regarding bandgap reference

[manissri]

how the bandgap reference voltage is independent of supply voltage.
as it is indepentdent of temperature.

 

[abcyin]

I think that what you mean is the PSRR of the bandgap reference, the method I know is to use cascode current mirror to shield the VDD variation.

 

[monte78]

Could you please clarify your question?

 

[Vamsi Mocherla]

The bandgap voltage is actually is captured as a sum of I*R and a VBE. The PTAT current into a PTAT resistor will capture a PTAT voltage. Now, VBE has a CTAT voltage. The bandgap circuit mixes both of them to get a temperature independent voltage reference, which equals energy bandgap of silicon at 0 K.

Now, the bandgap equations will show an independence of the VDD, as there is no term from VDD contributing to it. This is true in all the cases as long as the current generators have very high impedances. This usually happens in the saturation condition. Anyways, PSRR is determined by how deep these transistors are in saturation.

 

[leo_o2]

If the current thru bipolar, the BG will be independent from supply voltage.

 

[manissri]

i am using the low voltage bgap.
where i connect my pos and neg terminal of ERRor amp to the circuit.

 

[leo_o2]

If you use 1:8 bipolars, Pos is connected to 1 bipolar and Neg is connected to 8 bipolars.

 

[manissri]

what will happen if i connet the reverse order.
will it become the postive feedback if so then plz explain

Added after 5 minutes:

i just want to know the error amp polarity of pos and neg termnal is where to conect and why.

 

[stoned]

what will happen if i connet the reverse order.
will it become the postive feedback if so then plz explain

Added after 5 minutes:

i just want to know the error amp polarity of pos and neg termnal is where to conect and why.

For this diagram, the feedback coefficient is:
βn=(1/gmQ2)/(1/gmQ2+1/gm1+R2)
βp=[(1/gmQ1)+R1]/[1/gmQ1+1/gm2+R1]

make sure the total feedback is neg, the βp must be smaller than βn, the better choice is βp≈(1/2)βn, and the loop will have good tran. response when the load is large cap. .

Stoned

 

[manissri]

i am wrk on low voltage bandgap circuti
for error amp i am using the two stage ota( with single pole )
first stage is gm1/gm2 and second stage is common source amp
for this architeucture is am getting only 40dB gain
so is this gain is sufficient for error amp.
if not so can i go for two stage miller amp ==> but i doubt that for error amp this architecture can i use (as the stabiltiy problem may arise due to miller cap and resis)
please comment

 

[manissri]

please comment

 

[caosl]

low voltage??how low? this bandgap circuit's power supply at least larger than bg voltage itself.
surely, the two stage miller compensated opamp canbe used.Why you have that doubt? the opamp is used to regulate pmos' gate,
I think 40dB gain is not enough to supress the power supply noise, ie, PSRR is not high enough. Maybe in you spec, it can be torlered.