THE VLSI HOMEPAGE

Need for Logic Built-in Self Test (BIST)

Traditional scan requires large number of vectors to sensitize the design, runs at a maximum frequency of 50 MHz and is limited by number of channels supported by the tester. All these add to tester time that varies from 25 to 50 cents per second. Many designs integrate Logic BIST to overcome these limitations and reduce cost of testing.

Logic BIST, in brief words, involves driving control signals from an in-built controller, generating pseudo-random patterns on the chip, compact the responses from these patterns on the chip - All these occurs at-speed reducing the interface to the tester, the tester memory and also tester time.

Logic BIST Architecture

Logic BIST Architecture

The figure above shows the architecture of Logic BIST that is based on the traditional scan based architecture (known as STUMPS model). The primary instances in this model are:

• Pseudo-random Pattern Generator (PRPG) - this is implemented using linear feedback shift registers (LFSR) to generate pseudo-random patterns to stimulate the design. The LFSR is “maximal length” by nature which means that it visits each and every state before repeating the sequence.

• The Phase shifter block ensures that a large number of scan chains in the design are driven using a short LFSR by using phase-shifting techniques. This phase shifting also removes any inter-channel dependence between input channels. There are muxes at the input of the scan chains to select either traditional scan inputs (muxed-scan) or from PRPG to achieve more fault coverage.

• Space Compactor compresses the output of these scan chains using XOR logic before feeding the compressed outputs to Multiple Input Signature Register (MISR). The MISR outputs are then compared internally with an on-chip reference signature or are scanned out of primary pins.

• A BIST controller that controls the generation of clock control and scan enable signals apart from counters to track the shift cycles. A TAP interface is also integrated in the controller to initiate logic BIST through JTAG.

The shift pattern is determined by the longest scan chain path and also the number of capture clocks (usually one) in the design. Patterns from PRPG are shifted into the scan chains while simultaneously being compressed at the other end into MISR for better utilization.

The design requirements are stringent - no unknown “X” sources (like memories, non-scannable flip-flops), design should be pseudo random pattern testable with minimal area overhead. Any “X” sources can cause corruption of MISR outputs and hence control and test points need to be added in the design. The advantages far outweigh the disadvantages for complex multi-million gate designs - LogicVision’s LogicBIST and Mentor’s TestKompress are two well-known DFT tools for logic BIST.

Share This

Traditional scan based designs employ either Muxed-scan technique or Level Sensitive Scan Design (LSSD) techniques to achieve test coverage. In scan based designs, the registers are hooked up to form serial shift register chains - this aids in capturing all the combinational logic faults between two pipelined registers.

There is a shift phase during which the ATE pattern is shifted serially through the scan chain from the IO pins. Once the ATE pattern is shifted through the scan chain, there is a capture phase that allows the flops in the scan chain to capture the combinational logic output at their input pins. Following this capture phase, the pattern is shifted out serially and compared with the expected vector from ATPG.

Each clock domain can have it’s flops stitched into a single/multiple scan chains, based on the number of flops in the clock domain - stitching multiple scan chains is an advantage as it reduces tester time since we can parallely load all the scan chains. Usually mixing clock domains or posedge and negedge flops in a scan chain is not recommended as it can cause timing issues. Use of lock-up latches is advised if crossing from posedge to negedge flops in a scan chain.

Advantages of scan based design are several - high fault coverage with moderate increase in logic, the entire insertion is automated. The main disadvantage of scan based testing is that it runs at low speed typically 50 MHz - this is very slow for high speed designs and where at-speed testing is critical.

Muxed scan

The figure below shows a muxed-scan flop and associated waveform - the D pin is the normal functional input, SI is the scan input and SE is the scan enable to shift in and shift out data out of the flop. The Q output of the flop is connected to SI pin of the next flop in the scan chain. During the capture phase, the Q latches the D input and the data is shifted out.

Muxed scan flip-flop

LSSD scan

A schematic of LSSD scan flop and waves is shown in the figure below. The relationship between clocks is also shown - the SCK1 and SCK2 are active during shift phase while the CLK is active during the capture phase.

LSSD scan flop

Differences between Muxed scan and LSSD are several - in muxed scan, the data and test paths are the same i.e. the test path is muxed with the functional path and can add additional logic on the functional path making timing closure harder. Muxed scan flops are smaller in area and are faster but the same functional clock is used for shift/capture and can cause shift violations. In contrast, the LSSD scan paths are different from functional paths, have two different non-overlapping clocks for shifting data in and out and one capture clock. LSSD flops are larger in size but the timing closure is easier as we never run into shift violations.

Main disadvantage of scan based testing is that it runs at low speed typically 50 MHz - this is very slow for high speed designs and where at-speed testing is critical.

Share This

Overview

Design for Testability (DFT) is one of the critical requirements in a design - additional test logic is added to detect any manufacturing defects such as opens/shorts on a silicon die. A chip or a die can be tested at various levels ( at wafer level, package level, board level, system level or in the field). By detecting the manufacturing defects beforehand, the manufacturing costs can be lowered.

Various fault models exist to model the defects such as “stuck-at” faults - where a particular node is either stuck at zero or one and a vector should excite the node to it’s opposite value to detect the fault. Delay fault models help in detecting sequential faults that affect timing but not the functionality.

Controllability and Observability are key concepts to DFT. Controllability is the ability to set or reset each and every internal node in the design while Observability is the ability to observe (either directly or indirectly) any node’s state.

For example, one of the requirements for DFT is that all clocks and resets in the design must be controllable from the pins - this involves adding multiplexers with additional test control logic. This falls under the Ad-hoc testing category.

Designs also integrate testability features like traditional scan, Memory BIST (Built-in Self Test) and Repair, Boundary scan (JTAG), Logic BIST and IDDQ to ease testing at wafer/package level. These features are exercised using test vectors that are generated by ATPG (Automatic Test Pattern Generation) tools. Algorithms such as D-Algorithm, PODEM, SCOAP aid the ATPG tool in reducing the number of required test vectors and thus the tester time.

Fault coverage is the measure of effectiveness of test vectors exercised on the design. In most designs, a fault coverage of atleast 98% is mandatory before taping out.

We will look at each one of the DFT features in detail in another post.

Share This