bandgap current problem?

[iamxo]

as in the pic, we know that when the current through R2 R3 increases, the negative terminal voltage of the opamp increase quickly than the positive terminal, that is

I*R1+Vt*ln(I/nIs)>Vt*ln(I/Is), here neglect the Is mismatch and current mismatch.

so we can get the current should satisfy the following condition:

I>Vt*ln/R1,

so is there this limit of current value in real bandgap circuit?

please help, thanks.

 

[winsonpku]

In fact, the current flows through R2 and R3 is almost same
and the value is I=Vt*Ln/R1 ----(1)
the pic is a closed loop, and the loop is stable if those values are proper

So the current is only the same as equation (1), no limit as you said.

as in the pic, we know that when the current through R2 R3 increases, the negative terminal voltage of the opamp increase quickly than the positive terminal, that is

I*R1+Vt*ln(I/nIs)>Vt*ln(I/Is), here neglect the Is mismatch and current mismatch.

so we can get the current should satisfy the following condition:

I>Vt*ln/R1,

so is there this limit of current value in real bandgap circuit?

please help, thanks.

 

[pfd001]

what kind of current limet you want to know?

 

[iamxo]

In fact, the current flows through R2 and R3 is almost same
and the value is I=Vt*Ln/R1 ----(1)
the pic is a closed loop, and the loop is stable if those values are proper

So the current is only the same as equation (1), no limit as you said.

so, what do you mean of proper value?

what kind of current limet you want to know?

I want to know whether the current has a minimum value to insure the bandgap circuit work properly.

 

[engronger]

when the bandgap work stable , the current is the equation you give
but ,i think you can not say it is the limitation

as in the pic, we know that when the current through R2 R3 increases, the negative terminal voltage of the opamp increase quickly than the positive terminal, that is

I*R1+Vt*ln(I/nIs)>Vt*ln(I/Is), here neglect the Is mismatch and current mismatch.

so we can get the current should satisfy the following condition:

I>Vt*ln/R1,

so is there this limit of current value in real bandgap circuit?

please help, thanks.

 

[edwintsu]

what kind of current limet you want to know?

I want to know whether the current has a minimum value to insure the bandgap circuit work properly.

U can get the curve of BJT beta value vs. collect current by simulation, as below.If the collect current is too small or large, the beta will be changed. For the stable bandgap reference, u need keep the beta value level with different temp, corner and power supply. Typical i use 1uA at .5u process.

B/R
Edwin

 

[iamxo]

thank you very much, edwintsu.

I will try what you said.

 

[winsonpku]

I think you doesn't understant the principle of the bandgap.
You will find the anwser if you read Chapter 11 of razavi's book.

what kind of current limet you want to know?

I want to know whether the current has a minimum value to insure the bandgap circuit work properly.[/quote]