Art Project in danger: Servo and Led do not work together

[digistump]

nero - if it helps I used a standard Futaba servo, and powered it from 5v through the usb connection so it was about 4.7v on the device. What pin are you connecting your power to and what is the voltage?

[nero]

We used a Tower Pro 995 Servo, it is a high power servo.

We only have a power supply with 4,5V (4,65 in reality) or 6V, we try to get another one with 5V, maybe that is what caused the problem.
We did not want to put the servo on the digispark directly because the draw a lot of power.

The digisparks are powered over usb kables which are connected to our power source. Like this we hav a power supply connection an a holder in one but the power does not run over the onboard regulator. Therefore we used only 4,5V to power the system.

[Mark]

nero
Cheers for the answer.
The rapid led on/off may be introducing extra spikes on top of the servo.

One thing you could try, is a capacitor across the 5v near the controller, or the servo.

Some servos may have them included internally, and others may not.
The tower Pro 955 is a high torque but I can't see any current ratings for it..

Its a long shot, but it could be noise on the supply line, upsetting the micro.
In theory your oscilloscope should show this with the LED and servo running (just use a resistor in series with the led, and put across the 5v and ground.)
It will appear as noise on the 5v, which disappears if you undo the servo control wire.

The digispark only has a 4.7uF capacitor across the 5v supply. It needs 47uF or 100uF for any load.
For those interested, A general rule of thumb for supplies is 1000uF per Amp, and more is better, but don't forget the 0.1uF as well.

Mark

[nero]

Maybe we solved the problem, we use a laboratory(do you call them like this in english?) power supply now and with 5V we seem to get rid of the problem, no consistent results but on a first test the led and the servo worked together.
It seems that 4,6V were on the lower border of what th digispark needs and adding the led an the servo probably lowered the voltage so much that the digispark turned of an restarted itself.
We alos found out that a single servo draws about 0,5amps and we have 43 :O) normally they do not move at the same time but around ten is quite possible so we have to look for an even bigger power supply now.

Thank you all very much for your help and if you are around Darmstadt (Germany) between the 9. and the 19. of march (regular hours Friday, Saturday and Sunday 16.00-20.00) have a look at the Earlstreet Gallery where we put up our installation.

Earlstreet Gallery, Pallaswiesentrasse 25, 64287 Darmstadt, Germany

[digistump]

nero - if you are powering them over the USB connector you could even go up to 5.5v to help compensate for loss in the diode on the Digispark and line losses - you amperage should be fine.

[nero]

Oh good to know, with the new power supply (it is actually very old but solid) we can do that.
Thanks

[Mark]

Well done nero

I'd suggest you find some rather large capacitors, to hang across the supply line at various places.
This will help with the motors.
In theory the supply from a PC should be able to handle quite a few, and there are various references on web to running them without a motherboard.


Mark