Induction Problem in Microcontroller

Thread Starter

rmrps

Joined Feb 24, 2012
21
Hi Friends I am using Pic32mx460f512L controller my problem is when i switch off and on 230vac ceiling fan pic32 controller get disturbed and some times its get resetted i am tracing the power supply 24vac input via oscilloscope some unwanted hormonics coming from the 230v ac source when switch off and on 230vac fan i think i have to correct my power supply with best filter please give me your guidance to resolve this issue with best filter i have attached my power supply design please help to overcome induction problem
 

DickCappels

Joined Aug 21, 2008
10,152
That looks like a creative solution. Did you try it yet?

I think I would try putting a transient suppressor across the transformer and perhaps a low pass filter -a series inductor (some millihenries) then a good low equivalent series resistance capacitor (100 uf or so) to ground between the 1000 uf capcitor and the regulator. Also be careful about grounding so that noise from the transformer-rectifier-filter does not go through the circuit ground.

Sometimes noise coming in through a controller's I/O lines can cause problems with I/O ports themselves and/or program execution, to be aware of noise coming in through those lines as well.
 

Thread Starter

rmrps

Joined Feb 24, 2012
21
i am placed mov and 2.2mh 2amps twin coil compensator across the transformer secondary and i am used murata 103/100v suppression filter capacitor series with transformer secondary but no use the same problem stile there
 

AfdhalAtiffTan

Joined Nov 20, 2010
120
In my experience, filtered supply is not enough, you'll need to make sure any long wires attached is properly shielded and decoupled. Generally speaking, you might want to avoid making antenna.
 

cmartinez

Joined Jan 17, 2007
8,220
That looks like a creative solution. Did you try it yet?

I think I would try putting a transient suppressor across the transformer and perhaps a low pass filter -a series inductor (some millihenries) then a good low equivalent series resistance capacitor (100 uf or so) to ground between the 1000 uf capcitor and the regulator. Also be careful about grounding so that noise from the transformer-rectifier-filter does not go through the circuit ground.

Sometimes noise coming in through a controller's I/O lines can cause problems with I/O ports themselves and/or program execution, to be aware of noise coming in through those lines as well.
A few years ago, I installed an MCU for controlling the motor starters (contactors) of several big water pumps for a very large fountain. Sometimes the MCU would be reset after a couple of seconds... That ended having nothing to do with the large inrush current that the motors were drawing, but rather with the huge EMP produced after the magnetic coils kicked in. I ended having to shield the entire MCU in a grounded, metallic box (a faraday cage), and using a filtered isolation transformer to power up the MCU.
 
Last edited:

JohnInTX

Joined Jun 26, 2012
4,787
Echoing the others a bit - ensure you have good power decoupling at the uC pins, are not sharing power/GND with noisy loads, snubbing the source etc. My experience is that fast transients will sail through as many 3-terminal regulators as you want to put in the line (decoupling caps notwithstanding).
When all else fails, try a common mode choke in the power leads. I've really only needed one once but they can be a lifesaver when all else fails..
Google turns up lots of info.
Good luck. These things can be a pain.
 

Sensacell

Joined Jun 19, 2012
3,432
I have experienced PIC's resetting if a transient on an I/O line goes above or below the rails, current in the clamping diodes seems to do it.

Summary- look for spikes on the I/O lines as well as the power bus.
 

cmartinez

Joined Jan 17, 2007
8,220
I have experienced PIC's resetting if a transient on an I/O line goes above or below the rails, current in the clamping diodes seems to do it.

Summary- look for spikes on the I/O lines as well as the power bus.
Are you suggesting using TVS's in the circuit's powe rails, even if they're connected to regulated and filtered power supplies?... I'm saying this 'cause I haven't done that before... but they might help....
 

Sensacell

Joined Jun 19, 2012
3,432
Expand your investigation to consider more that just the power bus, look for transients on the I/O lines too.
Look for any other source of coupling where noise can enter the system.
 
Top