Burn-in socket is made by assembling contact pins and housing either manually or automatically. The assembled burn-in socket is then mounted on a PCB by soldering or compression mount. However, current burn-in socket technology hits the wall at about 0.27mm pitch.

New burn-in socket technology is presented for small pitches, up to 0.10mm.

Path to smaller pitch:

Reducing the pitch beyond the present limit requires paradigm shift in the burn-in socket manufacturing process. Traditional burn-in sockets are assembled first with contacts and housing, before mounted to the PCB. For the new burn-in socket, contacts are first soldered to the PCB. Socket housing is then added.

Solder paste is applied to the contact surface before affixing the contact pin to the PCB. Laser bonding will then solder the contacts to the PCB.

Very small pitch can be achieved with this process. The process can be automated, reducing the socket defects and cost.

Benefits of the new burn-in socket solution:

Reduced defect & reduced waste. Traditional burn-in socket would have to be discarded if there is solder bridge (short circuit) or cold solder joint with the PCB during socket assembly. Attaching the contacts to the PCB before adding socket housing allows for pre-inspection and repair of solder bridge and cold solder. Final defect rate can approach zero with the new process.

Saving material. Separator plates are used in traditional burn-in socket, to reduce electrical shorts and ease manufacturability. The new burn-in socket can eliminate the separator plate, to reduce cost. Fast socket deployment. The new burn-in socket system allows new socket development without any high-cost die cast molding. Fast new socket development, without high fixed cost, is possible. A new socket can potentially be made in one day.

As the increasing a requirement of artificial intelligence (AI) and a huge of data, the process chips, such as processor, memory, and integrated circuits (ICs), became a higher level of integration. And it has faced demand on enhanced specifications – a higher process speed, a more compact size, etc.

The probe needles for the wafer test and the socket for package test are challenge points to meet functions – more delicate design and faster data gathering speed between the package and the system board in the narrow pad pitches.

AAO (anodic aluminum oxide) is a ceramic material capable of precise anisotropic etching, which used as a mold for fabricating probe needles. By applying this new material to MEMS technology, precise shapes (e.g spring shapes) even with very narrow gaps can be manufactured vertically straight with desired thickness(~130um).

We would like to introduce new MEMS technology for fabricating precise metal parts such as fine-pitch wafer test probe needles and differentiated spring pins for package test which can respond to the demand for next generation semiconductor inspection.

The presentation introduces Schematic AI Extractor and how it assists the reviewer in reviewing and mining the data from schematic file. A schematic file usually contains lots of interfaces such as audio, display, memory, and so on. For each interface, there are details such as vendor, part number, voltage, technology type, and so on to be extracted. The Schematic AI Extractor can reduce some of the manual labour from the reviewer. The Schematic AI Extractor provides automatic extraction along with AI to interpret the extracted data of the components. Next, the details will filter using a similarity checking character by character with a threshold value method and a master list that contains details of the components. Once everything is passed, will be exported to a custom Excel sheet used by the reviewer.

Dr. Angie See Tien Ng is an AI Principal Engineer from Intel, who is the pioneer for Cloud based Remote Debug (CBRD), and spearheaded AI Chat bot. She has been in Intel for 14 years and in the industry for 26 years. Her career includes AI, R&D electronics design, RF designer, FPGA IC characterization, Electrical Validation, Functional Safety etc. Angie is always passionate in innovation and is honoured with Distinguished Invention Award by Intel. She graduated from Campbell University with BEng in Microelectronics/Physics in 1997, obtained her Master in Engineering from Multimedia University in 2014 and completed her PHD doctorate from University Science Malaysia in 2021 in area of designing Cloud-based virtual learning & testing for industry.

Siaw Yen Ming is a Cloud Application Engineer from Intel who is leading the Intel Penang Site Developer Cloud and involved in various AI initiatives including the AI Virtual Chat Bot. He has been in Intel for 8 years and in the industry for 10 years. (Motorola, Intel). His career includes Technical Support, Design & Electrical Validation, Firmware Validation and Software Development. He is graduated from University of Bradford UK with BEng in Electrical and Electronics Engineering in 2013.