you are here:   / News & Insights / Engineering Advantage Blog / What Was Not Lurking in the Heart of My Washing Machine

Engineering Advantage

What Was Not Lurking in the Heart of My Washing Machine

Washing Machine Cat | FEA Consulting
January 17, 2017 By: Chris Mesibov

Recently my son, happy to help out, decided to run the dishwasher and the washing machine at the same time. Normally, the only concern would be running out of hot water during the two wash events. Instead the circuit breaker tripped, causing the clothes washing machine to go dark. It failed to turn back on after power was restored. To make matters worse, I unfortunately failed to buy the extended warranty for this 2-year-old washer. The technician explained that the control card for the washing machine had failed when the power was cut and he indicated that I should have plugged the unit into a surge suppressor. 

Surge suppressors are typically used to protect sensitive electronics against transient AC power events external to your home or from some other electrically noisy appliance in your home, such as an air conditioner when it turns on or off. For these reasons, I make a point to plug my desktop computers and home theater components into un-interruptible power sources (UPS) in order to protect my equipment from these external events. What infuriates me about the washing machine event is that it was not an external event; the washing machine blew itself up when the power was cut!  How did that happen? 

The most likely scenario is that, at the moment the power was cut, the washing machine drum was turning a heavy load - in this case, heavy wet blankets, towels and sheets. The motor driving this load was drawing large currents and the motor, being an inductive device, is also essentially a reactive energy storage device, with the energy stored in the motor’s magnetic field. 

When the circuit breaker tripped, the AC connection to the washing machine opened and the stored energy had no place to go.  Normally, the stored energy is returned to the line. But, the AC line was disconnected, so the voltage on the line connections internal to the washer - and everything connected on that circuit experiences a voltage transient - (also referred to as back EMF), with the highest voltage appearing at the other components inside the washer. 

Inside many of today’s appliances are embedded computers which provide the user with the wash options control, and control the motor through the wash cycle.  This computer/controller requires an AC to DC power supply to supply various DC low voltages to the controller circuitry. 

An overvoltage can cause the following events at the input to a power converter:

1)    Post rectifier inrush into DC/DC converter filter capacitors.
2)    Ground/power noise coupling into sensitive circuits.

Inrush is an expected behavior at turn-on that would occur for every circuit that uses bulk capacitance to provide filtering, stability and local charge storage to a circuit. The input stage to most power circuits uses inrush controllers to meter the current into the capacitance (Figure 1), such that input fuses or other series devices do not get damaged when power is supplied. The inrush controller does this by monitoring the voltage drop across the sense resistor (Rsense typically set to milliohms) and controlling the MOSFET gate terminal accordingly. Figure 1 shows the path of the inrush current, from the power connector thru the transformer, rectifier, sense resistor, MOSFET and finally into the DC/DC converter’s input capacitors.

Figure 1: Example Inrush Current Path

At power-up, the circuit capacitors are at zero volts and appear as short circuits. As the capacitors charge, without any current limiting devices, they will draw huge currents until they are charged. The same is true if there is an over-voltage condition; even though those capacitors have an initial voltage, the surge voltage forces the capacitors to charge to that higher voltage, again at huge currents. A well-designed input circuit will have an inrush controller and a current limiting device such as a MOSFET that is slowly turned on to limit the energy dissipated by series devices such as rectifiers and printed circuit traces. The problem is that, for surge voltages, this scheme is inadequate for the following reasons:

1) Inrush controllers do not limit or clip the surge voltage.
2) Many inrush controllers only limit current at turn-on, secondary current surges are ignored.
3) Surge voltages on the load side of the DC/DC converter are not handled.

Typically, surge voltages are suppressed prior to the inrush controller by devices such as metal oxide varistors (MOVs) or Transient Voltage Suppression (TVS) diodes.  These components are similar to back- to-back Zener diodes. They clamp the voltage applied across the terminals preventing an overvoltage from damaging the downstream electronics.

I examined the damaged controller card (Figure 3 and Figure 4) and determined that it has a built-in motor controller.  This suggests that the damage occurred at the stator interface and not directly at the appliance AC input. Any overvoltage protection (clamping) must be applied either inside the motor controller chip or between the chip and the connector interface to the stator, as shown in Figure 2.

Figure 2: Motor Controller to Stator Interface

At this point, I concluded that the surge voltage occurred at the stator/controller interface and the damage was caused by one of the following conditions:

1)  There was no protection circuitry (i.e. TVS diodes) on the stator windings attached to the motor controller.
2)  The motor controller IC or PCB had TVS protection but it was defective.

Figure 3: Destroyed Circuit Trace Due to Overvoltage Surge

Figure 4: Washer Control Card Damage Integrated Circuit


A little bit of simulation could have gone a long way to preventing this problem. Figure 5 is a simple circuit developed using ANSYS HFSS that simulates the effects of a disconnected inductive load. A DC voltage source (V175) is fed into a switch that is controlled by a pulse source (V177). With inductance (L172) less than 10mH, huge voltages can be generated at the disconnection point of the inductor/switch node indicated by the voltage monitoring point (Vload).

Figure 5: Inductive Load Disconnect Simulation Circuit.

The surge voltage (Vin),could reach thousands of volts if not limited. Although the overall energy in the pulse might be limited, a high voltage could damage circuit card components causing them to short and subsequently draw large currents that would cause more damage, including burning up circuit card traces and destroying ICs, such as shown in Figures 3 and 4. The switch is opened and closed every half second. On switch closure a complete circuit current loop is established and the inductor charges.  When the switch opens, the current path is interrupted, with the voltage at the measuring point (Vload) surging negative as shown in Figure 6. Normally the waveform would show a 10V square wave, but the negative surge voltage transition is so large that it has expanded the voltage axis to the point where there is not enough resolution to see a signal with 10V magnitude.
 

Figure 6: Input Surge Voltage - Without Clamping


With the TVS clamp in place, as shown in Figure 7, the negative surge spike is limited to -10V, as shown in Figure 8. For sensitive circuits this still could be dangerous situation, but it illustrates, with the correct device selections, protection of circuitry attached to an inductive load can be obtained.
 

Figure 7: Circuit with TVS Diode Clamp

 


Figure 8: Surge Voltage - Clamped


Adding a properly selected TVS diode can clamp the surge voltage. The proper selection includes paying attention to the MOV or TVS device’s ability to absorb the anticipated energy, such that this protection device itself doesn’t get damaged. Using simulation to emulate your inductive device on disconnect can provide important data as to how much surge energy must be handled in the event of a power disconnection event. The SMDJ11CA TVS diode costs less than $0.50 in quantities of 1000. Had the appropriate surge protection device been installed on my washer’s control board, I could have saved hundreds of dollars of repair costs!

When you go out shopping for an appliance, it’s impossible to be sure how it is designed, where it was designed, or what components are installed (or not installed!). So, as a result, this experience has resolved me to consider my next washing machine to be simple with no internal computer. Have you ever experienced electrical failures like this that you would like to relate?  I would be interested in hearing your appliance tales of woe!