power line noise problems-help required

muruga86 · Apr 30, 2005

hai,
i have an application using pic micro 16f877 powered by 9v transformer via 7805 regulator 1000uF and 100uF bulk decoupling cap near regulator, and an RC filter 1ohm with 0.1uF. the problem is when any high voltage AC motor connected to the same power line with micro is switched on, during few seconds my micro behaves abnormally and displays junk value and then return to normal agter few seconds. can any one give me a suggestion about the easiest way to over come this power line noise problem.

thanks in advance

Murugan

DarkJedi · Apr 30, 2005

Well, you are connecting a high voltage AC motor to a pic line... that is a lot of noise. Try separeting the power supplies, you can use opto couplers to do so and separate the power from the digital circuitery.

One more way to solve it could be using a battery for the digital circuitery instead of the transformer, you can separate the power supplies that way, using two power sources.

Hope it helps.

IanP · May 1, 2005

In applications like yours RC low-pass filters are not efficient enough to protect microcontroller from AC/DC motors switch on/off interfirencies..

Try to use ≈1mH inductors between 1000µF capacitor and 100µF//100nF capacitors at the 7805 input and one on the 0V line, and ≈100µH inductor between 7805 output and microcontroller +V (both sides should be decoupled by 10µF tantalum // 100nF ceramic capacitors).
Also, you can use small EMC filter at the input to the transformer ..

Other thing to do is to use optical isolation on all digital inputs and outputs of this microcontroller, specially those which are somehow associated with motor(s) control ..

muruga86 · May 2, 2005

thanks for your advice, i will try that one can any one give me a typical rough schematic for ciruits which work in noisy environment..

thanks in advance

murugan

vdharan · May 2, 2005

my micro behaves abnormally and displays junk value and then return to normal agter few seconds

Where do you get the junk value..?.Is it displayed in an LCD..?..!

IanP · May 3, 2005

Attached is a picture with an example of noise filtering, ground separation, gnd return paths .. etc etc.
It is not exactly what I suggested to you in my previous post .. but pretty much in-line with it.

It is also a good guide for AGNG and DGND separation in mixed circuits ..

muruga86 · May 3, 2005

thanks for the schematic, but i can't understand the schematic clearly.

Added after 40 minutes:

thanks all for your advices,

i have one another problem the application which i have build calulate the speed of a 3HP AC motor by using M18 proximity sensor and display its RPM using 7 seg display. for each revolution the proximity gives a active low output which is coupled by using optocoupler pc817 and given to RB0 as interrupt. during normal running this arrangement works fine but during running when i switch on/off the motor during that brief moment numerous interrupts occurs and RPM display shows unusual values. (i found out by displaying the number of counts of interrupts occured). BUT MICRO NEVER RESETS .how to avoid this problem it is due to power fluctuation in 5V power to micro or 12V supply to proximity. please advice me how to condition dignals comping going to interrupts

thanks in advance

murugan

IanP · May 3, 2005

From what you descreibed the interference seems not to pass through 5V microcontroller supply, but rather from 12V and proximity switch. One of possible solutions will be to carefully screen cables to and from this proximity sensor ( try to find good shielded cables with >90% shielding).

Also, what you can look at is the implementation of solid state relay(s) to switch the motor On and Off. This will ensure that you will not have any "spikes" caused by a normal switch. Keep in mind what a microcontroller is capable of doing within, say, 20ms .. comparing with an action of a "slow" switch ..

artem · May 3, 2005

May be your sensor acts as transformator and get interfered with AC , due to fact that at motor startup you have highest current through motor coil and electromagnetic field strength should reach max , easely affecting sensor . Or parasitic path from metal shield of sensor through 9v transformator's coil capasitor .

oddbudman · May 3, 2005

Why not also add a transorb or an zenner diode across the +5V rails to help clip excess voltage peaks? This combined with and LC input filter may do the trick.

Also have you put some capacitors on the inputs to your opto circuit? It is probably excessive triggering due to noise coupled by the AC motor you mentioned. By adding some capacitors to this you should be able filter out fast triggering spikes. Also increasing the input resistance to the optocoupler may help reduce its sensitivity.

Hope this helps. Experiment a little with the system and see what happens.

oddbudman

muruga86 · May 3, 2005

thanks for advice,
the proximity sensor which i use is industry grade with metal schielded enclosure and used widely in industries no problem with that. and moreover i am not controlling the ac motor by micro, only monitering its speed and display RPM. the above mentioned motor is manually on/off by using a lever switch. so only during that on/off time of AC motor false interrupts were trigerred which the micro senses and micro goes mad. can i try signal conditioning circuits?

Added after 3 minutes:

i have also added 0.1uF ceramic disc on signal and gnd line of 12V side of optocoupler and another 0.1uF disc on micro side but problem still persists.

neuralc · May 3, 2005

Place also a Varistor at the input of the transformer of power supply of PIC

HH

NeuralC

artem · May 4, 2005

Just to ensure that n othing comes from AC supply line - use battery for power supply to hcek whether problem persist .

If it is the case - you know the cause .
If it is not - get a scope and check the sensor output when AC motor starts . You said that micro does not restart - that could also means sensor outputs garbage at AC motor startup .

power-twq · Jun 10, 2005

you can connect a inductor in front of the 5V regulator to supress the noise

that motor generated.

muruga86 said:
hai,
i have an application using pic micro 16f877 powered by 9v transformer via 7805 regulator 1000uF and 100uF bulk decoupling cap near regulator, and an RC filter 1ohm with 0.1uF. the problem is when any high voltage AC motor connected to the same power line with micro is switched on, during few seconds my micro behaves abnormally and displays junk value and then return to normal agter few seconds. can any one give me a suggestion about the easiest way to over come this power line noise problem.

thanks in advance

Murugan

coco · Jun 13, 2005

Try the subber circuit together with the optocoupler.

Welcome to EDAboard.com

power line noise problems-help required

Member level 3

Full Member level 2

Advanced Member level 7

Member level 3

Junior Member level 3

Advanced Member level 7

Member level 3

Advanced Member level 7

Advanced Member level 4

Member level 5

Member level 3

Full Member level 4

Advanced Member level 4

Full Member level 6

Newbie level 6

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor