why poly resister value cannot be large?

[shanmei]

poly resistor use high doped poly to achieve resistance,
so if we want large value resistor, why not use light doped poly,
but indeed, large value poly dosenot come into real.

 

[ankitgarg0312]

in 130nm tsmc , they have poly resistors with salicide and without salicidation...

with salicide -> resistance decreases
without salicide -> resistance increase ...


have a look on below thread ..

 

[varunmjman]

If we use a light value poly to make a large resistor then the length u need will be enormous....

 

[kwkam]

the problem is not the resistance, the problem is to control the variation, matching, ......

 

[dick_freebird]

Lightly doped or undoped poly will indeed be highly variable
(process, temperature). However sometimes you really
need an isolated resistor.

Poly can be more opportunistically routed and can be drawn
at much narrower widths than well resistors. So you may find
that it is not so terrible a raw layout density penalty as the simple
ohms/sq number might indicate (you could get, say, 4 legs
serpentine 0.2um poly in the same width as a 1um well
resistor and the #squares might be 20X in the same area).

 

[snafflekid]

There is usually an option for hi resistance poly, around 3kohm/square.

 

[bill66656]

There is usually an option for hi resistance poly, around 3kohm/square.

Agree, really I have seen that in about 2Kohm/square in 0.5um process. While in 0.18um logic precess, it disappears.