RU2537104C2 - Separation method of integrated circuits as to reliability - Google Patents

Abstract

FIELD: electricity.

SUBSTANCE: invention relates to electrical engineering, namely to methods providing quality and reliability of integrated circuits (IC), both logic and analogue ones. The essence of the invention consists in the fact that on a representative sample there performed is measurement of critical supply voltage (CSV) before and after the burn-in test (BT) with the duration of up to 100 h and after thermal annealing with the duration of 4-10 h at the temperature that is the maximum allowable for this type of IC; then, coefficient M is found and IC is separated as per its value as to reliability. $$M=\frac{{Е}_{CR}{}_{{}_{BT}}-{Е}_{CR}{}_{{}_{ST}}}{{Е}_{CR}{}_{{}_{BT}}-{Е}_{КР}{}_{{}_{ann}}},$$

where $${Е}_{CR}{}_{{}_{ST}},{Е}_{CR}{}_{{}_{BT}},{Е}_{КР}{}_{{}_{ann}}$$

- values of CSV before BT, after BT and after annealing respectively.

EFFECT: proposed method allows reducing a risk of damage to the tested circuits at the action of external test factors on them.

Description

The invention relates to microelectronics, and in particular to methods for ensuring the quality and reliability of integrated circuits (ICs), both logical and analog, and can be used both in the production process and during input control at manufacturers of electronic equipment.

A known method of allocating ICs of increased reliability [1] for the selection of ICs, conducted by the relative magnitude of the change in the critical supply voltage at normal and elevated temperatures. The disadvantage of this method is the high complexity of measuring critical supply voltage (KNI) at elevated temperature, as well as low accuracy.

The closest is the method [2], according to which a batch of ICs is subjected to preliminary separation by the critical supply voltage (KNI) method and is exposed to it by electrostatic discharge (ESD) with a potential that is half dangerous, and after exposure to ESR, an additional batch of ICs with increased reliability, after which annealing of defects is carried out.

The disadvantage of this method is that there is a risk of damage to the circuit when exposed to ESD.

The invention is aimed at expanding functionality.

This is achieved as follows. The batch is preliminarily divided by critical supply voltage. According to the results of statistical data, the KNI interval is determined, into which most (> 90%) values fit. For circuits that do not fall into this interval, a conclusion is drawn about reduced reliability.

Further, the ICs are subjected to electrical heating (ETT) for up to 100 hours and thermal annealing for 4-10 hours at a temperature maximum permissible for this type of IC. According to the results of measurements of KNI before, after ETT and after thermal annealing, the coefficient M is calculated, which characterizes the reliability of the IC:

$$M=\frac{E_{\mathrm{TO}R}{}_{{}_{ETT}}-E_{\mathrm{TO}R}{}_{{}_{N\mathrm{BUT}H}}}{E_{\mathrm{TO}R}{}_{{}_{ETT}}-E_{\mathrm{TO}R}{}_{{}_{\mathrm{about}t\mathrm{well}}}},$$

- значения КНП до ЭТТ, после ЭТТ и после отжига соответственно.Where $$E_{\mathrm{TO}R}{}_{{}_{N\mathrm{BUT}H}},E_{\mathrm{TO}R}{}_{{}_{ETT}},E_{\mathrm{TO}R}{}_{{}_{\mathrm{about}t\mathrm{well}}}$$

- KNI values before ETT, after ETT and after annealing, respectively.

The formula has this form, because with ETT, IP parameters degrade, and annealing allows you to restore the effects of ETT and bring the parameter values closer to the original values. The larger this ratio, the less reliable the integrated circuit.

The method is as follows. On a representative sample of ICs, both logical and analog, KNI is measured, then ETT and thermal annealing are performed, after which KNI is measured, after which there is a coefficient characterizing the reliability of the IS: $$M=\frac{E_{\mathrm{TO}R}{}_{{}_{ETT}}-E_{\mathrm{TO}R}{}_{{}_{N\mathrm{BUT}H}}}{E_{\mathrm{TO}R}{}_{{}_{ETT}}-E_{\mathrm{TO}R}{}_{{}_{\mathrm{about}t\mathrm{well}}}}.$$

Depending on the value of the criterion M, established experimentally for each type of IP, it is possible to divide the batch of IP by reliability.

1. RF patent No. 2365930, G01R 31/26. Published 08/27/2009.

2. RF patent No. 2230334, G01R 31/26. Published 06/10/2004.

Claims (1)

A method for separating integrated circuits by reliability, according to which a critical supply voltage (KNI) is measured on a representative sample, a coefficient M is found and the IS is separated by its value by reliability, characterized in that the critical supply voltage is measured before and after electrotraining (ETT) up to 100 hours and after thermal annealing for 4 to 10 hours at a temperature maximum permissible for this type of IC, and the selection of integrated circuits is carried out according to the relative value of changes in the critical supply voltage calculated by the formula:
$$M=\frac{E_{\mathrm{TO}R}{}_{{}_{ETT}}-E_{\mathrm{TO}R}{}_{{}_{n\mathrm{but}h}}}{E_{\mathrm{TO}R}{}_{{}_{ETT}}-E_{\mathrm{TO}R}{}_{{}_{\mathrm{about}t\mathrm{well}}}}$$

,
Where $$E_{\mathrm{TO}R}{}_{{}_{n\mathrm{but}h}},E_{\mathrm{TO}R}{}_{{}_{ETT}},E_{\mathrm{TO}R}{}_{{}_{\mathrm{about}t\mathrm{well}}}$$

- KNI values before ETT, after ETT and after annealing, respectively.