Preliminary Application Note 4

TEST PULSE WIDTH ISSUES FOR TLP

The 1/e fall time to 37% amplitude for HBM testing is approximately 150 ns. The rectangular pulse width for an equal energy TLP pulse would be 75 ns. The nominal rectangular test pulse for TLP has been about 100 ns. This pulse width may have become widely used because it provides closer correlation to HBM testing. This in turn may be because the temperature in the silicon for the HBM pulse peaks at about 230 ns and then begins to decay. (1)

There is another way that a TLP pulse could be produced that would give a similar temperature time to peak but still retain a flat voltage at the end of the pulse for accurate I/V measurements. This would be a combination of two different amplitude pulses to simulate the shape of the capacitor discharge found in HBM exponential decay pulse testing. For a similar time/temperature threat, the TLP pulse should provide a peak in temperature closer to 230 ns rather than a steady rise to maximum temperature at 75 ns provided by an equal energy threat. An initial high amplitude pulse followed immediately by a lower amplitude pulse that lasts a total of 230 ns would give a closer simulation of the temperature rise from an HBM pulse.

References:
1. Investigation on the Thermal Behavior of Interconnects under ESD Transients using a Simplified Thermal RC Network. P Salome et. al. 1998 EOS/ESD Symposium p. 191 Fig. 5

Return to Designer's Corner.