TIDA-010042: TIDA-010042 MOSFET BASED DESIGN- Is there any issue if I test MPPT code with DC Power supply

Hi Cyril,

Does second phase not be damaged? I think the converter will first enter MPPT mode and then after some time, if the current is higher than CC limit, it will go to CC, or CV mode. I guess the initial 30A current is too high for the FET, did you use 2 phase or one phase for this current capability? Does the phase 2 is running at the beginning?

You can try with lower current limitation on the power supply to make sure the FET is not overload.

Thanks.

Bowen

Hi Bowen,

In my design, TD1 ie, second phase will be ON when Battery current is greater than 3A. If battery current is less than 3A, only first phase will be ON ie, (TD0).When current reaches above 3A, 2nd phase will get turned ON. In my case, TD1 phase was damaged.

Thanks,

Cyril 

Hi Cyril,

I see, which MOSFET do you use in the system? And how long time after the detection of >3A current, the second phase will start?

One reason i can imagine is the current for first phase is too high, the second phase is not yet started, then too much heat in the first phase caused the damage, you can also try to start the two phases at the same time to exclude this possibility.

Thanks,

Bowen

Hi Bowen,

I am using bsc093n15ns5atma1-infineon mosfet..

After 0.2ms of detection of 3A current, the second phase will get started. I was trying to implement a regular battery charger from an input dc power supply and not from Solar.I have omitted mppt.

Is there any problem for implementing it based on TIDA-010042 design. Is there any limitation like response time while working with regulated power supply. In my design adc reading is taken every 0.2ms.Does response time of 0.2ms is good or bad.

Regards,

Cyril

Hi Cyril,

For implementing MPPT, i think 0.2ms is not slow. But for protection from OCP/SCP or implementing PI loops, 0.2ms is too slow.

For the MOSFET you used, it has very big current capability. So 30A current in very short time should be fine. Have you tried with a lower current limit of power supply to ensure the input current is low?

Thanks,

Bowen

Hello Bowen,

I am using meanwell powersupply CSP-300-120. In this I can either set voltage or current at a time. Therefore I set the voltage to 90V. Maximum current of powersupply is 30A.

In my code, I am starting duty from minimum value and  limited the upperthreshold of duty cycle. Gradually I increased the upperthreshold and I have seen a maximum charge current of 18A. At some point after programming the microcontroller to increase the upperthreshold, charger was not able to charge and when I checked the circuit I seen that half bride driver's low side part is showing low impedance ie, VCC to COM and VCC to LO(mosfet inside the driver is showing impedance of 200ohm).Please see the functional block diagram of IRS21867STRPBF of driver attached.

Also 2.21 ohm resistor(2.21ohm)  inbetween SW node of buck and Vs of driver was showing high impedance.

When I checked the gate to source of high side and low side moset,by removing it from circuit, it was showing open circuit(Drain to source of fet is not shortcircuit and body diode of fet is fine).

Regards,

Cyril

Hi Cyril,

It maybe the gate driver is damaged. Is this gate driver able to handle this scenario?

Looks like this is more like a hardware issue. Maybe you can try with a higher current capability driver.

Thanks,

Bowen

Hi Bowen, 

Thanks for supporting me. I have been using this gate driver for mppt operation and it was working well. I have not faced any damage for the driver till now with Solar Panel.

Regards,

Cyril

Hi Cyril,

I suspect this could be a hardware problem, is the bootstrap not damaged or only the inner driver FETs inside the driver is damaged?

Thanks,

Bowen

Hi Bowen,

Bootstrap capacitor is not damaged. Only the inner driver FETs inside the driver is damaged.

Do you have any suggestion regarding the MSP430f5132 for this design of charging battery from DC powersupply. Since we are using 182kHz switching frequency and we are responding to changes only after 0.2ms ie, around 36cycles of pwm, its reaction time is very slow right?.

What should be the minimum response time(control loop timing) for switching frequency of 182kHz. I have less experience in this field thats why I am asking.

Regards,

Cyril

Hi Cyril,

I believe if you sample every 0.2ms(5kHz), there's not so much you can do in software to improve the dynamic performance.

For 182kHz switching frequency, you can do loops at 15kHz if possible. But I am not sure if this device is able to support this loop frequency.

And put some hardware protection for OVP, OCP if possible.

Thanks,

Bowen

Hi Bowen, 

Can you tell me how you suggested 15kHz. If MSP430f5132 is capable, should I go higher than 15kHz or is there any calculation related to your suggestion of 15kHz for 182kHz switching frequency

Regards,

Cyril

Hi Cyril,

15kHz is my rough estimation. If the MCU is capable of doing this, you can do as fast as you can to improve the dynamic response of the converter.

Thanks,

Bowen

Hello Bowen, 

Can you tell me how much should be the adc sampling rate required for microcontroller for 182kHz switching.

Regards,

Cyril 

Hi Cyril,

The ADC sampling rate highly depends on how fast you want to control and protect the system.

You can use the same ADC sampling frequency as control frequency, maybe 15kHz.

Thanks,

Bowen

Hi Bowen, 

Based on Nyquist criteria, sampling frequency should atleast two times greater than the switching frequency right?

Why in TIDA010042, this is not considered.

Is it because power source is a Solar panel?

Regards,

Cyril

Hi Cyril,

I think the switching frequency current/voltage component is not what you want to control or observe in the loop.

So you don't need to implement sampling frequency twice as the fsw.

Thanks,

Bowen