[TowerTalk] Fwd: Relay lightening protection

[jimlux]

HansLG at aol.com wrote:
>  
> The diodes in ULN2003 are there to protect the switchers in the IC from too  
> high voltage when connected to relay coils etc. It works well provided the  
> "other end" of the diodes are terminated correctly (usually to Vcc). 

Which is the "diode across the coil" model that Tyco recommends against. 
  OTOH, as previously mentioned, it might be "good enough".

>  
> You usually don't need any protection diodes if you use mechanical  switches 
> and don't switch frequently, although protection diodes extend the  lifetime 
> of a mechanical switch.

Yeah.. but I'm not sure that life considerations are an issue here. How 
many diodes do people put across their doorbell switch (driving a big 
old inductor)?  Exactly. Low cycle rate, so if the rated life of 100,000 
operations is degraded by a factor of 10 to 10,000, that's still a lot 
of ringing doorbells.

Where it gets REALLY important is if you're building something like an 
autotuner, as opposed to an antenna switch.  Then, you can have tens of 
operations every time the PTT switch is closed.



>  
> I think you are better off if you put the protecting diode next to the  
> switch as you then take care of the inductance in the wires to the relay as  well.

I would agree.  What you're protecting here is the switching device, not 
the relay coil.


>  
> Normally you don't have to worry about the influence the  diode has on the 
> timing of the relay. If you need the faster release time I  suggest you connect 
> a resistor in series with the diode. Choose a resistor of  the same value as 
> the DC resistance of the relay coil. That's a good start. You  can be 
> sophisticated by adding capacitors etc. to a network around the relay but  I don't 
> think you gain anything substantial in this application.
>  


An RC snubber might be a good thing for other reasons (e.g. suppression 
of transients coming in on the wires from other sources)