WEBENCH® Tools/TPS53319: tps53319

Albert Dibelka

Part Number: TPS53319

Tool/software: WEBENCH® Design Tools

I need a circuit that can supply 10A @ 3.5V to an LED. Using the EVM board and modifying the feedback voltage (R10 and R12 also R23 and R24) to match my LED (3.5 V) I was able to get 10A of current at 3.5V to the LED. I than used webench to create a circuit, but that circuit did not work. The circuit can supply no more than 1.5A before it stops (shouts down). It will restart if the load is decreased, but it will not sustain anything above 1.5A. The major difference between the EVM and the Webench design is the inductor. the orginal EVM board has a .5uH inductor, but the Webench desing used a 1.8uH design. Can you help me figure out why the Webench design fails?

Thanks,

al (619) 838-9757 txt or voice

over 7 years ago

0 Harish K R over 7 years ago

TI__Intellectual 2320 points

Hi Albert,

Could you please let us know your input/output [Vin = ??, Iout=10A, number of series LEDs = ??, Forward voltage of each LED = ??] requirements to supply an LED so that we can assist you for your application ?
Also, TPS53319 is a voltage regulator whereas to drive an LED, we would need constant current regulator/LED driver.

Thanks & Regards,
Harish

0 Albert Dibelka over 7 years ago in reply to Harish K R

Prodigy 20 points

The nominal input voltage, Vin is 12VDC, but it can vary from 9 to 15V. There are three white Cree LED's in parallel and each LED draws about 3A at 3.5V. I understand that normally an LED driver is constant current, but since the EVM board worked so well setting the Vout to 3.5V, I was encouraged that it didn't need to be constant current. The major reasons I choose the TPS53319 part is efficiency, small size/parts count, including the inductor. The Webench design was very similar to the EVM, but used a 1.8uH inductor which is smaller than the .5uH inductor on the EVM. I think, but can not confirm that the TPS55319 is shutting down because of current, and not LED voltage. The voltage on the LED's is set to 3.5V, but at 1.5A (total for all three), it drops to 3.4V, still not enough to cause UV shut down. The output cap given by Webench is a single .1uF ceramic, but the EVM had three .1uF caps which I left in place. Also the ripple injector across the inductor was different from the EVM, but the equations seem to support the modified ripple injector values. Could either the output cap or the injector values cause the part to stop working at 1.5A?

0 Harish K R over 7 years ago in reply to Albert Dibelka

TI__Intellectual 2320 points

Hi Albert,

The EVM board for TPS53319 is optimized for Vin = 8-20V, Vout = 1.5V and Iout = 0 to 14A whereas you are looking for Vout = 3.5V.

Usually the inductor is designed in order to get peak-to-peak inductor current ripple of 30% of Iout. Hence for your condition [9V-15V to 3.50V @ 10A], you are getting inductor of 1.8uH instead of 0.5uH. Also I was able to get regulated Vout of 3.5V when I run electrical simulation in WEBENCH. Below is the snapshot FYR.