Actions
Forum Thread Details
  • State Suggested Answer
  • Replies 15 replies
  • Answers 1 answer
  • Subscribers 475 subscribers
  • Views 5270 views
  • Users 0 members are here
Related

Design Challenge : relay spike protection

Catwell
This is a common problem with relays. The relay coil acts like an inductor when power is removed, and it sends a spike of power back into your circuit. And let's say you are originally putting 5V through the coil. When the power is removed, in some cases, the coil could send a spike of any level, even in the range of 400V, back into your circuit. And the problems cause by that are many, from burning out circuitry to tremendous electromagnetic interference.
 
 
I have seen spikes that oscillate between hundreds of volts positive and negative. Like below:
 
image
The challenge is to eliminate the problem in the cheapest way, with as few components as possible. Let's just say space and money are scarce for this project. And the relay is connected to very sensitive circuitry.
 
 
Use the relay drawing here to draw your ideas.
image
The relay drawing here has NC & NO contacts.
 
Cabe
Attachments:
image
Parents
  • 0

    Hi Cabe,

     

    Apparently I missed the discussions on the relay coil switching. I don't have a lot to add, the most is said already but you never know...

    The suggested diode across the coil is absolute textbook proof. The type of diode however might make a difference. A mains rectifier diode (1N4001) is really slow and would not be my choice. I would go for a fast signal diode, like 1N4148 or similar (or with higher rating if needed). The diode is mainly needed to protect the switch that drives the relay. A semiconductor device (transistors are the worst) are killed by high voltages. Mosfet's are more robust, they have a avalanche diode (zenerdiode) protection the drain .

    Your oscilloscope plot might be influenced by the ground loop wire which pickes up a lot of noise. If you measure HF ringing use a short tip probe (or at least wrap the ground wire around the probe tip to have the smallest magnetic loop).

    Interference in your circuit is more likely induced by contact sparks (wide broadband noise) or the coupling between the wires (or tracks) of the switching contact to the rest of the circuit (high dI/dt). Placing an RC across the contact damps the spark and the emitted noise.

     

    Best regards, Enrico Migchels

Reply
  • 0

    Hi Cabe,

     

    Apparently I missed the discussions on the relay coil switching. I don't have a lot to add, the most is said already but you never know...

    The suggested diode across the coil is absolute textbook proof. The type of diode however might make a difference. A mains rectifier diode (1N4001) is really slow and would not be my choice. I would go for a fast signal diode, like 1N4148 or similar (or with higher rating if needed). The diode is mainly needed to protect the switch that drives the relay. A semiconductor device (transistors are the worst) are killed by high voltages. Mosfet's are more robust, they have a avalanche diode (zenerdiode) protection the drain .

    Your oscilloscope plot might be influenced by the ground loop wire which pickes up a lot of noise. If you measure HF ringing use a short tip probe (or at least wrap the ground wire around the probe tip to have the smallest magnetic loop).

    Interference in your circuit is more likely induced by contact sparks (wide broadband noise) or the coupling between the wires (or tracks) of the switching contact to the rest of the circuit (high dI/dt). Placing an RC across the contact damps the spark and the emitted noise.

     

    Best regards, Enrico Migchels

Children
No Data