ESD simulation models

from Wikipedia, the free encyclopedia

As part of ESD simulation models artificial and reproducibly electrostatic discharge ( english discharge electrostatic , ESD) generated by test equipment to protect against electrostatic discharges to consider the impact and evaluate the various degrees of seriousness function of the ESD protection device in electronic devices.

ESD discharges can destroy unprotected electrical components, especially microelectronic components. Protected electronic devices can be disturbed in their function by ESD and this can lead to temporary malfunctions such as a computer crash. There are various ESD discharge models for studying and evaluating the effects of ESD.

Artificial ESD discharge with a test device to the connections of a set-top box .

ESD discharge models

Stress methods with two galvanic connections

A circuit is set up in such a way that there is a complete line-bound current path for the discharge. The discharge current therefore flows into one component connection and out of another. Since, in principle, all combinations of connections have to be tested, a large number of combinations can occur with multi-legged components (1000 connections are not uncommon in integrated circuits today). The test can therefore take a long time, or it can be very time-consuming to find out the relevant (destructive) combinations.

HBM - Human Body Model

This is historically the oldest model, based on the discharge of a human body with less than 40  kV . The human body has a typical capacitance of around 100 pF to 300 pF and an electrical resistance R of around 1500  Ω (skin resistance, spark gap). When charging to z. B. 10 kV, 10  mJ are stored as energy and the charge is 2  µC . The rise time of the discharge edge is in the nanosecond range, the current peaks reach up to 20 A with fingertip discharge  .

Apart from the history, a certain capacitance is discharged in the HBM by a very large, ohmic resistance. The current is therefore determined solely by the resistance and not by parasitic capacitances and inductances, the curve shape is well damped and does not oscillate. HBM superstructures are therefore very easy to reproduce today.

MM - Machine Model

The basic idea here is the unloading of a machine against a component. Since a metallic contact is required, the resistance is very small in contrast to HBM. The capacitance, on the other hand, is somewhat larger and the line inductance has a strong influence. This results in a weakly damped oscillation with an average frequency of around 20-100 MHz. Because of the small resistance, the current peaks with MM are typically 10 times higher than with an HBM structure with the same precharge voltage. The parameters LRC of the MM are standardized and implemented less precisely than the dominant parameter R in HBM models, so the reproducibility is poorer.

System level model

There are still inconsistent proposals that are based on the idea of ​​"people with screwdrivers", for example when manipulating finished devices in the service area. In contrast to HBM, there is metallic contact from the device to the hand with a screwdriver, which causes a very hard, short, first current surge of up to 30 A, superimposed by the slower discharge of the remaining capacity of the arm and trunk. Accordingly, the model circuit contains two charge stores with corresponding resistors. Individual components cannot be protected against such brutal events or can only be protected with uneconomical effort. However, this is only required in exceptional cases because the investigation no longer revolves around the handling of the individual components, but rather that of finished devices or circuit boards with several components. Natural spark gaps and components such as capacitors or special ESD protective components can absorb most of the energy so that the extreme current peaks no longer reach the sensitive components.

Stress methods with displacement current

CDM - Charged Device Model

Discharges of a charged component - z. B. during assembly processes - against conductive parts occur in such a way that only one connection touches the conductive part (e.g. the machine). The circuit is closed by the capacitance of the component in relation to the machine, i.e. not galvanically, but by a displacement current. The capacitance of the resulting RLC series circuit is therefore relatively small and depends on the component itself and its distance from conductive surfaces. The parameters R and L are extremely small. The discharge is therefore extremely fast. Depending on the manufacturing technology, the internal protective circuits of the components may or may not react. In contrast to the other methods with two galvanic connections, the discharge paths in the component are characterized less by signal connections than by the supply connections. This also leads to other error patterns. The (slow) charging of the component can be achieved by wire or with the help of a field plate; this fact alone is irrelevant for the measurement. In contrast, the parasitic elements of the structure (LRC) are very important. For this reason there are different types of CDM devices. The peak value of the current and the curve shape are even less dependent on the precharge voltage and the measuring device, so that the results cannot be compared on the basis of the precharge voltage alone.

ESD discharge models for investigating radiation resistance

ESD guns

ESD test pistol

There are also "ESD test pistols" for generating flashovers and contact discharges in accordance with the IEC / EN 61000-4-2: 2008 standard and other harmonized standards. In practical use, the generated current pulses are very variable because the environmental conditions (distances to conductive surfaces, length and position of the grounding cable, etc.) vary greatly and the generated frequencies are high in relation to the size of the current loop. The discharge can be measured using a 2 ohm series resistor on a suitable oscilloscope .

Simulation of electrostatic discharges according to IEC / TR 61340-1

(Toggle switch LEFT) The capacitor C is charged via a resistor of approx. 1 MΩ to the direct voltage U dc ; (Toggle switch RIGHT) Object with capacitance C is fed into the DUT via a resistor R and an inductance L, i.e. H. discharged into the component / system to be tested

Table - Typical values ​​of models for ESD simulation

model R (Ω) C (pF) L (nH)
HBM 1,000-3,000 100-300 Parasitic
CDM <10 3 - 30   Parasitic
MM 8.5 200 0.5

Web links

Commons : Electrostatic discharge test equipment  - collection of images, videos and audio files