1
\$\begingroup\$

I tried to repeat the scheme for DS and general recommendations for the design of step-down regulators. enter image description here

enter image description here

My scheme: enter image description here

My PCB: enter image description here

It seems everything is designed correctly. But the TPS output gives 0V. (Pin EN floating)

What could I have done wrong?

\$\endgroup\$
  • \$\begingroup\$ Why do you say "Pin EN floating" when I can see the EN pin pulled low on your schematic? \$\endgroup\$ – Kevin Kruse Apr 24 '19 at 15:33
  • \$\begingroup\$ @KevinKruse I did not solder the resistor R30 \$\endgroup\$ – Denisok Apr 24 '19 at 15:35
  • 2
    \$\begingroup\$ What else doesn't match between your schematic and your actual design? Where do "Pow Level 1..5" go? \$\endgroup\$ – Kevin Kruse Apr 24 '19 at 15:38
  • \$\begingroup\$ @KevinKruse "Pow Level 1..5" are floating too. No more differences \$\endgroup\$ – Denisok Apr 24 '19 at 15:41
  • 1
    \$\begingroup\$ Did you try some other output voltage, like Vout=5V? Did you try to put some load, 10-20 mA? \$\endgroup\$ – Ale..chenski Apr 24 '19 at 16:19
4
\$\begingroup\$

You are trying to get 0.8 V out of 24 V input. This is 30:1 ratio, which theoretically requires 1/30 duty cycle for PWM at the fixed switching frequency.

Now, you selected 162 k set resistor, which will set the switching frequency to about 600 kHz, which is about 1700 ns period. At 1/30 duty cycle the up-time of switch should be about 55 ns. The datasheet, however, says that minimum controllable on-time is 135 ns, and this is at 12V input. This time will be even longer at 24 V, since the edge rate is limited too.

Therefore, your design needs 55 ns on-time, while the IC can do only 135+. This design can't output 0.8 V if Vin=24V. Try 3.3 or 5V.

And again, the narrow duty cycles require extreme care in designing high-current switching loops to minimize transients, which needs traces as short (and wide) as possible. Your layout deviates from the recommended one significantly.

Also, are you sure that the SW signal is not shorted to ground around C24 boost capacitor?

\$\endgroup\$
  • \$\begingroup\$ When you try to set the divider 3V or 5V and i try 12V input, nothing changes and the driver still gives 0V. I checked the short circuit on the pin SV, it is not. Magic( \$\endgroup\$ – Denisok Apr 24 '19 at 18:02
  • \$\begingroup\$ You say "layout deviates from the recommended one significantly". Could you advise me what is critical to fix in the design? \$\endgroup\$ – Denisok Apr 25 '19 at 16:05
  • \$\begingroup\$ @Denisok, some traces are longer than they could be, like the switching node SW, plus possible extra capacitive coupling to copper poor. The ground (return path) is weak. Thermal pads should be replaced with solid. It should technically work as-is, but you won't have desired efficiency or max current output. Component selection (ESRs) are very important. Try TI BEBENCH for your parameters and your components, and try to simulate this design. And just follow the suggested layout exactly. \$\endgroup\$ – Ale..chenski Apr 25 '19 at 16:21

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.