Hysteresis in electronics and electrical engineering
Hysteresis in electronics and electrical engineeringDate of publication: 04-12-2025 🕒 6 min read
Electronics engineers or automation specialists mainly deal with hysteresis in temperature control systems or when using a transformer or choke. However, this phenomenon occurs not only in electrical engineering but also in other fields, even those with little connection to electric current, such as economics, biology, or mechanical systems. This text is, however, addressed to people dealing mainly with electrical engineering, so it focuses on the aspects of the occurrence and application of hysteresis in electrical engineering.
The general definition of hysteresis states that it is a phenomenon in which the output state of a system depends not only on its current input state but also on past events. In other words, the system "remembers" stimuli and reacts depending on the direction of change. This general definition refers to some kind of "memory" and makes sense in relation to economics or biology, but in electrical engineering, hysteresis is related to physical phenomena and is much easier to understand than in other fields.
Hysteresis in electrical engineering
For example, let's imagine an electromagnet whose winding current intensity can be continuously varied. Suppose it is linearly changed from 0 to a certain value that ensures core saturation and then linearly decreased back to 0. It can then be observed that an object placed in the electromagnet's field will be attracted at a higher current intensity than when it is released. This happens because before the current flow is turned on, the core material is either demagnetized or residual magnetized.
As the current intensity increases, the intensity of the magnetic field (H) and the magnetic induction (B) increase. After reaching a certain maximum value dependent on the core material type, the material "saturates," and the magnetic field intensity stops increasing despite further current increase in the winding.
After some time, according to the assumptions given, the current intensity reaches its maximum value and begins to decrease linearly, resulting in a decrease of the magnetic field intensity. It can then be seen that the value of magnetic induction B does not return to its original value tracing the same curve depending on H intensity but defines a new curve shifted along the abscissa axis. The reason is core remanence, i.e., magnetization remaining due to its magnetization. To completely eliminate magnetization, an oppositely directed field must be applied to the core.
If this situation is illustrated on a graph, we get the so-called magnetic hysteresis loop, which—as many people think—is not just an academic term but a real feature of an object affecting its application in electrical engineering. As an illustration, two oscillograms will be shown. Photograph 1 shows the hysteresis loop oscillogram of a 470 μH choke, while photograph 2 shows the hysteresis loop oscillogram of an iron-core choke operating as a component of an interference suppression filter connected to the power grid.
Photograph 1. Oscillogram of the 470 μH choke hysteresis loop
In electrical engineering, hysteresis occurs not only when magnetizing the core but also during heating and cooling objects, switching transistor inputs on and off, charging and discharging capacitances, and other situations. Sometimes it is a desirable phenomenon (a wide hysteresis loop is useful, for example, in magnetic media such as tapes and disks), intentionally created in control applications (such as thermostats controlling heating devices) and sometimes causes energy loss and excessive heating of components (such as in transformers, chokes, electromagnets, electric motors).
Applications of hysteresis
The basic application of hysteresis that immediately comes to mind is avoiding oscillations at the switching threshold of the control system. As an example, consider the simple task of heating a glass of water to 40°C. To this end, an electric heater will be used, and the task of the designer, whether an electronics engineer or automation specialist, will be to build a device measuring temperature and controlling its switching. An automation specialist will likely use a ready module in which at least the temperature value can be defined, and sometimes also the hysteresis loop width. Alternatively, an advanced software-based PID controller may be used. For an electronics engineer designing such a module, the task would not be so easy.
Nowadays, an electronics engineer building a thermostat would rather use not an analog comparator but a microcontroller and an analog or digital temperature sensor. Due to the nature of the microcontroller, the temperature must be quantized, meaning it will be measured with some accuracy. Abstracting from disturbances and distortions in the system, let’s assume we are dealing with a temperature/voltage converter with 12-bit resolution and a range from 0 to 150°C. It is easy to calculate that the temperature will be measured with an accuracy of 150°C/4096 steps, approximately 0.04°C. If the control program in the microcontroller used in the thermostat contains the command
if (MEASURED_TEMPERATURE < 40°C) turn_on_heater() else turn_off_heater()
it will turn out that the heater will be turned on already at 40.04°C and turned off at 40°C. With a small amount of water in the glass, switching on and off will be very frequent, which may lead to undesirable oscillations, sparking of relay or contactor contacts, and ultimately to damage of the newly built thermostat or significant electromagnetic interference.
