely related to pC.

Decibels or dB is a unique scale developed by engineers to confuse everyone else. Seriously, it’s a great tool for simplifying relative measurements when there can be great differences. For example, the power output of the sun is around 300 dBmW (5×10²⁹ mW) and the power output of a laptop Wi-Fi signal is 15 dBmW (32mW) therefore the difference is 275 dB. You can see how makes it easy to work with big numbers and ratios. 

There are a few rules to remember:

a) Partial Discharge dB number can be negative. dB is a power ratio, not an absolute measurement. When we say “10 dB”, we mean 10 dB higher than something else.  When we say “-10 dB”, we mean 10 dB less than something else, not a negative level!

b) Partial Discharge dB number has a fixed reference. By adding a fixed reference, we can turn this ratio into an absolute measurement.  For example, dBmW is dB compared to a 1 mW reference.  So, 0 dB would equal 1mW.  A negative dB would be below 1 mW (-10 dBmW = 0.1mW) and a positive level is above 1 mW (+10 dBmW = 10 mW).

c) The dB number is logarithmic with the equation of dB = 10 log₁₀ (x/ref), where x is the signal being measured and ref is the fixed reference.  So, 10 mW = 10log₁₀(10/1) = 10.  Some common power level conversions are given below.

Table 1: Examples of Common Power Level Conversions

UHF PD Measurements (dBmW)

UHF measurements are made by a radio receiver picking up energy that is hitting an antenna.  Like most radio receivers, it measures power.  Therefore, UHF receivers tend to display in dBmW and use a scale like that described above.

TEV PD Measurements (dBmV)

TEV or Transient Earth Voltage is a measurement of a voltage on the outside of a grounded metal enclosure containing HV assets.  The partial discharge sends electrons to the surface of the enclosure where they go across the impedance of the ground system, thereby creating a voltage.  The TEV measuring device picks up that voltage through capacitive coupling.  The measurement is an actual voltage.  As the high frequency impedance of the ground system is unknown and the percentage of current flowing through that panel is unknown, and the coupling between the discharge source and that panel is unknown, it is impossible to consistently relate a discharge level to a TEV level.  So, the instrument displays the measured voltage on a logarithmic dB scale as described above.

dB is actually a power ratio, but TEV is measured in voltage.  So, when using the scale to indicate volts, the dBmV numbers are doubled from of the previous formula to dBmV = 20 log₁₀ (x/1mV).  This is because for a given impedance, the power is proportional to the voltage squared.  Below are conversions from the voltage to the dBmV scale.

Table 2: TEV PD Measurement Conversions

Ultrasonic PD Measurements (dBuV)

Ultrasonic measuring equipment takes the sound pressure level and converts it to a voltage via an ultrasonic sensor.  In similar fashion to TEV measurements, the voltage is presented in a logarithmic scale but now relative to 1 uV.  The formula is dBuV = 20 log₁₀ (x/1uV).  For reference, converting from dBuV to dBmV is as simple as subtracting 30.

Table 3: Ultrasonic PD Measurement Conversions

The ultrasonic levels are highly variable based on the location of the discharge, the location and direction of the sensor, as well as any blockage, reflections or absorptions of the signal.  The presence of clear ultrasonic discharge patterns, regardless of the exact level is a cause for concern.

So, how do I keep all these levels straight?

Remember these scales

• Offline PD = pC (inferred after calibration)
• Online HFCT based testing = pC directly measured
• Online TEV based testing = dBmV
• Online Ultrasonic based testing = dBuV

Several manufacturers and even the NETA MTS 2023 standard use the convention to just say “dB” as the units even though that is technically incorrect.

The severity of PD related to all these readings is best found in the NETA MTS standard and with manufacturers who have thousands of test results showing the relationship between the dBxx values and real-world PD.

If you are still confused, be glad you aren't in the voice telecom world, where it gets even worse with terms like "dBrnc0" (power level in dB above a reference noise source with a c-message weighting measured at zero transmission levelpoint).