Hi, I'm a mostly-digital guy who knows next to nothing about BJTs. However, lately I've started learning about current-mode logic, which has piqued my interest in BJTs.
In what appears to be **broken link removed** from Rabaey/Chandrakasan/Nikolic's book, they write:
the BiCMOS gate inherits one of the most important deficiencies of the bipolar technology: built-in voltages such as VBE(on) do NOT scale.
As lithography technology progresses, the features sizes of both MOSFETs and various BJTs (Substrate-Well-Diffusion, SiGe HBT, etc) decrease. In the case of MOSFETs, this tends to decrease the acceptable operating voltage (although other problems like gate leakage arise). It seems like this does not happen for ordinary BJTs made out of diffusion-well-deepwell structures in CMOS processes, nor does it happen for bipolar transistors in BiCMOS processes.
However there are other BJT technologies like SiGe.
Is it universally true that the voltage parameters of a BJT do not scale with feature size, regardless of the type of BJT? This would seem to imply that the MOSFET-BJT power consumption gap is only going to get worse as time goes on.
In CMOS technology mosfets threshold voltage depends to a gate oxide thickness, doping profile and channel length so its changes from technology node to node.
In bipolar transistor, a base-emiter voltage depends only to semiconductor material which made a junction - for silicon its around 0.6V, for germanium 0.4V, for schotkeys junctions 0.3V.
In bipolar transistor, a base-emiter voltage depends only to semiconductor material which made a junction - for silicon its around 0.6V, for germanium 0.4V, for schotkeys junctions 0.3V.