The slot on controller boards that accepts a Pololu/StepStick/TCMxxxx usually has a capacitor next

gplus
discussion
(Nathaniel Stenzel) #1

The slot on controller boards that accepts a Pololu/StepStick/TCMxxxx usually has a capacitor next to it. I read that it is a 100 uF capacitor on the motor power supply rail. If there are 5 stepper driver slots, do you really need 5 of them or could you use a 500 uF capacitor? I mean, they are all connected to the same power rail. After looking at the prices on one website, it looks like there are much more options for 100 uF and that the price is much cheaper. I wonder if that is why or if there is another reason.

(Duncan Gunn) #2

Stick with individual capacitors in this case, and keep them close to their associated drivers.

(Nathaniel Stenzel) #3

Does placement affect the functionality?

(Duncan Gunn) #4

The traces on the board add parasitic properties to the circuit making placement of such things something that should be considered. As a rule of thumb, electrolytic capacitors aren’t super critical but small high frequency decoupling capacitors are.

Just stick with the norm and you’ll come out ok. You should check out the EEVBLOG for a few hints on this.

(Nathaniel Stenzel) #5

I stuck it on the homescreen of my phone for future browsing.

(Ryan Carlyle) #6

Motor drivers ABSOLUTELY MUST HAVE two things:

  1. A small amount of low-ESR capacitance to dampen transient voltage spikes as the MOSFETs switch. The capacitor has to have very low internal resistance to absorb inductive kicks and switching noise so electrolytic caps don’t work. This must be as close to the driver chip as possible, for example and is placed directly on Pololu boards.
  2. A large amount of capacitance somewhere nearby to help buffer the Vmot supply voltage. The Vmot supply to the driver otherwise will rise and fall excessively as the driver H-bridge gives/takes energy from the motor coils (inductors) to control coil current, and also as the motor rapidly alternates between drawing more power from the PSU to accelerate and returning power back to the PSU to brake.

#1 keeps the motor from destroying the driver.

#2 keeps the drivers+motors from damaging the rest of the board. For example, if you’re using fast decay mode (reverse voltage on the motor coil) while the motor is aggressively decelerating the load, the motor is acting as a powerful generator, and will quite happily pump energy into your control board. Have a reverse-voltage protection diode on your board PSU hookup? Great, you just trapped that energy within the control board, and it has to go somewhere or you’ll over-voltage the board and fry everything. (This is why moving motors with the machine powered down can damage some controllers.) The 100uF bulk capacitance helps soak up excess energy return so it can be used elsewhere, like running a different motor or powering a heater.

It MIGHT work to have one 500uF cap for five drivers, but you’d have to position that capacitor farther away from some of the drivers (making it less effective) and you would probably make the overall ESR of the bulk capacitance higher (again making it less effective).

(ThantiK) #7

@Ryan_Carlyle , thank you for that brilliant explanation.

(Nathaniel Stenzel) #8

@Ryan_Carlyle It is #2 that I am curious about. Lets see…the copper traces can act as an antenna for the electrostatic discharges, right? That is why #2 should be as close to the driver as possible?
Now it looks like the selection of 100 uF is for cost, placement, low resistance and many options.
I think I will have the understanding that I was going for as soon as these two questions are answered.

(Ryan Carlyle) #9

@NathanielStenzel the traces have some small amount of capacitance, inductance, and resistance proportional to their length and also radiate and receive energy like antennas. Most of the time (like for heaters) you don’t worry too much about that stuff, but high-frequency effects make you care about tiny lags and such.

The long and short of it is, the longer the traces, the less effective capacitors are at doing their jobs. If the capacitor has to provide X amps for a tiny moment, and the trace has Y resistance, you’re losing X*Y volts through the trace. Or if there’s too much parasitic inductance/capacitance the cap can’t provide that much current in such a short time. We’re dealing with really small numbers here, but also very fast changes in current/voltage, so it can matter.

(Nathaniel Stenzel) #10

Thanks for all the explanations. I appreciate it. My curiosity on this matter is now quite satisfied.