Is it possible to design or simulate a 28 GHz of LNA by transistor using ADS 2022 version, i have been searching so far ,i didn't get the exact transistor, i am searching transistor with low noise capability which is available in ADS library.
Currently, we are no longer using ADS, only Cadence for CMOS RF design. To my best knowledge, CMOS manufacturers usually do not provide design library kits for ADS, only for CMOS design tools like Cadence, Synopsys, etc...
In case you want to design a discrete LNA you can find on the market encapsulated transistors for which manufacturers provide s-parameters plus noise parameters for very high frequencies. Probably, in some cases they provide a spice model. I hope this helps.
João Caldinhas Vaz Thank you for your response, Sir. May I know the reasons sir why designers choose Cadence rather than ADS for chip design. ADS also work on high frequency platforms. I am working on ADS and to make LNA design using Ka band, is there possible to get transistor from ADS to build LNA simulation.
Dear Arivazhaki, the main reason is the CMOS layout design phase. Design rule check (DRC), layout vs schematic (LVS) and parasitics extraction require design rules that should be provided by the CMOS manufacturer. Usually they provide these for Cadence, and not for ADS. In the past I used for a CMOS technology ADS, but the manufacturer only provided models for simulation phase, not layout phase.
Dear Arivazhaki Dravidaselvan , yes it is feasible, you have two options for the LNA:
1) CMOS implementation. For this we need the CMOS Foundry process design kit and the proper design tools to use it (probably cadence).
2) Discrete implementation on a PCB board. We buy package transistors, the passives will be mostly distributed elements like transmission lines, stubs, that you design on a proper PCB board for Ka-band. ADS is the right tool for this case.
You need to work with a foundry that provides a PDK (process design kit) for ADS. Start there and find the appropriate foundry. Unlikely that a lumped element design properly modeled with distributive affects will work out. I have done lumped element LNA's at 10 GHz using pHEMT devices and this worked out well. However, it does require extreme care and includes co-simulation with an EM tool.