[SOLVED] LM7805 output very not stable

Hello

I'm using a LM7805 to get a stable 5v from a 12v power supply.
I connected a 103 capacitor between Vin and GND.
and 103+220uF(16v) between the 5v and GND.

if I check the 5v with multimeter, I get a read of 5.05-5.06. which is good.
but as I use it for digital readings and had many problems with it, I checked it as a digital read and saw that the signal is jumping from 0v to 5v all the time. Very Very not stable.

I'm not sure, but I think it wasn't like this yesterday. Not sure about that.

Shouldn't I get a clean 5v output? Specially when 2 capacitors are connected at the output?


Thank you

I'd move 220uF from output/GND to input/GND, and rather than 103 (10nF) use 104 (100nF) on both pins (input/GND and output/GND).
Then, if the 12V does not fluctuate “significantly”, the LM7805 should produce very stable 5V (5.05 is good enough) ..

Also remember about heat generation: (12V-5V) - that's voltage drop, times the load current ...

:wink:
IanP

I agree with IanP recommendations for the input and output capacitors, but even with no capacitors the output of the 7805 should not go between 0V and 5V. Check for a bad joint or disconnection to the 7805.

What is the current rating of your circuit, is it more than 1A, sometimes the output of regulators fluctuate when the current requirements go more than the rated current of the regulator or if there's any short circuit, so check your circuit for shorts and then see what is the rating or current requirement of your regulator.

Thank you for your answers.
after trying changing the capacitor to 104 and changing the position of the bigger capacitor and even changing the L7805, I still got fluctuating signal.

when sampling at a rate of 1000 times a second, I got 86% 5v at the best try.
when using the 5v of the Arduino microcontroller, I get 100% all the time.

The problem with the Arduino is that the max current is more limited, and I need to connect few components together

- - - Updated - - -

BhavaniT, if I'm sure that there are no short-circuits, can I know that the current is ok, or should I limit it somehow?

I use it for digital readings and had many problems with it, I checked it as a digital read and saw that the signal is jumping from 0v to 5v all the time. Very Very not stable.

Click to expand...

How did you checked it (digital)?

with oscilloscope?

Shachar85 said:

Hello

I'm using a LM7805 to get a stable 5v from a 12v power supply.
I connected a 103 capacitor between Vin and GND.
and 103+220uF(16v) between the 5v and GND.

if I check the 5v with multimeter, I get a read of 5.05-5.06. which is good.
but as I use it for digital readings and had many problems with it, I checked it as a digital read and saw that the signal is jumping from 0v to 5v all the time. Very Very not stable.

I'm not sure, but I think it wasn't like this yesterday. Not sure about that.

Shouldn't I get a clean 5v output? Specially when 2 capacitors are connected at the output?


Thank you

Click to expand...

If your multimeter reading says 5.05 volts, then that is the average voltage. But then you report some kind of "digital" reading jumping between 0 v and 5 v. This is not possible. If the average voltage is 5.05 volts then either the signal is stable at about 5 v or the signal is jumping between 0 v and something much higher than 5 v, perhaps 10v. So I question your "digital" instrumentation methodology. I am not sure what you think is happening is really happening. Can you get some kind of tie-breaker in there to look at the output? Perhaps an old-fashioned analog scope?

The readings were made with the analog input of the arduino.

I might be wrong about the readings. It seemed strange to me too. It acted like PWM, which doesn't make sense at all.


What I'm sure about is that:
1. Multimeter reads 5.05v-5.06v
2. Reading as analog input to Arduino gives me many different numbers. Most are 0 or 1023

now, the readings are between 0 to 1023, so I'm not sure exactly what it reads if I pass the 5v

anyway, I'm confsed
and I wish I had a scope, but...I dont
(the number are int

It sounds like the power supply is fine and your problem is software. Connect a battery to your analogue input and see what happens, e.g. 1.5V.

Keith

The readings were made with the analog input of the arduino.

I might be wrong about the readings. It seemed strange to me too. It acted like PWM, which doesn't make sense at all.

Click to expand...

Well i must say there's nothing so spooky about your power supply fluctuations, i have come across some linear regulators who behave in such a manner whenever we start drawing their max current at the output, in fact i have seem those power fluctuations on an oscilloscope and yes they seem exactly as you said - a PWM waveform.
reasons for such a thing to happen is
1 - when you don't give an input voltage equal to required "output voltage + Dropout voltage" given you are drawing maximum current at the output.
2 - When your circuit requires more current than your regulator can provide and in this case your output voltage of the regulator will even fall to somewhere 2 to 2.5V just to obey Kirchoff's laws.

BhavaniT, if I'm sure that there are no short-circuits, can I know that the current is ok, or should I limit it somehow?

Click to expand...

If you are sure that you don't have any short circuits then do a power budget analysis of your circuit, see what current each and every IC requires(worstcase maximum), sum it up and see if it falls below rated current, if yes then everythings fine, or else you don't have any option than to change the regulator.

As far as current limiting is concerned, you are not supposed to limit the current as it effects the circuit's performance, if a circuit requires a given amount of current then you don't have any option than to supply that amount of current. current limiting is carried out in cases where you don't want a high current input source to destroy your circuit.

Just to be on a safer side replace the regulator with a newer one and see if it also behaves the same way.

And i would suggest you to stick to this circuit until and unless you find out what is the fault and rectify it, believe me it would be a great learning experience(been through same scenario;-)) and share with us if you rectify the problem.

Moreover i would suggest you to use electrolytic capacitors with the values suggested in the datasheet (0.33uF & 0.1uF) try this first then the other things i have suggested.

Shachar85 said:

The readings were made with the analog input of the arduino.

I might be wrong about the readings. It seemed strange to me too. It acted like PWM, which doesn't make sense at all.


What I'm sure about is that:
1. Multimeter reads 5.05v-5.06v
2. Reading as analog input to Arduino gives me many different numbers. Most are 0 or 1023

now, the readings are between 0 to 1023, so I'm not sure exactly what it reads if I pass the 5v

anyway, I'm confsed
and I wish I had a scope, but...I dont
(the number are int

Click to expand...

I feel the problem is with arduino. Some times, the output of 7805 may go bit higher about 0.5v due to the design/quality of the regulator.
I think the voltage above 5.0v is the problem.

Check as keith said..

I checked with a new circuit and new 7805.
I only connected the 7805 and the 3 capacitors, so I'm sure I didn't need more voltage (Vin = 12v).
Though, I didn't put any resistors either...could it lead to a very high current that reached the 7805's limit?

I will also check if the problem is the 5.05v instead of maximum 5v.
I'll keep you posted when i try (unfortunately it will be only in Sunday...)

OK guys, I got it figured out, and to be honest...it is pretty embarrassing.
Common ground.

The arduino and the L7805 must have common ground (the embarrassing part is that I knew it)

Thanks for all the help!