regulator output: overvoltage

[burningman]

I am reading nearly 15.1volts on the output pin unloaded. I was wondering what is the cause of my 7815 going over voltage? Is there a way to correct this or is it a bad regulator?
Thanks.

[anchovie]

What's the accuracy of your multimeter? Has it been calibrated this year? 

You're measuring 15.1V on the output of a 15V regulator, right? A difference of 0.1V? I'd say there's nothing to worry about! 

[rustypinto]

Look at the 7815 datasheet. They come in different output tolerances like 2% and 4%. The manufacturer will also give values in the datasheet for specific input conditions and the resulting output for all their regulators in the series. You'll notice the tolerance reflects the minimum, typical, and maximum output possible.

Sounds like a perfectly good regulator to me.

[liquids]

I might be jumping ahead of you, but its probably about the risk/benefit ratio.  This may be helpful to to others even if it doesn't apply so I'll note it anyway....

I'm assuming you are either very concerned about getting what you paid for in regard to it being a 15v regulator in principle,  and/or you are concerned with precise regulating because you have a component that is being fed said regulated voltage, and 15v is at the 'max' of the datasheet range.  If so, you could probably view it a few ways...

1) that you are pushing said part TO or OVER it's maximum voltage level, which could cause premature or inevitable failure, and you might want to either drop it down to say, 12v regulation, for example, to ensure reliability of the component and to be 'within' the datasheet for said part, which specifies it's own range of what it can 'tolerate' and how it may respond...

or,

2) Realize on some level, the datasheet is merely a promise/contract from the manufacturer...and say, 15v is all they want to promise it can handle max, which means realistically, some manufactured devices can handle more that that without or with varying levels of reliability unit-to unit, but none should fail up to that voltage.   You can hence go for the gold hoping you may be okay at that threshold with a worst-case-scenario component, and determine if it's worth the risk....factoring in selling it to a customer, reputation, repair/reliability balance and cost, if calculating all of this is worth your time and effort, or if you personally need peace of mind and guaranteed reliability a live gig in order to sleep at night, etc, etc...which, in such cases, see #1.   

At that, if you buy an assortment of 15v regulators, you are bound to find one that puts out 14.9v which is within tolerance, say (see datasheet) etc, which may ease your concerns...?

[burningman]

I am feeding the input roughly -25volts so it is well above the dropout voltage. From your answers my guess is that it just has to do with tolerance and that the regulator is fine. I just wanted to make sure that it wasn't a symptom of something else.

[Gus]

Check the reg spec often you need a load of a few ma for the reg to work correcly

[burningman]

The datasheet mentions a quiescent current is 4.4ma - Is this what you are referring to? I can't find any other min values for load. I am using the NTE 969

[Gus]

4.4ma and 15VDC
Ohms law
15VDC/.0044A=3409K  You could use a 3.3K
The use a DC power equation for the wattage needed

When I test a power supply before connecting it to a circuit, I often build a resistive load set at about the current the circuit operates at and connect the load to the PS and take some measurements.

[PRR]

Entirely within specs:



Note that 7815's UN-loaded output is not specified. I have never seen one soar (some adjustable regs will go wonky when unloaded), but I would not worry about the exact output until it had some load (they assume >5mA, or 3K at 15V). I would only look to see if it was way-off in case I had made a wiring mistake (bad GND pin will pass raw supply voltage). If it was "near" 15V I would connect my system.

The LM7815 will take 20V to 35V and deliver a smooth output over 14V but under 16V, usually very close to 15V. That's plenty good for almost anything except precision DC test systems (and some analog synths). You round-up from your signal needs and round-down from your chip ratings. for example: 4558 is rated +/-18V max, we rarely ask for more than 10V signal, it has 2V loss: the power rail must be between 12V and 18.0V. The 7815's 14.4V-15.6V spread is plenty precise for such use.

If you need 15.0V, you need to trim. If you need 15.0000...V, you need a Precision Voltage Reference. And beyond the third or fourth digit requires calibration against Standard Cells or a transfer reference.