What about emi - i had some experience, if being more concrete , input selector for amplifier ,made from relays ,had some clicking noise appear at signal patch ,when relay switched on off. Solution was small filter , resistor from supply +12v and capacitor 100uf to ground ,also 4148 diode across winding . It was small signal relay with 5v coil . Regardless off presence of diode ,click was audible through speakers. Maybe it get through ground .
Snubbers or just small capacitor across winding is ok,but if across contacts ,depends on what relay contacts are switching - if low level signal ,snubber is not a choice . Actually if diode is present , voltage spike across winding is small , not exceeds diode forward voltage ,lets say 1V . But it will have sharp edges,so snubber there is a good idea .
With high side switching, at the moment of turnoff, the relay coil node shoots towards minus infinity, and the zener diode conducts current from the +15V supply, through the zener, to the relay coil. In other words the +15V supply sources current into the relay coil during flyback ... exactly as it does when the switch is closed and the relay is engaged.
Thank you guys for all these advices. I found this on stackexchange, and it compares different methods :
https://electronics.stackexchange.com/questions/100212/why-dont-relays-incorporate-flyback-diodes
I would love to know which technique create the least amount of noise and EMI to close signal and power nodes.
https://electronics.stackexchange.com/questions/100212/why-dont-relays-incorporate-flyback-diodes
I would love to know which technique create the least amount of noise and EMI to close signal and power nodes.
All diodes turn on fast enough, not the least because of the high capacitance of a forward biased junction accepting current before the junction fully turns on.
And the diode doesn't have to turn on very well, just clamp the voltage to a few volts. The only real criterion is that the voltage and current handling are adequate.
And the diode doesn't have to turn on very well, just clamp the voltage to a few volts. The only real criterion is that the voltage and current handling are adequate.
A zener plus a 1n4004 is better than just a 1n4004.
With a single diode the current going through the diode makes the relay release slowly and can cause arcing.
With a single diode the current going through the diode makes the relay release slowly and can cause arcing.
The normal solution is just a diode. You add a zener only if you for some reason want a faster turn off.
A TVS diode is another possibility - one device instead of two if its a symmetrical TVS "diode" (which is basically back-to-back zener structure AFIACT) However you have to ensure the TVS doesn't conduct at the supply voltage.
If you want fast turn off, no arcing etc., just use a SSR.
And nobody has posted speed curves of the same relay with different shut off times using different types of diode, or whatever.
It really is spring loaded, which is the main part.
The electronics are to reduce arcing, and protect the coil.
And nobody has posted speed curves of the same relay with different shut off times using different types of diode, or whatever.
It really is spring loaded, which is the main part.
The electronics are to reduce arcing, and protect the coil.
Have you tried to find the FETs and opto drivers for SSRs lately? No stock with 2-year lead times is the rule now, except for older types that are about 3X overpriced.
The shut off time of a mechanical relay depends on how fast you can get the coil current to collapse. With just a diode it can take longer for the current to drop than it takes for the spring to cause it to open. By adding the zener, significant power is dissipated so it has somewhere for the stored energy to go so the coil current can drop faster. Then the relay opens.
The shut off time of a mechanical relay depends on how fast you can get the coil current to collapse. With just a diode it can take longer for the current to drop than it takes for the spring to cause it to open. By adding the zener, significant power is dissipated so it has somewhere for the stored energy to go so the coil current can drop faster. Then the relay opens.
The relay is being switched how?
Transistor, or RC network?
That also decides the open / close time...the diode is there to protect the coils, and also is involved in the working action of the relay.
We get ready SSR here...2 terminals for load, and two for supply, 5-32 V DC.
The opto-couplers and so on are inside the unit, no need to source them.
Look at Unison, and others in India.
Unison also has SSR with mains input, a direct replacement for mains operated single phase contactors, now they are cheaper than the old electro-mechanical contactors.
They may have competitors, just an example.
3 phase SSR are also sold.
Now Omron SSR are easily available, some PCB relays are long delivery!
https://www.omron-ap.co.in/service_support/technical_guide/solid_state_relay/index.asp
These are from a Schneider group company: (looks like they are for DIN rails)
https://www.eurotherm.com/in/produc...-solid-state-relays-in/ssl1-slim-socket-ssrs/
But the basic thing is that drop in SSR seem to be available, and can be considered for such applications.
If you feel like it, search for PCB Mounted SSR.
Even Mouser India has them, and many ready boards are available.
Omron has them in SMD too.
Transistor, or RC network?
That also decides the open / close time...the diode is there to protect the coils, and also is involved in the working action of the relay.
We get ready SSR here...2 terminals for load, and two for supply, 5-32 V DC.
The opto-couplers and so on are inside the unit, no need to source them.
Look at Unison, and others in India.
Unison also has SSR with mains input, a direct replacement for mains operated single phase contactors, now they are cheaper than the old electro-mechanical contactors.
They may have competitors, just an example.
3 phase SSR are also sold.
Now Omron SSR are easily available, some PCB relays are long delivery!
https://www.omron-ap.co.in/service_support/technical_guide/solid_state_relay/index.asp
These are from a Schneider group company: (looks like they are for DIN rails)
https://www.eurotherm.com/in/produc...-solid-state-relays-in/ssl1-slim-socket-ssrs/
But the basic thing is that drop in SSR seem to be available, and can be considered for such applications.
If you feel like it, search for PCB Mounted SSR.
Even Mouser India has them, and many ready boards are available.
Omron has them in SMD too.
Last edited:
Don't remember if someone recommender simplest solution - just resistor across winding. Its widely used in automotive 12V relays , coil draws about 80-120ma at 12v ,and resistor inside relay is 680 ohm .Negative side-effect is increased current consumption ,but speed up relay release in comparision to diode .Also this don't make relay polarity sensitve .Found this tester and comparision ,made with it. And this variants .
Not all relays are made the same, and some relay contacts can exhibit bounce, where the bounce 'performance' depends on the way the coil energy/current is managed by the 'flyback diode' or circuitry used. Contact bounce may or may not be perceived as an 'emi' issue, depending on what the contact is used for. As an example, contact bounce can be a very significant problem for applications such as high current-voltage control, so there has been a lot of R&D and characterisation on that aspect.
Turning off a relay coil may cause a voltage rail current disturbance/step. That disturbance may somehow couple to the audio signal through poor power supply management or control of the relay coil. That disturbance may also couple to audio circuitry via the rapid change in magnetic field of the coil.
The sad aspect of this thread is that the OP has provided no background to why they want this-or-that type of performance, and so are just spraying out naive questions imho.
Turning off a relay coil may cause a voltage rail current disturbance/step. That disturbance may somehow couple to the audio signal through poor power supply management or control of the relay coil. That disturbance may also couple to audio circuitry via the rapid change in magnetic field of the coil.
The sad aspect of this thread is that the OP has provided no background to why they want this-or-that type of performance, and so are just spraying out naive questions imho.
Exactly...
Car relays are sometimes 3 wire, the wires to coil and input are common. Earth switching at times.
Car relays are sometimes Battery (common), Output, and Earth...internally, the supply goes to coil and fixed contact.
Energizing coil makes contact to the output side. Sometimes Earth is switched.
Those are for high loads, like 30A working for headlights and so on.
Switching speed and arc suppression are not critical.
An idea of the end use will be better..
Car relays are sometimes 3 wire, the wires to coil and input are common. Earth switching at times.
Car relays are sometimes Battery (common), Output, and Earth...internally, the supply goes to coil and fixed contact.
Energizing coil makes contact to the output side. Sometimes Earth is switched.
Those are for high loads, like 30A working for headlights and so on.
Switching speed and arc suppression are not critical.
An idea of the end use will be better..
Last edited:
Are you saying the application is to switch low level audio signals using the relay contacts? Are you saying the relay is physically small - can you identify a model? Why are you concerned about 'emi', as emi typically relates to high frequency interference that could be conducted or radiated? Why do you want 'the best' (ie. no matter the cost) outcome - do you have limits on performance or experience on what may not be good performance? It's like pulling teeth on a technical forum.
The application is switching low level signal, that is why I am concerned about psu and coil EMI. I like to learn techniques that are not biased by the cost, it basically allows everyone to explain his preferred technique without being constrained by a limiting factor. As for getting too technical, I like this side of experimenting 😉
Thank you.
Thank you.