PIC + Servo , Noisy initial power supply

Hi everyone. I am currently working on a circuit with a 18F452 PIC. I have developed my code and everything is working fine on breadboard + oscilloscope.

Next step was to connect the servos on the circuit. I noticed that whenever I connected the servos my circuit stopped working properly. Disconnected them and the circuit is working fine again. The response of the PIC showed that something was wrong with the power supply of the circuit so I connected the oscilloscope on the 5V (after the 7805).

What I noticed is that:
Servos DISCONNECTED - Powering up takes few ms and it is a clean line from 0 to 5V
Servos CONNECTED - Powering up is extremely noisy taking about 80ms+. Many times during those 80ms, the power signal drops 1V+ and then up again until it settles.

Questions:
1) Why this happens? I have few ideas but I think they are wrong
2) In order to fix it I tried to place a capacitor between + and ground after the 7805 (additional to the ones of the 7805 datasheet). The capacitor was 0.33uF. There was NO improvement. Do I need to buy a bigger capacitor?

I will put up a couple of pictures I took from the oscilloscope when I go home.

Oh and something else. The power supply was adjusted to 0.5A. I raised that to 1.5A and the circuit is working fine. The circuit uses about 8mA

Thanks for reading! Any advice would be really appreciated!

Sorry, but the answer is simple and obvious - the servos are taking too much current and pulling the supply down - this is why you usually power them seperately.

With no load on the servo, it should be drawing about 7 to 12 milliamperes at rest. As soon as a command to move it occurs, the current will jump to a higher value. How high? depends on the load but can easily approach 1 ampere for a fraction of a second.

Click to expand...

hope this helps

No need to answer like that Nigel, I am sorry I don't know that.

I just thought that since the servo does not need that much current in order to run (i.e. that power is temporarly needed) maybe it could be fixed with a capacitor/diode/(something else) trick.

Anyway, thanks for your answer.

be80be said:

hope this helps

Click to expand...

Excellent! Thanks man...

demestav said:

I just thought that since the servo does not need that much current in order to run (i.e. that power is temporarly needed) maybe it could be fixed with a capacitor/diode/(something else) trick.

Click to expand...

Not really, it's a powerful DC motor, and takes considerable power everytime it moves, for as long as it moves. Even staying in one place will take power as well, if there's any mechanical stress on the servo - such as holding the steering on a model car in any place except the centre - and even at that bumps in the terrain will cause it to take power to hold itself in t's fixed position.

So the solution is to always power them differently?

0.33uf pwr supply smoothing cap is a joke, try 300uf, or even 3300uf ;-)

demestav said:

So the solution is to always power them differently?

Click to expand...

Pretty well, it makes it far easier that way.

Its not a good ideal to run any motor from the regulated supply and it don't hurt to use optocouplers for your control