SM72445: SM72445 panel mode with UCC27712 gate driver

Hi Andras,

thank you for using the E2E forum.

The SM72445 is just providing the digital control signals for the H-Bridge Driver. So it depends on the H-Bridge Drive on how the high side gate signal can be maintained when not switching.

May you can open another thread with the UCC27712 as main part so that it gets assigned to the right supporting team for that device if you would need more details on the UCC27712.

Best regards,

 Stefan

Hi Stefan,

As you can see the picture below the bootstrap supply needs periodic on signal from the low side mos to recharge itself and it only depends on the driving signal because I do not think that there is any intelligence inside UC27712 that knows that we want to keep on the high side fet indefinitely:

Although I found some info on panel mode in the datasheet of SM72445 when we use the normal FETs for 100% (important especially the last two sentences):

"Three different frequencies for the PWM operation of the H-bridge as well as two different implementations of the Panel Mode switch can be set on the ADC input channel 2 (pin A2). The table below lists the different conditions that a user can select on pin A2. Each frequency has a different associated dead time for the operation of the synchronous switches. When dedicated PM switch modes are used, the unit will stop switching the converter upon entering PM mode and the PM_OUT pin will switch at a high frequency to provide activation of a dedicated Panel Mode switch.

When the H-bridge modes are used, the unit will keep the H-bridge switching at half the operating frequency (to reduce switching losses) and with a total input to output ratio of 1. The dead times are unchanged during this phase. "

I will also ask the same question with the part number of the gate driver but in the meantime if you have someone else who might know the answer please let me know.

Thanks and regards,

Andras

Hi Stefan,

I also see the following sentence in the datasheet of SM72445:

The SM72445 does not require a dedicated switch to implement Panel Mode. The four buck-boost switches can
be controlled to implement PM. A dedicated switch may be used for higher efficiency. Setting the voltage on pin
A2 selects between the options.

Based on this and the previous red sentence I would like to know how the four switches of the SM72445 work (what is the switching algorithm) to achieve panel mode that does not hit the 100% duty of the UCC27712 gate driver.

Thanks again,

Andras

Hello Andras,

In PM, the PM_OUT creates a 440kHz signal, would it be possible to use this signal to refresh the driver?

On the other side, the device is only for 60s in PM and then goes back into MPPT mode to check if the tracking is still where it should be. Normally I expect at least some switching cycles for all 4 switches during this time.

Best regards,
Brigitte

Hello Brigitte,

The SM72445 datasheet states that no dedicated PM switch is needed for panel mode and I can use the h-bridge as well so I would take advantage of this .

I was asking about the panel mode since the UCC27712 has bootstrap supply on the high side mosfet which require certain amount of switching on the low side to recharge the bootstrap cap so I would like to know how the panel mode switching is commanded from SM72445 to make sure that the high side stays on? Do not you have any measurement taken where panel mode is active and you can see the gate drivers from SM72445?

I found a thread where Youhao Xi has replied to Feng Xu about the different operation modes:

https://e2e.ti.com/support/power-management-group/power-management/f/power-management-forum/762432/sm72445-the-mppt-operation-of-sm72445/2826601#2826601

In his first answer he says this:

"The controller sense the input and output voltage to decide the operating mode.  For instance, the MPPT voltage is found to be Vin1, and your output voltage is Vo.  If Vin1 >Vo, then the controller will operate in buck at a duty cycle of Vo/Vin1.

If Vin1<Vo, the controller will operate in boost at a duty cycle of (Vo-Vin1)/Vo.

If Vin1 is close to Vo, the controller will operate in the buck-boost mode, at appropriate duty cycle. "

Understanding the red line means that no matter how close the two voltages (Vin and Vout) the SM72445 will perform switching.

Again I am only asking this because the there is the bootstrap can that has to be refreshed not to hit UVLO for the high side gate drivers.

The other thing is that you need to enable a high side of the other wing of the bridge that is not in operation. For example when you do bucking you have to enable to high side of the boost side constantly, right?. How do you achieve that? Or you just leave it for the body diode conduction?

Thanks and regards,

Andras

Hi Brigitte,

Understood, thanks!

Andras