Part Number:TPS2400
Dear TI group,
I am evaluating the SLVA163 application note, fig. 3, as a basic scheme to obtain UV and OV protection for my circuit with TPS2400.
With nominal input voltage of 24Vdc, Cout=680uF and Rout=360R as output load, I set:
- R1=100k, R2=30k to obtain UV around 15V and OV protection around 32V;
- Q2= Nexperia PBHV8540Z (500V 0,5A npn);
- R3=1k, C1=10nF, to obtain an inrush current around 260mA;
- Q1= Diodes DMN10H120SE (100V 3,6A nMOS), low-side configuration.
At startup, the problem is that the circuit does not limit the inrush current as expected. Most of the times, the dV/dt on Cout is either fully limited by the power supply impedance, or split into two parts: one limited by my power supply, the latter effectively by R3 and C1 values. Very rarely I observe the correct dV/dt since the beginning. I also played by removing Cout and powering the circuit abruptely, but nothing changed. The situation seems to rectify only by adding a 1uF capacitor between the base of Q2 and GND, but of course this is not a solution, because it would slower the response time of the OV protection.
It seems there is something wrong with TPS2400 at startup, that pulls the gate of the mosfet slightly above its threshold value, before being absorbed later by the Miller effect (provided by C1). As a result, the output capacitor is quickly charged to a fraction of the input voltage, the remaining part being completed at the desired costant current value. To be more consistent, I tried several other values for C1 (up to 470nF) and R3 (up to 10k), but the problem appeared anyway.
Now, I temporarily solved by adding a open drain, active low voltage detector to keep the TPS2400 output low till a voltage slighly above the UV threshold. In this way, the TPS2400 starts with the output shorted to GND and the (undesired) peak current sourcing from this pin at turn-on is diverted away from the gate of the mosfet. But I don't like this solution for the number of extra components required and because I don't know how long TPS2400 can stand with a short on its output, despite it has a low current source inside. Another solution could be removing (C1, R3) and using Q1 to drive a load-switch like Infineon ITS4140N to perform the inrush current limitation. But I still expect TPS2400 to manage everything on its own.
Is the above behaviour normal? What do you suggest to do?
Thank you in advance,
Simone