Switching noise of ac motor

Go To Last Post
64 posts / 0 new

Pages

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Read the manual for your scope and have a play.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

engmado wrote:
sorry, i didn't get this how to do do it my oscilloscope is  gw instek 1104b it's new. what do you mean by Trigger to fire once on negative edge. Trigger on voltage just under VCC?

 

You got to learn how to use your scope. As far as I can see, its a DSO type.

 

I have no experince with this specific model, I suppose you have a manual?

 

Your scope most likely have a trigger menu or other ways to select triggering. Set trigger to single event, negative edge. If you have VCC=5V, set voltage level to 4V. If 3.3V set it to about 2.5V. Wait for an event that fires the trigger. Se on the screen what you got.

 

I am serious about the wires. Open loops generates and collects noise. Twisting wires in pairs reduces both. Try it!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I did that and there was spike when relay turn off but i am not sure if voltage level was 4v . The only wires now are for output signals and lcd i can't put it on the board.
I noticed that Vcc was constant during far switchings.but if i switch relays fast every second , there are spikes on vcc so i think that it may be the diode of relay coil is slow i replaced it with 1n4148 fast diode and i will see its effects
How does open loop look ,give example please to compare it with my case.

Last Edited: Wed. Mar 21, 2018 - 03:30 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

engmado wrote:
I did that and there was spike when relay turn off but i am not sure if voltage level was 4v .

 

4V is just a suggested trigger level. A voltage drop of a volt or more got to make trouble.

 

engmado wrote:
I noticed that Vcc was constant during far switchings.but if i switch relays fast every second

 

If your scope is running free, then chances of seeing a glitch increases with number of glitches, but the scope will quickly owerwrite the screen. Try again to set proper triggering to catch a single event. I mean this:

1. set trigger as described to single event, falling edge, probe on VCC or reset as close to IC1 as possible. Your users guide may use other words, but it should be described in trigger menu.

2. your scope will now display a totally blank screen.

3. kip the relay, and bang, the scope will display the glitch. Scope screen will not refresh! Remember: a DSO scope can show not only what happens after triggering, but also what happened before it.

4. once captured, the scope can measure how deep the voltage drop is, and how long it lasts.

 

engmado wrote:
so i think that it may be the diode of relay coil is slow i replaced it with 1n4148 fast diode

 

You cannot outpace a diode with a relay, no matter how slow the diode is! 1N4148 is a small signal diode, not suited for the purpose. It might blow. You want a diode that can take some beating.

 

engmado wrote:
How does open loop look ,give example please to compare it with my case.

 

You already know how it looks from your own pictures. Any 2 wires carrying the same current, and having a distance between is an open loop, asking for trouble.

 

3 things troubles me in your setup:

 

1. power for the motor. When the relay opens to stop the motor, current is likely still running through motor, wires and relay contacts. The motor is an inductive load, which means that energy is stored in its magnetic field. When relay breaks, this energy will create a spark over the relay contacts. That again turns your motor wires into antennas, emitting electromagnetic noise. This is a primary candidate for your problems. Twisting the motor wires together and keeping them as short as possible reduces the problem. A triac that breaks only at zero crossing could be much better than a relay.

 

2. same goes for relay coil, but to a lesser degree.

 

3. your PCB have a ground plane horseshoe shaped with a giant hole in the middle! Even worse, the 2 GND pins of Atmega328 are connected to each leg of the horseshoe, but not directly to each other. Your ground plane is an antenna! That could easily pull the carpet under the atmega.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

how to improve ground plane ?should i split it? should i connect the two gnd pins of the atmega8 to each other?

I twisted the motor wires together, i replaced 1n4007 with 1n4148 , i noticed improvement and it's working since yesterday. it is working with no resets or random characters on lcd.

i set up scope as you mentioned there are spikes sometimes there aren't when relay off on vcc and reset pins but its magnitude is less than with 1n4007.

Last Edited: Thu. Mar 22, 2018 - 09:12 AM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

engmado wrote:
how to improve ground plane ?should i split it? should i connect the two gnd pins of the atmega8 to each other?

 

YES! Exactly! Connect the 2 GND pins of Atmega directly, but don't split the ground plane.

 

engmado wrote:
I twisted the motor wires together, i replaced 1n4007 with 1n4148 , i noticed improvement and it's working since yesterday. it is working with no resets or random characters on lcd. i set up scope as you mentioned there are spikes sometimes there aren't when relay off on vcc and reset pins but its magnitude is less than with 1n4007.

 

1N4148 may have a short life in this application. High current or voltage spikes can kill it. If you think a fast diode is better, then chose a fast rectifier in the 1Amps range, or a diode made for noise suppression.

 

I think you could benefit from reading up on electrodynamics. Try to understand that wires acts as antennas. Wires both emits and receives noise. Proper arrangement of wires greatly reduces noise.

 

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

How to eliminate spikes then?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

engmado wrote:
How to eliminate spikes then?

 

I real world you cannot eliminate spikes and noise.

 

You can reduce noise. You can reduce the effect of noise in your circuit.

 

Understanding electrodynamics is essential. Read up on it!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

engmado wrote:

How to eliminate spikes then?

 

Do you understand what is causing the spikes? If not, now is the time to learn...

 

#1 This forum helps those that help themselves

#2 All grounds are not created equal

#3 How have you proved that your chip is running at xxMHz?

#4 "If you think you need floating point to solve the problem then you don't understand the problem. If you really do need floating point then you have a problem you do not understand." - Heater's ex-boss

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I've suggested that OP should read up and study EMI related problems somwhere on the top of this post. He seems to ignore most of the advice given here and repeat questions like:

engmado wrote:
How to eliminate spikes then?
This leads to a very tedious guiding of OP to partial solutions and is why this is now post #61. I find reading / following / posting in threads like this a waste of my time, and am not inclined to participate further.

http://www.catb.org/~esr/faqs/smart-questions.html

Paul van der Hoeven.
Bunch of old projects with AVR's:
http://www.hoevendesign.com

Last Edited: Thu. Mar 22, 2018 - 04:36 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

 

1N4148 may have a short life in this application. High current or voltage spikes can kill it. If you think a fast diode is better, then chose a fast rectifier in the 1Amps range, or a diode made for noise suppression.

 

 

The reverse-voltage supression diode in this application only needs to handle the current that the relay draws when it is energized.  Remember, an inductor (the relay coil) simply attempts to maintain a constant current through it.  If a relay coil draws 20ma when energized, when it is de-energized that same 20ma is imposed across the diode, and only for a very short time at that.  1N4148s are extremely fast diodes and will work fine, assuming the diode voltage rating is also considered.

 

edit:

 

To blather on a little more, when an inductor de-energizes, the voltage rises as high as it needs to until the current through the coil reaches what it was previously.  If the suppression diode is "slow" like a 1N400x, the reverse voltage in theory, can rise much higher than the diode's rated forward conduction voltage because of the time needed for the diode to go into conduction.

 

 

Tom Pappano
Tulsa, Oklahoma

Last Edited: Thu. Mar 22, 2018 - 06:03 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I think that the problem is solved because of twisting the motor cable and arrange cables.
Thanksv every body

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

There is one problem still exists. Sometimes the lcd get blank also when I tried to connect capacitor directly on its vcc it get blank but the mcu is doing the code

Pages