Looking for a cheap way to prevent back voltage

[swuster]

I'm trying to build a circuit that can power multiple sources with a single battery (rated at 18.5V but the charge is pretty unstable and could be anywhere from 16-20V depending). Among other things, the battery is powering several servos which are prone to producing back voltage when activated...what's the best way for me to fix this? I was planning on adding some sort of buffer, but thought it might be risky or simply impossible to power the buffer with its own input (the whole point of the system is for it to be able to run off a single battery). Appreciate any suggestions out there as I'm not very well versed with this sort of thing. Thanks!

 

[berkeman]

I'm trying to build a circuit that can power multiple sources with a single battery (rated at 18.5V but the charge is pretty unstable and could be anywhere from 16-20V depending). Among other things, the battery is powering several servos which are prone to producing back voltage when activated...what's the best way for me to fix this? I was planning on adding some sort of buffer, but thought it might be risky or simply impossible to power the buffer with its own input (the whole point of the system is for it to be able to run off a single battery). Appreciate any suggestions out there as I'm not very well versed with this sort of thing. Thanks!

Can you diode isolate each consumer of power? You would have your power supply feeding each consumer through a power diode (a power Schottky if the voltage drop matters).

 

[vk6kro]

Servos are usually connected across a power supply of about 6 volts.

They generate a small reverse voltage when power is applied but this is easily countered by the incoming power.

However, when the power is removed, they have considerable power stored magnetically and they can deliver a large positive going pulse towards the regulator.

It is common to put a diode across the regulator if this is a problem on power down.
The diode is placed with the anode to the output and the cathode to the input.

This avoids the regulator having a high voltage on its output and no voltage on its input which could be a fatal situation.

 

[dlgoff]

It is common to put a diode across the regulator if this is a problem on power down.
The diode is placed with the anode to the output and the cathode to the input.

I would put the diode across the servo at the source of the pulse. Less noise that could couple from the wiring to the rest of the circuit. IMO

 

[vk6kro]

The problem is that the pulse is positive going, so where would you put the diode and which way around? The supply is positive so it can't be cathode to ground or it would short out the supply.

I confirmed that there is a large positive going pulse with a test setup here.

Putting a diode across the regulator at least protects that during power down. Having a bypass capacitor between the servo and the regulator might help a little.

Putting a diode in series with each Servo would stop reverse voltages getting back to the regulator, but also cause a drop in voltage of 0.6 volts which would have to be allowed for by generating 6.6 volts instead of 6 in the regulator.

 

[swuster]

I'm actually powering the servos at their 4.8V mode, so the lines are regulated to 5V. If I put the diode across the regulator, that means the back voltage goes all the way back to the battery. Seems like that might be a possible problem...

 

[vk6kro]

I'm actually powering the servos at their 4.8V mode, so the lines are regulated to 5V. If I put the diode across the regulator, that means the back voltage goes all the way back to the battery. Seems like that might be a possible problem...

This is a voltage generated when you remove power. The regulator would possibly be vulnerable to reverse voltage but the battery would just see it as a momentary charging voltage without much current capability.

So, having the voltage bypass the regulator seems like a good idea and the battery should not have much of a problem with it. It might hardly increase the battery voltage at all.