5
\$\begingroup\$

I recently shorted out my AMS1117 5 V; it turns out the output is directly connected to input now, that is, I now get direct 12 V supply onboard.

I wanted to know if it is specific to a brand or the if the built-in short-circuit protection doesn't work for anyone.

Edit:

I short-circuited the output of the AMS1117 to ground. The short-circuit protection did not work. Instead, now the input and output of the regulator are internally connected; this is very dangerous for my onboard components.

AMS1117 datasheet here

\$\endgroup\$
4
  • 1
    \$\begingroup\$ How did you short it? Output to Gnd? Input to Gnd? Input to output? Is there load on output, or no load at all? How much capacitance (total, not just what is immediately next to regulator) on output, and input? \$\endgroup\$
    – Justme
    Commented Nov 1, 2020 at 12:04
  • \$\begingroup\$ You damaged that regulator so throw it away. You probably did something that damaged the regulator. When used properly (not exceeding maximum ratings, see datasheet) it is quite difficult to damage such a regulator. \$\endgroup\$ Commented Nov 1, 2020 at 14:29
  • \$\begingroup\$ There are varying quality versions available. A properly specified device would be expected to be properly short circuit proof. If it failed it failed - quite possibly due to being an inferior product - but that's not certain. Vin max is 15V so the 12V should not have been an issue. \$\endgroup\$
    – Russell McMahon
    Commented Nov 3, 2020 at 12:33
  • \$\begingroup\$ I also just accidentally shorted the output of an AMS1117-5V to GND (very shortly, screwdriver slipped) and afterwards it was also forwarding its 12V input to the output, smoking several components on the PCB. Painful experience :-( This page was the first thing I googled, I guess I'm not alone. The AMS1117 came with my PCB from JLCPCB. \$\endgroup\$ Commented May 8, 2023 at 9:11

3 Answers 3

2
\$\begingroup\$

Well I've just had a similar experience with AMS1117: I have some peripherals that are powered by a Nodemcu ESP32's own 3V3 LDO, I was distracted and connected that to the ESP32, swapping 3V3/GND.

So the peripherals (3x I2C IO expander, and a LCD) got reverse polarity on their power supplies. There were a few electrolytic decoupling caps in there, so they probably limited the voltage to less than a volt.

Everything survived except the AMS1117 which just gave up. It somewhat kinda worked after that, delivering an unstable output, so the ESP32 would still boot, then immediately go into an infinite loop of reboots as soon as the chip tried to start up the power-hungry Wifi.

Here's what the datasheet says:

The AMS1117 series of adjustable and fixed regulators are easy to use and are protected against short circuit and thermal overloads.

Well this chip costs 3 cents, so I guess that's what it's worth. Although the chip could also be a counterfeit, with these super cheap modules, you never know what's on them. Maybe it's not an AMS1117 at all, or it came from the reject pile that somehow "disappears" during night shifts, so it wouldn't be surprising that it doesn't perform as per spec. If you bought chips from aliexpress, wish, etc, just assume they're counterfeit.

Since the thing was powered by a PC USB port, in theory that limits current to 500mA, so it blew way below its rated current.

I replaced it with ST LDL1117 I had lying around, and the ESP32 now works. I would trust the short circuit protection of LDL1117 a lot more. Also I got it from Farnell, so I'd assume it's genuine.

Basically, these super cheap regulators are not a good deal. LDL1117 and many others cost a few cents more but they deliver much better transient response with a much smaller and cheaper ceramic cap, whereas AMS1117 requires a large tantalum cap. So total cost is probably higher with the cheap regulator. Unless the tantalum cap is also counterfeit, of course.

\$\endgroup\$
0
\$\begingroup\$

You are dropping a lot of voltage. Maybe the regulator was already hot and having a sudden increase in power dissipation might have caused rapid thermal stress and the protection wasn't fast enough.

\$\endgroup\$
0
\$\begingroup\$

I was looking this thread, and today I had the some problem with a 5 V 1117 regulator; pin 2 and 3 are shorted and my Arduino died with 13 V across it.

I put a 5.1 V Zener diode together with a 1 Ω resistor between the "+" and "-" dots. With this scheme, if the regulator fails, the diode and the resistor keep the voltage at 5 V.

\$\endgroup\$
1
  • \$\begingroup\$ How effective this will be for overvoltage protection depends on how much current the unregulated supply can source, and how much power the zener and resistor can dissipate. A better protection circuit might be a power SCR and the 5.1V zener to its gate, which constitutes a "crowbar". This would normally be combined with a fuse or circuit breaker on the input which will trip before the SCR overheats or other components (such as wiring) exceed their ratings. \$\endgroup\$
    – PStechPaul
    Commented Nov 16, 2022 at 0:28

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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