TPS51601A: Power management forum

Another question - my application will leave the TPS51601A in a static state for long periods of time (hours or days) in either high or low state.  What must be done to maintain the boost voltage?  The datasheet doesn't mention anything, except for the boost FET being there.  Any further info is appreciated.

- SteveP

Hi Steve, 

1. Usually high voltage spike or ringing occurs on switch node because of bad layout. It can also occur when high-side FET Cgs is very small and the high-side FET is turned on quickly. Both R1 (boot resistor) and R2 (gate resistor) in datasheet figure 19 helps to slow down the turn-on of high-side FET, hence to reduce voltage spike/ringing on switch node. 

I would suggest to add placeholder for R1 and R2. The initial suggested value is R1=2.2ohm, R2=0ohm. 

2. I'm not familiar with TEC application. What's special requirement in TEC application?

3. TPS51601A can not maintain the boot voltage if high-side FET keeps on for long time. This could be a problem for your application. 

Thanks

Qian

Thanks for the R1/R2 info, Qian.  

A TEC (ThermoElectricCooler) is mostly resistive, but likes a steady current to operate efficiently.  So an H-bridge is a nice solution to control the current in a switched manner. (A motor controller can be used if it fits the application, but the internal FETs usually don't handle the steady-state 3-5A current so external FETs work better.)  To change the heat flow direction one changes the direction of the current (i.e. swaps the voltage across the TEC).  So a dynamic H-bridge on one side of the TEC driven by a PWM signal controls how much current flows (i.e. the heat transfer rate, therefore the temperature of the target side of the TEC) and a static H-bridge on the other side controls the direction of heat flow (i.e. heating or cooling).  A SW loop regulates the PWM signal to reach and hold the target temperature.  An LC filter changes the H-bridge output into a steady current, pretty much like a buck DC-DC converter, but without the voltage regulation.  However, the LC filter causes large current variations in the H-bridge, which depends on the switching frequency.  But the switching losses are higher at higher frequencies.  So one side is an actively switching H-bridge, but the other side is mostly static, hence the boost voltage issue.

Is there any info on the minimum switching activity needed to maintain the boost voltage and what factors are involved?  I imagine it depends on the leakage of the boost capacitor, but I'm not sure if that's a significant part or if other factors are more important.  

Is there more info available on the Skip input function and use?

I found info and a solution in SLVA444 so this thread can be closed.   It would have been nice if the forum had referred me to this App Note.

Hi Steve, 

Thanks for sharing what you found.  This information will be useful for other customers. 

Thanks

Qian