The volt it needs to work
The AMS1117 needs its input about a volt above its output, more as the current rises. Below that it stops regulating and the output follows the input down. USB's 5 V is enough for 3.3 V out. A lithium cell is not.
About a volt, more under load
The transistor inside the chip cannot open all the way. Even flat out it drops a volt or so, and that sets the lowest input that still gives a steady output. UMW's datasheet gives these, for the 3.3 V board's chip:
| Current | Typical dropout | Worst case | So VIN must be at least |
|---|---|---|---|
| 100 mA | 1.00 V | 1.20 V | 4.3 V |
| 500 mA | 1.05 V | 1.25 V | 4.35 V |
| 1 A | 1.20 V | 1.30 V | 4.5 V |
Slide the input down and watch where the output leaves 3.3 V:
Right of the knee the output is flat: the chip has headroom and uses it. Left of the knee the chip is as open as it goes and the output is simply the input less the dropout. Nothing is wrong with the board; it has run out of volts.
Which supplies are enough
- USB, 5 V. 1.7 V of headroom. Enough, with margin, at any current the heat allows.
- 9 V or 12 V. Plenty of headroom — far more than needed. The next page is about what that costs.
- One lithium cell. 4.2 V fully charged, and most of its charge is spent near 3.7 V. From 3.7 V at 240 mA the output is about 2.7 V: below the 3.0 V an ESP32 is rated to run from, and not far above the 2.43 V where it resets.
- Two AA cells. About 3 V. A regulator like this only ever lowers a voltage.
USB is not always 5 V
USB promises between 4.75 V and 5.25 V at the port, and a long, thin cable loses more on the way. 4.75 V still leaves 1.45 V of headroom, which covers the worst-case dropout. A cheap cable under a heavy load can take that margin away, and it shows up as an output that is fine on the bench and low in the project.
When it does not work
That is dropout, not a fault. At 3.7 V in and a couple of hundred milliamps out it needs about 4.3 V, so the output is the input less a volt. A single cell needs a regulator specified for a few hundred millivolts of dropout, or a buck-boost.
Two things grow with current: the dropout, and the voltage your supply loses in its cable. Measure VIN while the part is running. If VIN is under about 4.4 V, the input is the problem — a shorter or thicker USB cable, or a better supply.
No. Two fresh cells give about 3 V, and a regulator like this can only lower a voltage, never raise it. The output would be around 2 V. Raising a voltage needs a boost converter.
Just, while they are fresh and the current is small, and not for long: three alkaline cells fall below 4.4 V well before they are flat. Four cells, or USB, leave margin.
Every volt above the output, times the current, is heat. How much the chip can take.
Where the extra volts go →Edit this page — content/books/ams1117/the-volt-it-needs.mdx
Questions about this product
See what other owners have asked, and read their solutions.
AMS1117 LDO Regulator Kit, 24 Boards (18 × 3.3 V, 6 × Adjustable)
Loading discussions…
Discuss this article
Ask about this page. The answer stays here, on the page it belongs to, for whoever hits the same wall next.