Other Parts Discussed in Thread: TLV172, TPS60400
With LM2776 I want to feed the negative supply of two TLV172 from +3.3V. If 3.3V is fast rising everything works well but at slow rising supply the LM2776 starts too late and the current trough the op amps creates a small positive voltage at VOut of LM2776. The chip cannot start properly in this context, sometimes there ist about -1.7V with very large ripple of some kHz at VOut, in other trials there is just +0.6V (one pn drop). The forum shows several flavors of this problem (e.g. temperature dependency, other kinds of latchup problems). I tried several combinations of pump and output capacitors and to my surprise a stable circuit was established with a far too small pump capacitor of 100nF while keeping the 2.2uF at the output. This configuration starts up properly with any supply rise time and op amp activity. This solution cannot feed large currents but I think most op amp circuits are not that hungry anyway. The LM2776 works in some kind of burst mode with this solution and therefore the output ripple is a bit ugly, but finally 10uF at VOut and one additional RC filter worked for me. I have never seen this problem with other inverting charge pumps in the past. Can someone explain this effect and it's boundaries?
