Roberto Rosales

Test Lab Manager

HOME

ABOUT ME

LAB ACTIVITIES

LINKS

 

The SoC Test lab is mostly intended for the test of micro chips designed here at the UBC SoC lab, but is also used for all other kinds of experiments. Since its opening in 2001 we have tested more than 57 analog, mixed-signal and digital ICs. The pictures on this page provide examples of activities in the lab. 

The original setup of the lab for the grand opening in 2001. The lab today looks the same but with more equipment. In the picture above Daniel is working on Sameer’s electro Optic Sampling System.

There are two manual probe stations in the lab. A Cascade RF-1 and a Signatone S-1160 with a hot chuck. For these we have a collection of probes covering the range from DC to 67GHz.

The majority of our ICs are manufacture through CMC scheduled runs. The samples we receive back are either packaged, loose dice, or loose dice mounted on glass slides.

For packaged ICs we frequently design and build our own test fixtures. The IC on the board is a an interface circuit for an ultrasonic medical imaging system.

Typical setup: (Left) setup for a 5GS/s 4bit flash Analog-to-Digital Converter ADC. (Right) Microphotograph of the IC with the probe tips in contact with the IC PADS.

Some students like Shahrzad (smiling on the left) and Andy (sitting on the right talking to Jim) are assigned permanent work areas in the test lab. Most other students have their desks in the cubicle areas of the group in rooms MCLD332 or K4025. 

The picture on the left shows the instrumentation used to test a self-timed interface IC for crossing clock domains. The circuit under test provides nearly two clock periods of skew tolerance and operates correctly for any initial phase offset between the transmitter an the receiver. The picture on the right shows the IC under test on the probe station.

The SoC test lab is one of the main attractions in our department for promotional photo shoots and visits. On the left is Nima posing with a heat chamber and his high temperature resisting sensors for pulp and paper. But some times things get really messy … on the right is Jihong trapped by the mess in a corner.

Lately we have been experimenting with die on board as a test strategy. On the left is a test setup I recently made for Neda’s 60GHz VCO. A loose dice is mounted and wire bonded directly on a PCB (right) to provide for the DC bias and low frequency control signals, while the 60GHz output is accessed with a GSG probe.

When everything goes well we cannot resist posing for pictures by the test setup. Above, Michael (left) is testing his on-chip termination circuit using active inductors and Samad (right) one of his flash ADC converters.

Back to Top