I'm building a self balancing unicycle which uses 5v and less than an amp for it's "brain" circuitry. However, I also have a 50v supply which is supplying power to a brushless motor at about 30A. The grounds of the two power supplies need to be connected (of course), but I'm worried about noise from the motor affecting the 5v circuitry.

Do I need to put a giant cap between the power and ground rails of the motor power supply? Or perhaps a giant inductor between the two?

Any help is greatly appreciated! Thank you in advance!

 

Better to put a nice fat electrolytic capacitor on the 5 volt supply, in parallel with a .1uf capacitor to catch the high frequency components. Then be extremely careful about how both the currents travel in the ground system. Don't let both circuits share a length of wire. Attach them both at a single point on the battery.

 

put a nice fat electrolytic capacitor on the 5 volt supply, in parallel with a .1uf capacitor to catch the high frequency components.

Can you define "nice, fat" in terms of microfarads?

Then be extremely careful about how both the currents travel in the ground system. Don't let both circuits share a length of wire. Attach them both at a single point on the battery.

I should've mentioned this before, but the 5v PCB is running off a 9v battery + regulator. The 50v supply is a LiPoly battery pack. Can you explain in a bit more detail about currents traveling through the ground system?

 

Put an inducor followed by a cap in series with the + line feeding the clean supply.

 

I think about 100 uf should work.

A single point grounds the LiPo pack, the motor, and the 9 volt battery. A seperate wire is run from each of these devices to the ground point. That way the motor current doesn't develop little voltage spikes on the ground wire and make the 9 volt supply get all jumpy as the current surges happen.

 

A single point grounds the LiPo pack, the motor, and the 9 volt battery....

By "a single point" you mean like...anything besides a long ground trace or long wire, correct?

 

By a single point, I mean a nut and bolt in free air, or one lug of a lead acid battery, or the only spot on the chassis that the motor is grounded to. A single point is not a splice into the side of a wire that is carrying intermittent 30 amp surges on the way back to the LiPo pack or a second bolt threaded into the frame.

 

By a single point, I mean a nut and bolt in free air, or one lug of a lead acid battery, or the only spot on the chassis that the motor is grounded to. A single point is not a splice into the side of a wire that is carrying intermittent 30 amp surges on the way back to the LiPo pack or a second bolt threaded into the frame.

So... You're saying the following diagram is a bad idea? I currently have it configured such that the boards are connected through a simple molex connector.

 

Also have some ceramic capacitors in parallel with the electrolyte capacitors.

 

So... You're saying the following diagram is a bad idea? I currently have it configured such that the boards are connected through a simple molex connector.

That's exactly what #12 is saying - Yes it is a bad idea to wire the ground of the control circuit via the high power board.

 

That's exactly what #12 is saying - Yes it is a bad idea to wire the ground of the control circuit via the high power board.

So... Is this what everyone is suggesting?? Or is there supposed to be a cap in there somewhere?

 

Close. The purple wire from the LiPo pack should go all the way to the screw, not attach to the ground wire from the Mosfet board. The motor ground should also have its own private ground wire, all the way to the screw.

Each controller board ahould have an electrolytic capacitor in parallel with a ceramic capacitor about 1% the size of the electrolytic. The are connected from the voltage supply of the board to the ground on the board. This keeps the noise down on the circuit boards all the way up to the MHz range.

 

Gotcha, thanks.

The GND wire coming from the low voltage board can be a small wire, right (22 AWG)? The one coming from the MOSFET board needs to be large though, like 14 AWG?