will caps eliminate back EMF

[Thunderchild]

I have a project that requires occasionally pulsing a latch relay, as the pulses available are quite short I thought of putting a capacitor in parallel with rhe relay to help boost the pulse time and while they are at it would they stop back EMF ? as with the cap the voltage would fall slowly after the actual pulse has ended

 

[MrAl]

Hi there,


It's a little hard to predict exactly what might happen if you use a cap instead of the usual diode or zener or even resistor. It could resonant and cause a problem like that too, but even worse might be that the cap might delay the opening of the contacts so that they dont break as quickly as they should and that would lead to a shorter contact life.

Is there any problem with using a diode or well-chosen resistor?

 

[MikeMl]

In a word; NO! The capacitor resonates with coil inductance to make things worse. Look at the attached. The RC one can be tuned to a specific case, either with added external resistance, or by utilizing the coil's own resistance.

 

[Thunderchild]

ok, well the initial reason for the cap was to make the control pulse more robust, I had read something about caps being used in the past for back EMF so I though maybe I could save the space, it looks like I best use the usual diode and put in a full length pulse

 

[dknguyen]

Not a capacitor on it's own, but a series R-C network (aka an RC snubber) will do the job. It behaves differently than a clamping diode though. It reacts much faster since it's technically always on and doesnt have to switch like a diode, but slows the rate of rise and absorbs the energy rather than clamping it to a preset level. You need a high quality capacitor (a ceramic at the least and preferably a film though this is just a relay so it's not a critical high power application) and a good low inductance surge tolerant resistor (like a carbon resistor).

 

[Thunderchild]

well Iobviosly need to just use a diode and set the correct pulse length in my software instead of trying to cut corners

 

[nickelflippr]

In a word; NO! The capacitor resonates with coil inductance to make things worse. Look at the attached. The RC one can be tuned to a specific case, either with added external resistance, or by utilizing the coil's own resistance.

If I could be so bold and ignorant, what would the combined effect of the diode with a parallel cap in the 10uf and 47uf range look like? This type of combination seems to keep a slow changing signal from contact chatter.

 

[Thunderchild]

well this will be the occasional single 30mS pulse, i think the diode will be the best solution