gaom9
Full Member level 4
resistor floating bulk
Hi, I am working in a bandgap reference voltage circuit design.
As the large area needed for the resistor of it, I chose the rpplus (P+ diffused resistor w/o salicide) as the resistor in the bandgap. As the two resistors of this circuit with large values, one is 54K and the other is 640K, the rphpoly resistor will cost a large area but more accurate.
But I met a problem in DC simulation with the rpplus resistor, the cirtuit could not get convergence which result in a wrong result when the three terminals of the rpplus were all connected (the bulk terminal connected to VDD, the bulk of rpplus is NWELL) in some parameters setting but some not.
But when I float the bulk terminal of it (float the NWELL), the circuit works well, and could get a good result without any convergence error.
The cirtuit is correct and the convergence error does not come out when I use the rphpoly resistor instead of the rpplus.
I am worried about the effect of the floating NWELL bulk to simulation result and the layout performance.
My process is TSMC 0.18um CMOS
Can anyone give me some advice, please? Can I float the NWELL of rpplus in simulation and layout?
Thank you.
Best regards!
Hi, I am working in a bandgap reference voltage circuit design.
As the large area needed for the resistor of it, I chose the rpplus (P+ diffused resistor w/o salicide) as the resistor in the bandgap. As the two resistors of this circuit with large values, one is 54K and the other is 640K, the rphpoly resistor will cost a large area but more accurate.
But I met a problem in DC simulation with the rpplus resistor, the cirtuit could not get convergence which result in a wrong result when the three terminals of the rpplus were all connected (the bulk terminal connected to VDD, the bulk of rpplus is NWELL) in some parameters setting but some not.
But when I float the bulk terminal of it (float the NWELL), the circuit works well, and could get a good result without any convergence error.
The cirtuit is correct and the convergence error does not come out when I use the rphpoly resistor instead of the rpplus.
I am worried about the effect of the floating NWELL bulk to simulation result and the layout performance.
My process is TSMC 0.18um CMOS
Can anyone give me some advice, please? Can I float the NWELL of rpplus in simulation and layout?
Thank you.
Best regards!