bandgap startup problem

Hi all,

Could someone help me with the startup problem in my bandgap circuit?

I have startup problem for lower temperatures .i.e if i simulate ckt for weak -40.
I get output waveform as attached ..

Thanks in advance
rampat

can you please say whether the circuit gives proper band-gap voltage without start-up circuit(start-up circuit disconnected) at weak -40 degree Celsius?

yes it gives proper bandgap voltage at weak,-40 in OP.But when simulate transient with supply ramp,output does not settle to bandgap voltage ?

-rampat

How much current flows through the starttup circuit - it migh tbe you are not totaly turing it off and it loads your bgap.

could you please say us how you fixed?

I guess as Teddy said either your startup is not completely off or the loop stability parameters (with ota in the loop of bgap) are not proper...(may be bad PM)

I think start-up loop have too big gain. So pls try to reduce some. And pls check start-up loop stability.

Startup needs to learn when to let go. It looks to me like
it is cutting out early at this corner and the main loop is
still too weak to pull it in. Maybe you need to use another
means of determining when the bandgap core is properly
lit, rather than a crude VT vs VBG comparison. Such as,
tapping the guts of the op amp at a place where you have
a clean indication of the gain stage being in the linear range
(hence likely closed loop). This might appear as "not railed"
and involve a pair FETs, pair of loads and a little more logic.
Find your largest signal excursion as you sweep the bandgap
head current with a broken loop. That will be the easiest to
work with. Assuming your op amp starts even when the
bandgap doesn't, which is also a thing to check.