Photograph 2. Oscillogram of the hysteresis loop of a suppression choke – an element of a power network filter
The remedy is to implement the Schmitt trigger hysteresis model, often used in digital technology and electronics. In this model, switching on occurs at one threshold value and switching off at another, and the gap between the two threshold values is set according to the control system accuracy requirements. In thermostats, an accuracy of about ±1°C is often sufficient, which can be written as:
if (MEASURED_TEMPERATURE < 39°C) turn_on_heater()
else if (MEASURED_TEMPERATURE > 41°C) turn_off_heater()
In this example, the threshold values are: turn on at 39°C, turn off at 41°C, and the nominal programmed hysteresis loop width is 2°C. The switching-on and switching-off thresholds lie significantly farther apart (about 50 times the measurement resolution), resulting in much less "nervous" thermostat operation, fewer switchings, and reduced level of EMI disturbances.
Similar techniques are used in other stabilization systems, but it must be admitted that while they work well in the simplest cases, more advanced control systems require continuous supervision and uninterrupted proper parameter adjustment, so other regulation methods such as PID algorithms and heating power regulation by PWM will be applied there.
A very popular system using threshold switching is the widely used Schmitt trigger gate in digital technology. While at the dawn of digital technology it was either a separate integrated circuit (e.g., SN7414) or the presence of so-called Schmitt inputs in a given component was emphasized, with the introduction of CMOS circuits, practically every integrated digital circuit has such inputs. This allows avoiding oscillations during slow voltage level changes at the digital input. To see this, it suffices to glance at the datasheet of any popular digital IC, such as a microcontroller.
Figure 3 shows a fragment of the table containing electrical parameters of a popular microcontroller ATmega328.
Note that:
- The minimum switching voltage (state change from low to high) is 70% of the supply voltage.
- The maximum input voltage for a high level is VCC + 0.5 V.
- The maximum voltage for a low-level input is 30% of the supply voltage VCC.
- The hysteresis width is 5% of VCC.
Supplying the mentioned microcontroller with 3.3 V results in:
- VIH MIN = 2.31 V; VIH MAX = 3.83 V,
- VIL MIN = -0.5 V; VIL MAX = 0.99 V,
- Hysteresis 0.165 V.
Thus, the input state change from low to high occurs after exceeding 2.31 V, but from high to low below 0.825 V due to hysteresis caused by saturation of input transistors. Between 2.31 V and 0.825 V is the so-called forbidden zone where the input does not switch. This simple way avoids oscillations during slow voltage changes at the input. Usually, these considerations are not very significant because care is taken so that the voltage at the digital input of the microcontroller changes as quickly as possible from near ground to near supply voltage, but connecting longer cables or cooperating circuits with large input capacitance can flatten signal edges.
It is easy to guess that without artificially implemented hysteresis, many systems would react too sharply even to small signal changes, including those caused by minor disturbances. Properly and consciously applied hysteresis introduces a dead zone or reaction delay, stabilizing operation, preventing too frequent switching, and increasing resistance to noise and interference.
Figure 3. Fragment of the electrical parameters table of the ATmega328 microcontroller
Transfer Multisort Elektronik (TME) is one of the world’s largest global distributors of electronic components, electrotechnical parts, workshop equipment, and industrial automation. The catalog includes over 1,500,000 products from 1,300 leading manufacturers. TME’s modern logistics centers in Łódź and Rzgów (Poland), with a combined area of over 40,000 m², ship nearly 6,000 packages daily to customers in more than 150 countries.
TME also invests in the development of knowledge and skills of young engineers and electronics enthusiasts through the TME Education project, and supports the tech community by organizing the TechMasterEvent series, promoting innovation and experience exchange.
