LM2596 buck converter

Build and break

1. 1BuildNext up

   Strip to the bare board

   A blue 66 x 34 mm board with four M3 holes. Input on the left edge, output on the right, both with the plus terminal at the top. The big pad in the middle-left is for the regulator's tab: it is ground and the heatsink.

   Do it
2. 2

   Place and reflow the SMD parts

3. 3

   Fit the trimmer and the display

4. 4

   Fit the capacitors and terminals

5. 5

   Solder the through-hole joints

6. 6

   Wire the pack to IN+ and IN-

7. 7

   Put the meter on the output

8. 8

   Switch the pack on

9. 9

   Switch the display to OUT

10. 10

    Turn the trimmer down to 5 V

11. 11

    Power off, then wire the motor

12. 12

    Power the motor

13. 13

    Turn it down to 3.3 V

14. 14

    Ask it for 12 V

15. 15

    Break it: pack backwardsBreak it

16. 16

    Break it: C1 in backwardsBreak it

17. 17

    Break it: cold joint on the trimmerBreak it

18. 18

    Break it: run the pack down to 5.5 VBreak it

19. 19

    Break it: hold the wheelBreak it

20. 20

    Break it: short OUT+ to OUT-Break it

21. 21

    Break it: a 44 V inputBreak it

22. 22

    Exploded view

Each fault is one that catches real builders. Trigger it, read the symptom on the bench, then fix it.

• Pack connected backwards

  reverse polarity

  Trigger
  Why it happens

  Symptom: The display stays dark, the output is dead, the regulator gets hot fast and C1 may hiss. Often the LM2596 fails as a short, so the module is gone even after the leads are swapped back.

  With minus on IN+, the protection diodes inside the LM2596 and the meter's regulator conduct straight across the pack, and C1 is reverse-biased. A LiPo can push tens of amps through that. There is no diode, fuse or MOSFET on this module to stop it.

  Fix: Swap the leads. Next time, a keyed connector (XT30) on the pack lead, or a P-channel MOSFET or a series Schottky in front of the module.

• Input capacitor in backwards

  part in backwards

  Trigger
  Why it happens

  Symptom: C1 warms, may hiss, then vents through the cross on its top. The switching pulses now come from the battery wires.

  An aluminium electrolytic only works one way round: reversed, its oxide layer breaks down, it conducts, heats and builds pressure until the vent opens.

  Fix: Replace C1 (a vented capacitor is never reused) with the stripe toward the board edge.

• Trimmer wiper open

  cold solder joint

  Trigger
  Why it happens

  Symptom: The meter jumps from 5.00 V to 7.6 V: the output follows the input. The motor runs fast and hot. Turning the trimmer does nothing.

  With the wiper disconnected, FB sees only R1 to ground: 0 V. The LM2596 thinks the output is far too low and turns its switch fully on. This is the dangerous failure: whatever is on the output gets the battery voltage.

  Fix: Reflow the trimmer's pins. For anything that cannot take the input voltage, add a zener or a crowbar on the output.

• Pack run flat

  brown-out

  Trigger
  Why it happens

  Symptom: The output sags with the pack: 4.7 V from 5.5 V in. The motor slows. The display keeps working down to about 4 V, the converter to about 3.6 V. Nothing warns you.

  Below Vout plus the dropout (about 0.8 V here) the switch stays on and the output is the input less the dropout. The module has no under-voltage lock-out, so it keeps draining a 2S pack far below 6.0 V, the point where the cells are damaged.

  Fix: Charge the pack. On a robot, add a cut-off: a LiPo alarm, or a Nano that reads the pack through a divider and switches the load off at 3.3 V a cell (the cut-off task).

• Motor held

  overcurrent

  Trigger
  Why it happens

  Symptom: The motor current jumps to 1.25 A and the pack current to about 0.9 A. The meter still reads 5.00 V. The LM2596 warms to about 70 °C.

  A held motor has no back-EMF: it is just its 4 Ω winding, 1.25 A at 5 V. That is far under the 4.2 A the converter can give, so the output holds. A small motor does not collapse this rail; a big one, or two in parallel, would hit the current limit and pull the output down.

  Fix: Let go. In a robot, stop the motor in code when it stalls (the current rises, the encoder stops) before the gearbox or the module overheats.

• Output shorted

  short circuit

  Trigger
  Why it happens

  Symptom: The meter reads 0.00 V. The display, on IN, still reads the input, because the meter circuit runs from the input. The SS34 heats; after a few seconds it fails and the output stays dead.

  The LM2596 limits its switch to about 4.5 A peaks, 4.2 A on average. Into a short the output is near 0 V, so the switch is on less than a tenth of each cycle and the diode carries the current the rest of the time. TI asks for a diode rated at least the current limit; the SS34 is rated 3 A, and it overheats.

  Fix: Remove the short and replace D1. Fit a fuse on the output, or a diode rated 5 A (SS54) if shorts are likely.

• Input over 40 V

  wrong voltage

  Trigger
  Why it happens

  Symptom: The diode overheats and fails; often it takes the regulator with it, and a regulator that fails short passes the full input to the output.

  While the switch is on, the diode sees the whole input voltage in reverse. The SS34 is a 40 V part, and TI wants 1.25 times the highest input. The module is sold as 4 to 40 V, which already leaves no margin; 44 V is past it.

  Fix: Stay under about 32 V with this diode, or fit a 60 V one (SS36 or SS56).

Live status

Check yourself

• After: Solder the through-hole joints

  Unlocks after: Solder the through-hole joints

• After: Power the motor

  Unlocks after: Power the motor

• After: Ask it for 12 V

  Unlocks after: Ask it for 12 V

• After: Break it: short OUT+ to OUT-

  Unlocks after: Break it: short OUT+ to OUT-

• End of bench quiz

  Unlocks when every step is done