bandgap start up testing and consideration

[whlinfei]

Hi All,

I have a bandgap reference circuit, which works fine in simulation. (Both DC temperature sweep and transient sweep give satisfactory results.)

As this is my first fabrication, I am a bit concerned about the start-up and also the stability of the circuits.

startup
the startup circuit I used is the one based on capacitor. such cap is connected to Vdd and a NMOS gate. once the circuit is functioning, the gate of aforesaid NMOS will be pulled down, i.e. discharged the gate capacitance and the capacitor.
It is basically a standard capacitor based startup circuit.

The other concern of mine is the diode. supposedly it is biased around 0.65 or 0.7 So it is on at that point. But I don't know how I can ensure that it is actually on. Because even Vbe is less than 0.7 or 0.6, the diode can still conduct current.

I want to ask if there is any standard tests for bandgap reference and its startup circuit ?
What are the special consideration for them ? what is the usual problem after fabricatioin ?

Thanks in advance .

Rgds, whlinfei

 

[keith1200rs]

The test I usually do to make sure my startup circuit actually works is remove it, then deliberately add a high value resistor to create a leakage which stops it working (assuming it as a potentially bistable circuit). The replace the startup circuit and make sure it starts. In you case I would also find out how slow the power supply rampup can be before it fails to start (with the deliberate leakage added).

Keith