The SKF Tked1 is an instrument designed to detect discharges of electromagnetic radiation in the High Frequency Band.

Any pulse of energy be it Mechanical or Electrical is a sharp peak of short duration and this can be seen in a spectra (FFT ) which creates High Frequency Harmonics typically in the MHz range.

The SKF Tked1 manual shows the Instrument is tuned to detect frequencies between 50 to 200 MHz.

As Energy in high frequency is emitted it attenuates very quickly and is very directional.

If you compare Low frequency sound (eg Bass Speaker) it travels a lot further and spreads making it difficult to detect the source.

With this concept in mind when using the SKF Tked1 bearing currents which are emitting this energy at boundaries are very directional and localized.

This localization can be in a very small area, therefore when scanning a quick wave around is likely to miss a potential problem. The full surface of the asset needs scanning at slow speed, take your time to conduct a good job.

An Example: I have seen this Bearing current run through just one motor foot, and in only 2 cm width of that foot trying to find it’s path back to earth. All other areas on that Motor had zero counts.

Secondly the wand of the SKF Tked1 is designed to pickup radiated energy meaning it’s in the air. Do not touch the Asset with the Wand head as this now will pick up conductive energy giving you a false reading. Keep the wand a consistent distance of approximately 2cm from the asset.

Where do you find EMF Counts and what levels are bad?

The source of these bearing currents come from the Inverter in the form of Common Mode and Differential Mode Current, and enter the motor winding.

Depending on Parasitic capacitance most of this energy is reflected from the Inductance of the first few turns of the coil, this attack decays the insulation propagating a winding failure.

Ensuring a motor is inverter duty helps slow this attack, but there are many techniques to ensure inverter duty some better than others. Wire grade is quite dominant.

The first method I adopt is scanning the Motor Cable at boundaries, these boundaries are Isolators, Motor Termination, and any joins in the conductor. The reflected energy often in phase with the supply amplifies and this voltage can be more than twice Line Voltage.

Many Inverter duty Motor have wire resistant to 1600 V which again slows the decay.

This reflection of voltage is often localized approximately 30 cm back from the termination (Impedance Change) and is an area to ensure thorough scanning around the circumference of the cable.

Note: SWA Steel Wired Armour cable that doesn’t have an extra screen is magnetically porous and will be extremely noisy. This cable is NOT recommended for inverter applications and causes a lot of the underlying problem.

Many OEM manuals state that Cable runs of more than 75 feet (25m) are prone to these reflective voltages. In these cases grounding is of the utmost importance.

The counts for a 10 second period (Recommended Standard) on a well grounded screened cable should be ZERO.

This grounding of the cable can be measured and verified by transfer impedance measurement and 3Phi Reliability have developed a method to measure this which taught during skill transfer.

Conducting a similar scan around the cable terminations of the Inverter will also indicate how well the grounding has been made. The counts in this location will be much higher as the emission attenuates quickly down the cable. A bad reading is normally in the 10 of thousands for 10 seconds.

Care obviously needs to be taken when measuring in this area, as flash over risk is much higher.

Once these emissions enter the Motor the path of bearing currents is determined by Capacitance.

Capacitance due to Xc = 1/2FC means capacitance is a dead short to High Frequency current.

The path of the current is desired to be Motor Winding to Motor Frame (Largest Capacitance in the Motor), so grounding the motor frame is important.

The second parallel capacitance path is the Bearings which is a smaller capacitance. Therefore most bearing current can be controlled through verified grounding diverting this HF Current to earth through the frame. Badly grounded motors suffer bearing currents.

The current path of a badly grounded motor flows frame to shaft (Through the Bearings) and therefore grounding brushes can increase bearing current by making this path lower impedance. This current path is described in research Oct 2023 Darmstadt University.

If the SKF Tked1 picks up counts at the motor frame but lower or no counts at the bearing housing then suspect a badly installed or maintained brush.

Insulated Bearings also have capacitance due to the insulating material but this decays under HF Current and numerous failures occur in a short period of time.

Insulated bearings are best suited to stop circulating current that occurs from Impedance Imbalance.

Scanning the bearing housing at approximately the Outer race boundary is recommended, then move lower to the motor feet. Also if possible scan the coupling as badly grounded motors push current to the least impedance that could be the driven asset.

Note: For various reasons the NDE Non Drive End Housing tends to be more prone to Bearing currents, even with insulated couplings. This is covered in 3Phi’s skill transfer.

The root cause of these Bearing currents is Poor Grounding and should be addressed before other devices are fitted. Most of the Bearing Current problems are poor installation or design, Cable selection.

Whats a Bad count?

Using a 10 second reference time period, Motor size tends to be a determining factor in severity.

Larger Motors eg >200 kW should have ZERO counts at the Motor Frame/Housings.

Smaller Motors tend to be less prone and a 100 count may not trigger alarm. Eg SEW MoviMots are noisy and a few 1000 counts is bad, it’s difficult to get assets to ZERO.

I have been in a facility where high counts were detected on the building frame, verifying the Earth Loop Impedance showed the Electrical earth was poor.

It is recommended to follow the PE back scanning each termination point, any counts is bad.

What to do if Counts won’t go to ZERO after remedial work?

Stopping the Drive emissions entering the motor is by far the best strategy. Brushes or Insulated Bearings will help (Duration dependent) but do not stop the Motor Insulation Attack.

Installing EMF Cores at the drive stops these residual emissions, and also stops cross talk between cables in the same cable tray.

These EMF Cores are cheap permanent fixes that aren’t prone to maintenance requirements.

Limitations of the SKF Tked1

The SKF Tked1 is an instrument that detects emitted pulses of energy across a boundary.

When that boundary becomes conductive the impedance of that boundary drops and the emitted energy also drops.

The analogy is an electric fence. If you stand on the wire pushing it into the mud and voltage drops close to zero, but the actual Current increases.

The recommended method of measuring bearing current is measuring the Current which excludes any issues around boundary changes. Eg Blacken grease becomes conductive and if severely decayed will see less energy pulses detected on the SKF Tked1.

This method uses a High Frequency Current coil and a Scope to directly measure this current.

The method is taught in 3Phi skills transfer.

I have been using the SKF Tked1 for more than 25 years, and the experience during that period is summarized in this guide.

Mark Gurney Motor Analyst

3Phi Reliability SARL (The GO TO bearing fluting specialists)

www.3phi-reliability.com