The meanest of servo jitter!

[fritsl]

Working on a project, where I want a servo to rotate/position a laser that is reflected to a mirror, that is vibrated by magnetic pulses.. every day stuff here on LMR :)

Earlier I posted this:

- And now it was time to hook it up.

However strange things happened. When I was pulsing the mirror, the servo was moving!

I thought that

* I had hooked up something wrong
* I needed (more) seperate power supplies
* Wires where too close to each other, or spun around each other so the servo got signals that way
* The Microcontroller got some feedback from the magnets, and reset

Eliminating one after the other, I ended up hooking the servo up 100% isolated, only giving it power on the 2 lines, on a stand alone power supply.

Then, by another battery pack, I simply tried to give voltage to the coil, and see what happens on the video!

It even "works" when the coil is getting very little power (on the video it's just getting hard core 5V).

The materials between the coil and the servo are non-magnetic, and the coil is 10 cm away from the servo!!

Now I have proof tested this, and aparently, a coil and a (strong) magnet - which is what is in many motors(!) can cause the servo to freakin move!

Dammit, I have to find another way of pulsing the mirror, this is the hardest, most evil case of servo jitter I have ever experimented! I would not believe this if I did not see it, but it is true, unfortunatly.

[jka]

I have experienced servo

I have experienced servo jitter when I hook up a servo to a power supply without any control signal. The servo is moving a bit, when I put in the plug. Try to connect the control line of the servo to ground to suppress noise or try to let the picaxe hold the servo still and have the coil on a separate power supply. The magnetic force of the coil might pull the motor and there is nothing to hold it back, because the servo is not told to hold the servo at a specific position when you only hook up the power lines.

[fritsl]

Thanks, but I am way beyond

Thanks, but I am way beyond this;

First I tried to lock the servo by all sorts of servo-controllers, then hook it up, but keeping the control pin low…

No matter what; if the servo get’s power supply, and the coil is activated, servo moves!

I COULD of course unhook the servo completely, and then pulse the coil… but then every time I hooked up the servo again, it would jitter, by the power up. And it would make a quite strange setup in general

The coil and the strong magnet and the servo just cannot live next to each other, unless I invent a magnetic field-thingey to put in between!

[rik]

Just to be clear
the servo horn does not turn when there is no power to the servo at all? No matter how hard you click the mirror with the solenoid/coil?

[fritsl]

No, i is clearly that the

No, i is clearly that the servo - if powered - internally is getting a wrong message.

If it at the same time is getting “real servo pulses”, then it just gives a hard yank (not wank, yank) … while doing it’s normal servo business.

If it at the same time is not geting any signal, only power, it does at shown on the vieo.

Not powered at all, nothing happens.

[wohn]

Wrap the servo in "Mu

Wrap the servo in “Mu metal”. It is the best for magnetic and emi. Its a little expensive but will dot he trick no mater what. Just some pointers on using it, bend slowly, DO NOT solder it and remember it is metal and will conduct if it contacts anything. Heat alters the effectiveness of it so try to avoid getting it hot.

Also what may be happeneng is is when the coil becomes de-energized it may be sending a pretty big spike past ground and moving the reference points around for the microcontroller and it may go into a frozen state where the uC stops sending the pwm signal to the servo for a brief period of time. Try using a diode to catch the flyback voltage.

[fritsl]

Tried diodes on all

Tried diodes on all wires!

It’s inside the servo that something is going on.

Thanks for the info on “Mu metal” - I am not going for it though, as I have made a new solution wihout the coil / magnet