5V regulator outputting 5.9V and more.

[Simon]

I made a small board with a µC and a 5V regulator. On power up i was surprised to find that it outputs 5.9V. I thought maybe it does not have enough load so i put a 510 ohm resistor on its output as the datasheet says that all the specs given are at 10mA. But now it outputs 6.1V.

is it a dodgy part? did I overheat it on re-flow? It's a LM2931ADT-5.0RKG

[Wimberleytech]

I made a small board with a µC and a 5V regulator. On power up i was surprised to find that it outputs 5.9V. I thought maybe it does not have enough load so i put a 510 ohm resistor on its output as the datasheet says that all the specs given are at 10mA. But now it outputs 6.1V.

is it a dodgy part? did I overheat it on re-flow? It's a LM2931ADT-5.0RKG

Measure the voltage on all three pins (right at the pins of the regulator).  Let us know what you read.

[Simon]

that is the voltage on the regulator pin. There is a 330nF capacitor either side as close as practical.

[andy3055]

Can you post the schematic you used?

[Wimberleytech]

that is the voltage on the regulator pin. There is a 330nF capacitor either side as close as practical.

I would like to know that pin 2 (the ground pin) is actually zero volts.
You should also look at the pins with an oscilloscope to be sure that there are no oscillations.

[andy3055]

I assume you have this:
https://www.onsemi.com/pub/Collateral/LM2931-D.PDF

[Simon]

Well relative to that 6V output the ground pin is 0V I took the measurement across the regulator. I'll shove the scope on it.

[Simon]

I assume you have this:
https://www.onsemi.com/pub/Collateral/LM2931-D.PDF

Yes

[andy3055]

So, which  circuit did you use?

[thinkfat]

It might oscillate. Try with a larger output capacitor.

[Simon]

So, which  circuit did you use?

Regulator, 330nF either side

[Simon]

I see they don't suggest any value of capacitor, but some example diagrams have what appears to be 100µF

[wraper]

I see they don't suggest any value of capacitor, but some example diagrams have what appears to be 100µF

Really? Sometimes it's worth to read the datasheet besides simply looking at example schematics. Also note that LDO are not 78xx. You cannot simply slap few hundred nF on output and expect them working properly.

[james_s]

330nf is much, much smaller than I would typically use. I bet if you scope it you'll find that it's oscillating, I ran into that once when I used too small capacitors on the input and output of a 7815.

[Simon]

OK, fair cop I never saw that way of explaining it before.

[wraper]

If you look at max ESR value, It's probably best to use MLCC or LOW ESR tantalum. Keep in mind that if you choose electrolytic, ESR most likely will be above mentioned 0.4 Ohm for 10-22 uF capacitors even if they are LOW ESR type.

[Simon]

what abut my existing 330nF with an electrolytic?

[james_s]

It's a common technique to pair an electrolytic with a ceramic capacitor, not as necessary as it used to be but it will still work.

[Simon]

Just remembered, i have some nice 150µF polymer capacitors at work.

[floobydust]

LM2931 is an old 1st gen LDO, finicky about the output capacitance (value and ESR) and I would use something else. Ensuring it stays stable over temperature can be a challenge when including the O/P capacitor's behaviour. 150uF solid polymer may be outside the stable range for ESR at room temp. alone.

[MagicSmoker]

LM2931 is an old 1st gen LDO, finicky about the output capacitance (value and ESR) and I would use something else. ...

Yep, this. I like the NCP5501 series of LDOs because they are pin-compatible with the 78xx series and have much more relaxed ESR/min. capacitance requirements. The LM2931 and xx1117 series should be left to the surplus bins and stocked only for replacement purposes.

[james_s]

There's nothing really wrong with them, regardless of what part you use it's prudent to study the datasheet carefully. All components have limitations and this is a classic case where RTFM would have avoided the whole issue.

[floobydust]

The LM2931 has more traps than usual. Noobs use a MLCC or polymer output capacitor and have troubles, especially in cold weather.
Jim Williams did a nice write up on using short PCB traces a few mm long to add ESR, have to dig for that LT app note.

[wraper]

Apparently LM2931 from different vendors are not equal. Say according to ON semi datasheet it does not have any problems with extremely low ESR (MLCC). But according to TI, it should not be to low. On the other hand, TI can tolerate double ESR on the high side. IC schematic in the datasheets has some differences too.

[wraper]

The LM2931 has more traps than usual. Noobs use a MLCC or polymer output capacitor and have troubles, especially in cold weather.
Jim Williams did a nice write up on using short PCB traces a few mm long to add ESR, have to dig for that LT app note.

More like little problems with half decent MLCC (given added series resistance or using LM2931 from ON semi) and big problems with aluminium electrolytic at low temperature according to those charts. It actually recommends poly cap (OS-CON).