As i know there are three main different bandgap references:
1-BJT core Bandgap reference
2-Fully Mosfet bandgap reference
3-Mixed BJT and MOSFET core Bandgap Reference
I wanna know which of these three are better for designing a sub-1v and up-1v bandgap references with less temprature coefficient for output voltage(less ppm/c)?
Or better i ask can we judge these three in this way?
by accident, I was also trying to ask the same question. I already finished the design of many band gap reference voltage and current based on the parasitic bipolar transistor introduced by the CMOS technology. the only problem of this design is they have large area occupied by the bipolar transistors as well as the resistor. therefore, you will save much of area in case of fully CMOS bandgap.
but here is my question, does both has the same quality??. I was about to post this so may be who answer you will answer this. attached is a two pictures of CTAT circuit, in the first one is based on MOS and the other on the bipolar transistor. What is the actual or real advantage and disadvantage between them ????????
As my experience helps designing an up-1v BGR without using BJT for low ppm/c output voltage is not possible , but for sub-1v BGRs (around 0.1v to 0.3v) it is possible to design Fully mosfet with low ppm/c for output voltage.
Now about your question, As i have mentioned the CTAT core which is fully mosfet is more suitable for ultra sub-1v designs arond 0.1 to 0.4. so the left photo is more better for sub-1v design (if PTAT core is also Fully Mosfet).
But As i have got from my experiences for designing a 0.8 to up-1v BGRs BJT core is much better (BJT CTAT core + BJT PTAT core). for 0.4v to 0.8v a mix BJT and Mosfet is better (BJT CTAT core + MOSFET PTAT core OR MOSFET CTAT core + BJT PTAT core). for sub 0.4v Fully MOSFET is better (MOSFET CTAT core + MOSFET PTAT core).