LTC3896 doesnt work

  simulation result 

Hi Bulent Kupeli ,

I've moved this to the LTspice forum, where we can examine it for model or simulation problems.

mike

Hi Bulent Kupeli ,

Can you upload your simulation files?

mike

Hi all,

I had a similar problem when I was designing my LTC3896 circuit. Right now, I'm working on designing an inverting buck converter. This will produce a maximum of 5V and 10A of current, and it will be powered by a +24V rail.
When I did the first test with a 10-Ohm load, I saw the same behaviour in the output rail. Your post made me want to use LTSpice to simulate my circuit too, and I got exactly the same graph for the output rail. After doing some research, I looked at the datasheet again and found the LTC3896's REGSD feature. I connected the EXTVcc pin to VOUT, just like you did with your design. This means that the EXTVcc switchover is disabled. The REGSD feature now monitors the EXTVcc pin's voltage and the SS pin's voltage at the same time. After powering up, the SS pin is pulled with 10 uA until the voltage is above 2.2 V compared to VOUT. If the EXTVcc voltage hasn't been switched over by this point, the SS pin will be pulled down with 5uA until the voltage has dropped by 0.2V. If the missing switchover doesn't go away, the regulator will switch off. This will discharge the SS pin with 1uA and keep the switching off until the SS pin voltage is about 200mV. At this point, the controller will try to start up again, and the cycle will repeat.
The datasheet provides a solution to this problem. You can do this by connecting the SS pin with a 330-kOhm or less resistor to INTVcc. This will override the 5uA pull-down current. This means that the regulator will not be disabled. Remember that this feature is there to protect the LTC3896 from overheating. If you override this feature, the chip might get a little hotter. I used gate resistors (8.2-Ohms) to limit the MOSFET's gate currents, and the controller only gets to about 30°C.

Below are the waveforms for VOUT (yellow, CH1) and the SS pin's voltage (green, CH2) of my real-life inverting buck converter. First without a pull-up resistor, and then with one. I measured the voltages with respect to VOUT, this is why the are displayed as positive values.






Best regards

Can you check sense resistor value? I think you may have reached the sense voltage limit.

The sense resistor value is 10mOhm, not that different than yours. I left the ILIM Pin floating such that the sense voltage limit is set to 75mV.

Below there are the startup sequences for different time scales.

Even with an acquisition rate of over 5MSa/s, I cannot see current limiting. The rather high current ripple of about 3A is by design. I am aware that with the current ILIM setting and choice of sense resistor value the output current of 10A is not achievable. The sense resistor I had laying around and will only be used for testing.

The voltages are now measured with respect to system GND.