[MOVED] Iref generation from Bandgap Vref

[mburakbaran]

Hi guys,

I'm designing a DAC where i need current sources. I've previously designed a bandgap voltage reference Vref=400mV. (1.2Vdd, 90nm tech) where sub threshold operation is used. My problem is that my DAC is based on current sources (i need 8 current sources of 16uA each and one current source for the opamp at the end - it is acting like a summing circuit-). I'd like to make this circuit operate in a wide temperature range let's say -120 celcius - 100 celcius. That is why i've used bandgap Vref. How can I convert this Vref into Iref without losing temperature independency? The bandgap voltage reference topology is attached.

Thank you all in advance.

 

[dick_freebird]

Re: Iref generation from Bandgap Vref

Add a buffer amp with a FET source follower that has a
precision source resistor. Close the loop at the resistor so
Vres=Vbg and Ires=Vbg/Rres. Other sources are replicas
of the master, the buffer amp drives all gates common.

 

[mburakbaran]

Re: Iref generation from Bandgap Vref

Add a buffer amp with a FET source follower that has a
precision source resistor. Close the loop at the resistor so
Vres=Vbg and Ires=Vbg/Rres. Other sources are replicas
of the master, the buffer amp drives all gates common.

You mean like the stage in the new attachment sir? What about the internal bias current source of the opamp itself?

Thank you very much

 

[checkmate]

Since you already have a bandgap, it must have its own bias current generator. Just mirror an additional branch for the opamp.

 

[mburakbaran]

Since you already have a bandgap, it must have its own bias current generator. Just mirror an additional branch for the opamp.

Well the thing is, i am trying to operate the circuit in a wide temperature range. Through the range, the current varies as you can guess (Vth temp dependency). I wonder if it is possible to keep the current constant just like the voltage? Thank you very much for your responses.

 

[checkmate]

Well the thing is, i am trying to operate the circuit in a wide temperature range. Through the range, the current varies as you can guess (Vth temp dependency). I wonder if it is possible to keep the current constant just like the voltage? Thank you very much for your responses.

I'm referring to the opamp bias current you mentioned earlier.

 

[mburakbaran]

I'm referring to the opamp bias current you mentioned earlier.

So opamp bias current wont be temp independent... I am not sure i need that too though. Thanks

 

[checkmate]

As long as the open loop gain is high enough, changes to gain attributed to bias current changes will translate to only very small output errors..
Furthermore, your bandgap opamp is working on exactly this "temp independent" current bias too!

 

[mburakbaran]

As long as the open loop gain is high enough, changes to gain attributed to bias current changes will translate to only very small output errors..
Furthermore, your bandgap opamp is working on exactly this "temp independent" current bias too!

Hmm well thanks for the tip. Could you please check the first message on this thread and tell me where will I mirror the current from? For Opamp I mean...

If you say that I'll be taking it from the gate of M2 or M9, I must tell you that the current is not temp independent. The current is summed and then realize the temp. independency...

 

[checkmate]

The current in a bandgap is not temp independent, only the bandgap voltage is.
But as long as the opamp is designed to work within the full current variation range, it's fine.

And from your bandgap, I see I2 and I9 are derived from the same wilson mirror. IR being the sum of the 2, will also share the same temp dependence. In other words, you have no other choice. You either use this, or have to come up with another bias current generator.

And lastly, your bandgap doesn't seem to have a startup circuit.

 

[mburakbaran]

The current in a bandgap is not temp independent, only the bandgap voltage is.
But as long as the opamp is designed to work within the full current variation range, it's fine.

And from your bandgap, I see I2 and I9 are derived from the same wilson mirror. IR being the sum of the 2, will also share the same temp dependence. In other words, you have no other choice. You either use this, or have to come up with another bias current generator.

And lastly, your bandgap doesn't seem to have a startup circuit.

Well thank you very much. The circuit in fact has a startup part. M11 does the job.