LM317 voltage regulator question

[BDuguay]

Now, I'm no rocket surgeon but, a voltage regulator is supposed to keep voltage at a constant right? The power supply I built for my pedalboard uses an LM317 and was putting out 8.9 volts dc. I had some pedalboard problems recently and during my trouble shooting I measured over 11volts dc coming from my power supply. I removed the 1.5k setting resistor, which still measured in spec, and replaced it with a 1.2k resistor which brought the voltage down closer to 9. I'm wondering if something is wrong with my power supply. Can a regulator drift? Is this a sign that it might be failing? I know that the few extra volts shouldn't be a problem as fresh 9volt batteries can put out 10volts or more but, my tuner stopped working and I wonder if it's because it was getting to much juice.  Does anyone have an idea what's wrong here?
Thanks,
Brian

[H.Manback]

How have you measured it? Did you measure the voltage under load, or did you measure the open circuit voltage? I'm not sure if a LM317 will put the regulated voltage out without load, but I built one recently and it did put out 9 V exactly without load.

[Hal]

I used one of these a while back, and I don't remember well, but isn't it set by 2 resistors?  If you're pulling too low current on the resistor set, it could become instable.

[R.G.]

The LM317 is designed to be self loaded by the set resistors so that it works properly all the way down to zero load on the output. The handiest way to do that is to make the top resistor 240 ohms and set the voltage with the bottom resistor.

The LM317 needs the specified caps near the input and output to be stable.

LM317's don't drift noticably, but resistors and cold solder joints do.

[BDuguay]

Both measurements were taken without a load. What I'd like to be clear on is once a voltage regulator is set, it's job is to constantly deliver that voltage correct? I'm glad you chimed in R.G. as the power supply I built was from your website. Thanks for your replies gentlemen.
Brian

[Paul Perry (Frostwave)]

What I'd like to be clear on is once a voltage regulator is set, it's job is to constantly deliver that voltage correct?

Absolutely right.
But, it won't do it, unless the voltage of the raw dc (at the 'input' pin of the 317) is at least 3 volts higher than the output you are expecting.
Sometimes, aas you draw more current from the 317, this overloads the transformer plugpack supplying the raw DC, so input drops to where the regulator hasn't got enough headroom to work with.
I learnt this the hard way, like most things :oops:

"low dropout" regulators can work with less headroom.

[BDuguay]

One last question. Make that 2. R.G. has made reference to the setting resistor possibly drifting in my power supply. Which way would it drift to cuase what's happening? Lastly, the resistor I had in there was a 1.6k that actually measured 1.5k. Would that resistor need to be under a load to achieve a more accurate measurement?
Thanks again!
B.