How accurate is an L7806 voltage regulator ?

[Jolly 7]

Is an L7806 regulator supposed to provide a constant output of 6.00 V ? At various times, I'm measuring outputs of between 5.9 to 6.2V. Is this variation acceptable ? Could mains voltage variation be behind this small fluctuation ?

[cmjones01]

The 78xx series regulators are not precision devices. Glancing at the data sheet for a 7805, its output voltage at 25 deg C is specified as 4.8-5.2V, and it will also vary a bit with input voltage and output load, as well as temperature.

However, if you're seeing that much variation from one regulator under constant conditions, something is probably wrong. Has it got good decoupling capacitors on the input and output? Or is the meter faulty?

Chris

[turretslug]

Does it have sufficient input headroom- under load, it may be ducking-out on ripple troughs, resulting in the meter giving a different averaged reading.

[Jolly 7]

Originally Posted by cmjones01

The 78xx series regulators are not precision devices. Glancing at the data sheet for a 7805, its output voltage at 25 deg C is specified as 4.8-5.2V, and it will also vary a bit with input voltage and output load, as well as temperature.

However, if you're seeing that much variation from one regulator under constant conditions, something is probably wrong. Has it got good decoupling capacitors on the input and output? Or is the meter faulty?

Chris

I have a 103 cap connected from input to ground and another one from output to ground.

[cmjones01]

Originally Posted by Jolly 7

I have a 103 cap connected from input to ground and another one from output to ground.

That should be OK, though a larger value (100nF) would be conventional.

In my experience, a lot of '7806' regulators are really cheap devices aimed at mains adapters for radios and so on. I don't know if this means they're any worse than the more common 7805, 7812 and 7815 from the bigger manufacturers.

Chris

[G4YVM David]

The OP measured less variation than the o e spec'd. 5.9to6.2 is tighter than 4.8 to 5.2...isn't it

[turretslug]

4.8 to 5.2V output variation refers to the +- 4% spread accepted across devices marketed as "7805", rather than output variation of a sample device. I'd expect similar percentage tolerance to apply across a spread of 7806 types. +- 4% variation in the output of an individual 78xx device over its specified operating regime would be rather disappointing!

[G4YVM David]

Ok, understood. Thanks.

[Biggles]

My experience of these types of regulators is that they are usually within a couple of hundred millivolts at worst to the specified output voltage. They are still my first choice when knocking up a quick circuit even though they have their problems. I wonder if you have a poorly made clone of the genuine device?

[ajgriff]

L7800 series devices (cloned or not) are quite sensitive to high junction temperatures. Adequate heatsinking is important to maintain the specified output voltage especially when significant current is drawn.

Alan

[Jolly 7]

The 7806s were purchased from Spiratronics in the UK a couple of years ago. I have another of these, I might set up a temporary circuit to see if it behaves the same way. Nothing too critical is being connected to either, just a 6V radio circuit and possibly other bench experiments in the future.
I have to say though this is far better than a Paget wallwart I have. It's reading a whopping 9.1V at the 6V setting !

[Herald1360]

Cheap older wallwarts are usually unregulated and only deliver anything near to as low as the rated voltage when supplying the rated max current along with plenty of ripple!
Cheap new ones will be switchmode and usually deliver the right voltage but with plenty of RF noise along with it.

Older ones which feel suitably heavy (50Hz TX inside) are worth having to feed any ripple resistant application, but pick one which delivers the right voltage on the actual load rather than just rely on the nominal rating. For low current (relative to the wart's rating) loads where ripple matters, there's usually enough overvoltage from one rated at the voltage you want to feed your own downstream 78xx or other regulator's headroom requirements.

[SteveCG]

I agree with Herald1360. I would just add that if you are trying to power a LW/MW/SW portable radio then the older transformer ones are very desirable, except that a small, say 0.01uF capacitor, across the secondary winding may be needed to reduce the rectifier diodes switching noise.

[ChristianFletcher]

I was also going to say you get variations from one device to another but individually they are normally very accurate better than your getting. I would also check the headroom and decoupling

[Terry_VK5TM]

Sounds like not enough capacitance at the input and output if you are getting that sort of variation on the one device.

[Mooly]

Lack of a minimum load current can affect the negative 79 series devices... might just be worth you adding something like a 560 ohm across the output as a test.

A scope check will also be more revealing than static voltage tests and as has been mentioned, check that the input voltage troughs never drop below the minimum specified input voltage for the device.

Whatever the voltage is 6, 6.1, 5.9 etc, it should be 100% consistent and repeatable and never alter from that.