2

I have two LM1117T through-hole low dropout regulators that I bought from two different sources. Both are labeled as the 3.3V variant.

I'm reading about 4.2 V across Gnd and Vin. One reads 3.27V on Vout, and when I hot-swap the other one instead, it reads 3.9V on Vout.

Is this just a case of bad quality components or could there be something else going on? There is no load on this test circuit, nor do I have any capacitors or anything else connected - just the +/- from my 4056 battery charger module going into Vin and Gnd on the LM1117T.

My goal is to power an ESP32 which needs regulated 3.3V. I'm using a 14500 3.7V Li-ion battery with the 4056.

JRE
  • 67,678
  • 8
  • 104
  • 179
Tom Auger
  • 265
  • 1
  • 9
  • 2
    Should "it reads 3.9V on Vin!" be "it reads 3.9V on Vout!"? Also, the datasheet *does* specify a minimum load current; can you verify if you've met it? – nanofarad Jul 03 '22 at 17:46
  • Can you specify, what mean by hot swap? You swap it on a live circuit with power applied? – Ilya Jul 03 '22 at 17:51
  • Sorry for the error in my post. I absolutely meant I was reading two different voltages across Gnd and Vout. By "hot swap" yes I meant that I left the +/- leads connected on my solderless project boards and just removed and popped in a different LM1117 component. – Tom Auger Jul 03 '22 at 17:55

2 Answers2

8

The problem is that you have, as you put it, no "capacitors or anything else connected".

There are two problems with that:

  • The LM1117, like most linear regulators, requires those capacitors for stability, or it may start oscillating, basically rapidly turning itself on and off. This will give incorrect readings on a multimeter, and probably damage the regulator over time as it's not meant to be used like that.
  • The LM1117 requires a minimum load current of at least 5 mA (see page 7 of the datasheet, table entry "minimum load current") to guarantee that it will keep regulation over its entire rated conditions.

Fix both of those, and your problem will likely be solved.

Hearth
  • 27,177
  • 3
  • 51
  • 115
  • I'm not great at reading datasheets yet. It looks like 10µF Tantalum cap on both sides (Vin and Vout)? I've also seen someone else' schematic that has a 100µF electrolytic in parallel to a smaller ceramic cap on the Vout. Where can I get definitive information on what is required? – Tom Auger Jul 03 '22 at 19:04
  • 1
    @TomAuger The datasheet will tell you. The application information section is a good place to start; in this case, you want section 9.2.2.1.3 (TI *loves* their section numbers) on page 16. That gives you minimum requirements; you can always add more capacitance if needed. – Hearth Jul 03 '22 at 19:29
  • 2
    @TomAuger Note that tantalum capacitors are no longer as widely available or as cheap as they used to be, since modern large ceramic capacitors have taken over their former market segment (mainly low voltage, low profile capacitors around 10-47 μF). I would use a ceramic capacitor for this, but since the datasheet says that the capacitor needs to have an ESR of between 0.3 and 22 ohms, put a small resistor in series. Ceramic capacitors have a much lower ESR, around 0.005 Ω or less, than tantalum or aluminum ones. – Hearth Jul 03 '22 at 19:33
  • @TomAuger (this is a development within the last decade or so, which is why many datasheets still specify tantalum capacitors--they don't get revised that often, and those revisions rarely if ever change that much) – Hearth Jul 03 '22 at 19:34
4

As well as your immediate issue (lack of capacitors), note that you are very close to the dropout voltage of the LM1117, which is not really a low-dropout regulator, more of a "medium dropout".

So (particularly) when the battery voltage drops you may not get the full 3.3V out, since the regulator needs 1.25V or 1.2V to work reliably over temperature (for 500mA max or 100mA max respectively).

Even a true LDO won't necessarily give you regulation over the useful output voltage range of the battery, which might extend to below 3V.

Spehro Pefhany
  • 376,485
  • 21
  • 320
  • 842
  • Is there a better LDO component I should be considering - or perhaps a completely different approach to powering my 3.3V projects from LiPo or LiFePo batteries? I have a number of all-in-one modules but they are very bulky and I'm going for trying to keep things compact. – Tom Auger Jul 04 '22 at 14:23