An Interview with Shahriar from The Signal Path - Quisquous Quivering Quadripole

 

Welcome Shahriar from The Signal Path! If you’re not already, you should definitely be subscribed to the YouTube channel.

• Shahriar currently works in the ASIC design lab that is part of the Murray Hill building at Bell Labs.
• The millimeter wave researchers often say "GHz is basically DC"
• Making higher speed complex devices can be difficult but there was a 2006 paper that had a 81Gb/s timer.
• ST Micro has one of the leading SiGe BiCMOS high speed.
• The ASIC group has some of their research work fabricated at TowerJazz because it is approved by the government.
• The government has been focused on III/V semiconductor technology - InP and GaAs, GaN
• The PA/front end of cellphones/wifi uses GaAs (instead of Si) because of breakdowns of submicron processes.
• Chip designers get design kit from the fab and these are used in simulations. Here is an example of one from TSMC, provided by MOSIS (discussed previously on the show)
• Shahriar usually tests new design kits with a LNA/PA.  He also designed a 160 GHZ PLL as a student.
• 3 um x 3 um capacitor is 12 fF.
• The ASIC designers use 3D EM simulation from Cadence.
• Design kits will have variations: DC characteristics like transconductance (gm) might be ok but the designers know the capacitance will have uncertainty as they characterize the design kit more.
• Dave recently did an EEVblab about what a hobbyist should know. Shahriar agrees and things knowing to test is very important. https://www.youtube.com/watch?v=9iXFhKUa1BU
• Chris loved Shahriar's cryo experiments because they really showed the (extreme) effect of temperature on components. https://www.youtube.com/watch?v=MpCQsB1VK24
• Testings for THz can require complex/custom test gear. Often it is a network analyzer with a front end. OML, VDI make extenders for Anritsu, Keysight.
• One Signal Path episode used a Keysight/Agilent 62GHz oscilloscope, it was on loan from the dealer. https://www.youtube.com/watch?v=dx596o8t_TY
• On scopes you start to run into the noise floor of the ADC (b/c jitter). This limits around 100 GHz. At 110GHZ the connectors become limiting (1 mm). If the connectors go to .8 mm, it could go up to 160 GHz.
• Shahriar has a range of research papers: http://scholar.google.com/citations?view_op=new_group&hl=en&nun=Shahramian&imq=author:%22Shahramian%22&authorid=2011075341937338871
• Peer review is critical for journals. Bell Labs splits research between keeping it private and publishing it.
• There are a range of technical conferences for millimeter wave research: ISCCC, SiRF, SSYCs, IMS conference
• A phased array is a grid of antennas. It allows steering the signals towards the source/receiver.
• Different high GHz bands are great because of the open bandwidth.