PTAT current reference help

how can i design a PTATcurrent reference in 1.8v supply.The circuit is just the one in the page391(Razavi's "Design of Analog CMOS Integrated Circuits").Thank u!!

Dear copoler,
VDD=VDS(M3)+VGS(M1)+VBE1
VBE1 is a function of quicent current in it. It decreases with decrease in I. So you could chooes proper I. this leads to VBE of 0.5V at nominal temp. VBE has a Tempco of -2mv/C. So for -40 you would hit VBE of .625V
VDS of M3 could be between 0.15 and 0.2
So we are left with VGS(M1)as 1.8-.2-.5=1.1
this must be equal to Vth(m1)+Von(m1).
See that we have things working in supply corner (1.8±10%)
If R is ideal your worst case would be
Temp=-40, SlowN slowP and 1.62V as VDD.
But as R is not ideal you could choose a R that has negative TC so as temp. decreases R would increase, this would decrease Ias I=ΔVbe/R. this would decrease VBE across Diode BJT' and also lower Von requirment for M1 and M3.
I=ΔVbe/R=(VBE1-VBE2)/R
(VBE1-VBE2)= Vt*ln
Vt=thernal voltage=25.2mV for 25C
Hope it helps

Thank u Ambreesh!
But now i have another question,when i .dc R2 with temp sweep,i found that Vout got to be constant with R2/R1*ln<17.6,and Vout equaled about 1.05V.How this happened?
I simulate the circuit by Hspice2004.W.

Added after 28 minutes:

Another question
If i use the circuit of Figire11/18,employ a error amplifier instead of 2 nmos,how i bias the amp?I even haven't a Vref,how can i get the current source?

Regards!

Dear copoler,
I could not get what you are trying to explain to me.
You have done a DC sweep of temperature and what did you observe?
You r using equation 11.37 to get Vref.
You are getting the ratio of R2 and R1 by differenciating 11.37 (u could refer to baker Le and boyce also)
Please elaborate your problem.
As for your second question,
The amp output drives all the 3 top PMOS. Now this also drives another PMOS to which u connect a diode NMOS load. From this mirroe the current into tail current of amp.
Could refer to the attached paper for the same. You shall understand. And feedback controls the bias current.

If you cannot dowmload it you could mail me at ambreeshb@kpitcummins.com

I need ptat too!!!

Message is unavailable.

How to improve the variety rate of the ptat voltage/current?

nuiscet said:

How to improve the variety rate of the ptat voltage/current?

Click to expand...

I have the same question
We design a PTAT circuit, and simulation is fine, current nearly independent of PS.
But the actually chip current is proportional to PS.

bandgap is a difficault design case.