In the next simulation (Figure 6), we represent a full switch-mode power supply (SMPS) voltage rectifying and smoothing circuit, applying a 440 V peak to a load of 290 Ω. We can place switches that will deactivate the PTCs once the capacitances are charged at the level of 95 % of the voltage amplitude. And as the reboot model runs 20 times faster than the previous ones, the presented results are achieved after a few seconds (a real time of 1 s was simulated here at a speed of 200 ms/s, thus results were ready after 5 s per run). Monte Carlo tolerances have been introduced for the PTC (± 30 %) and for the DC link capacitor (± 20 %). Ten runs have been performed.
We will represent the temperature variation of the two PTCs (Figure 7), the voltage rising around the capacitors (Figure 8), and the current into the resistive loads (Figure 9).
We see in Fig. 7 that the capacity charging will take from 0.25 s to 0.5 s, driving the PTCs to temperatures between 70°C and 90°C (under the switch temperature). When the voltage has reached 95% of 440 V (Fig. 8), the switches close and the current is brought to the load (Fig. 9).
We can now push all knobs in the red by simulating the charge of a big 4700 µF capacitance in the same circuit as Figure 6 and visualizing the circuit behavior. After a very short time, the two PTCs switch above 160°C (Figure 10), but do not pass 170°C (limit of the validity of our new model). Then a bit surprisingly, the charging of capacitors, which seem to saturate, continues (Figure 11) while the PTCs cool down (Fig. 10). Current switch on the loads occurs between 25 s and more than 1 min.
Now, we will end this fireworks show of curves and waves with a circuit network featuring up to eight PTCs (Figure 12). The smoothing capacitances C1 and C2 have been put into short circuit (with infinitesimal R3 and R4 resistances) in order to induce a huge current to make the PTCs switch.
Figure 13 shows the temperature evolution of each of the eight PTCs (note that an initial ambient temperature of 70°C is chosen, a very limited case).
Half of the components switch after 0.7 s, ending up at a temperature around 165°C. This was completely expected. If you consult your PTC electrical engineer technician, they would tell you that one element in every PTC network branch will switch, and that is the component with the highest value. The series element with this switching part will cool down.
In conclusion, the simplified reboot model not only confirms what the PTC technical specialists had already stated, but it also allows us to produce very fast, predictive results.
There is no need to stare at the screen for hours, no wasted time, and no glitch in the simulation anymore; it’s the ideal way to prepare experimentation.
As always, the models and simulation shown in this article are available by request at firstname.lastname@example.org.
- Simulation Notes for SPICE Modelling PTCTL, PTCCL, and PTCEL, https://www.vishay.com/doc?29180
PTCEL data sheet https://www.vishay.com/doc?29165
- 2021: A Simulation Odyssey for Thermistors
- 2021: A Simulation Odyssey for Thermistors, Part 2
- Designing with temperature sensors, part two: thermistors
- Thermistor and DMM Measure Temperature
- Thermocouples: Simple but misunderstood
- Thermocouple physics: Even Jim Williams got it wrong