Smoothing Supply

I am using one of these XP Power SMPS ECE10US12 to run the Arduino that controls my setup, I have been having some issues that may or may not be related to the SMPS. I thought smoothing the supply couldn't do any harm but as I know very little about these things I was hoping somebody could recommend the most appropriate configuration for the smoothing

The manufacture says the maximum capacitance the SMPS should see is 550uF, this maybe overkill but what would I know. In the reading I have been doing it seems that having a resistor involved is a good idea, but calculating the correct value is a complete mystery. I imagined a RC or a CRC is probably what I am after. All suggestions would be greatly appreciated.

I can afford to loose a couple of volts from the supply if necessary, as the Mega will run on less than 12V
 

Attachments

  • SF_ECE05-10.pdf
    381.9 KB · Views: 57
Last edited:
A Power Amplifier draws it's current from the smoothing capacitor.
In a normal (unregulated) power supply that smoothing capacitor is usually in the many mF range. It can supply lot's of current for a very small change in supply voltage.
The transformer in the PSU does not supply the transient currents.

However in an SMPS where the output capacitor is strictly limited in maximum value, the amplifier and speaker current demand must be met by the SMPS.

That is a problem. Very few SMPS give useful information on the transient current capabilities.

To solve the problem without the manufacturer behind you holding your hand, you in effect become the SMPS designer.

If you can't design the SMPS, then don't try to modify by exceeding the manufacturer's recommendations.
 
I have found the attached Design Note which relates to smoothing the ripple coming from a switching DC/DC convertor, which probably (wild guess) would do the trick smoothing the supply coming from a SMPS.

My problem is my lack of maths skills, the Design Note has a couple of things in the calculation that I am unsure how to apply them, I have highlighted the offending items, If anyone can enlighten me to how they should be applied in the calculation I would greatly appreciate it.
 

Attachments

  • Output filter LC design.pdf
    76.6 KB · Views: 32
How are you controlling on/off state? Is it electronics doing the latching of the state or is it something in code that switches an output on/off and the on/off state on the arduino pin is held constant?

If it is the latter and you have some sort of timer/counter, my guess is that there is a problem where the variable runs out of memory and sets back to 0.
 
How are you controlling on/off state? Is it electronics doing the latching of the state or is it something in code that switches an output on/off and the on/off state on the arduino pin is held constant?

If it is the latter and you have some sort of timer/counter, my guess is that there is a problem where the variable runs out of memory and sets back to 0.

To start the system three relays are turned on, the relevant arduino pins are sent high to switch the the relays on. the problem mainly seems to be with it randomly switching the relays on, by initiating the start sequence (more happens than just the relays). I've eliminated the IR Receiver, either the code has some weird bug that initiates the start sequence or some electrical interference does it, there my two guesses. I thought if I smoothed the supply to the arduino it may help.

At least it couldn't hurt.

I'll just chuck a capacitor in and see what happens.
 
Certainly can't hurt, I don't think it'll solve the problem. It does sound like a code bug to me.

I'd be checking for variables related to timing/counting, they might be exceeding the size of the largest value able to be stored in them. Perhaps try sending serial events from the arduino to a computer to track state of code as different events occur, may give you a clue to where things are going awry.