This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

TPS92515HV: COUT with shunt dimming?

Part Number: TPS92515HV
Other Parts Discussed in Thread: TPS92661-Q1, TPS92515

Dear TI,

I would like to use the TPS92515HV with the TPS92661-Q1 to provide 700mA and fast dimming through the matrix manager chip.

I used Webench to produce a preliminary design, but am confused on several points when I attempt to make it consistent with the datasheet and my usage.

1. As I understand it, the output capacitor should dump significant current into the non-shunted LEDs when one is removed from the chain. I do not believe that Cout should be there for shunt dimming, and my inductor ripple is already within the specification I want for my LEDs by a large margin. Does Cout provide functionality beyond reducing output ripple? From the datasheet it appears that if I can tolerate a current ripple in my LEDs equal to the current ripple in the inductor that it can be removed, but I am unsure if it may impact the current supply stability.

2. Since this design would requirer no external PWM signal to the TPS92515 pin, the datasheet says to simply connect it to VCC. In the Webench design it is instead connected to a resistor divider with Ruvlo1 and Ruvlo2 with Cdim. I believe this is to implement undervoltage lockout, but this did not appear to be possible to disable in Webench. In an application with 48V at Vin and an absolute highest output voltage of 42V with typical operation at 36V or less, is this necessary?

3. I would like to provide a potentiometer to provide analog dimming. In this way I would like to make it so that you can use a knob to turn down the brightness while the shuntFET array in the TPS92661-Q1 provides PWM dimming independently. In this mode, I believe that I can simply remove Riadj1 to connect IADJ with Vin and instead provide a voltage between 0-2.4V or so. Is this what is intended?

4. For the TPS92661-Q1 it claims 225 mohm per switch and I plan to use four of the provided switches. To calculate Roff2 I assume that all four switches are conducting and so the Vshunt is 0.63V. Webench calculates an inductor ripple of 63.5 mA, an inductor of 82 uH, a Coff of 470pF, and as I understand it Vin is always 5V. With these values I calculate Roff2 of 37.4k. Please let me know if I am missing something, I could not find a Vf for the TPS92661-Q1 other than the Ron.

Thanks very much for your help,

  • Hello Brian,

    Webench provides a decent starting point, but it doesn't take everything into account or have full functionality to change things. But hopefully I can answer your questions here:

    1. The output cap is just for reducing current ripple in the LED. It isn't required and in fact one should not be used with shunt FET dimming. As you suspect the peak currents could get very high in both the LED(s) and the shunt FET. This device was designed to run without an output capacitor for the purpose of shunt FET dimming.

    2. It is a UVLO function and it is not required. If you don't need it you can just tie PWM to VCC as mentioned.

    3. Yes, IADJ can be driven directly with a voltage.

    4. That is correct, you calculate using the current and total Rds(on).

    Regards,

    Clint

  • Fantastic, thanks Clint!