Reverse current of the servos

Depending on some servo types various customers claimed about undervoltage entries in the event log although the power supply was supposed to be strong enough. They used good known BEC or a 2s LiPo directly connected. The issues could be reproduced in some cases with collective movements on a single servo without any load!

Using a fullsize VBar instead of a Mini some of this problems seemed to be solved - but hard to explain. The only difference is a small additional capacitor in the fullsize sensor. Should this be our key here? - And if ... what might be the reason?

So, thats why we started another bench scope measuring session ...

We used:

- Mini VBar
- 1 Servo Graupner HBS 860 BB MG
- Futaba S-Bus
- 2s 2200mAh LiPo direct connected to 3 slots of the VBar
- Measured with a scope, current measuring with a shunt

 

The result was disappointing. It seems to have servos on the market which have a bad internal electric design, they will provide reverse current. Shown here as peaks above the zero line (blue line, zero is named with "2" at the left side). You wll see the voltage at the yellow line too:

 

Here we have the log. No discussion, this is not acceptable!

 

Servos which will show this behaviour could be a possible noise source - disturbing all the model electronics, possibly causing a crash. If you will see this kind of issues the further usage of the model is forbidden until it is solved.

Although we have used a 2s LiPo direct connected with 3 cables it seems not to be possible to work as a buffer here. The internal resistance of the battery (new!) is too high. A battery is always better in this point instead of a BEC, but it seems not to be enough here.

What should we do ... we tried with a additional capacitor  connected to a unused port of the Vbar. A capacitor can handle such current very fast, look a this screenshot:

 

And we got a clean log:

 

 

So it is solved from our point here. The servos have some bad electric design. A servos must just execute orders and should not disturb any kind of model electronics! - Most of the servos are ok in this point, but not all. Please avoid such servos - or - take care about the reverse current and add capacitors or whatever will be needed to avoid any risk here.

In our test example it was enough to add one capacitor with 6800 uF for all servos (3 swash, one tail). In some forums they discussed other solutions using special diodes but we could not get a visible effect on the scope with this method.

Using a  Mini it may be more visual in the log as with a Standard VBar. The additional sensor has a additional capacitor which helps a bit here. But .. keep in mind ... the capacitors in VBar are not made for fighting against reverse current of the servos.