5
\$\begingroup\$

relay in situ circuit diagram

Here I have an automotive relay (from a scan of a workshop manual I'm afraid) which is rather more complexly wired up than the relays I'm familiar with. In this diagram:
L/Y - +12V
L - From the relay's switched live (a switch in the car)
L/B - Gets grounded if the car's ECU 'allows' the relay to be on
B/G - holds high if another condition is met (engine temp too high), to prevent the relay energising. (As far as I can tell...this is the correct logic for sure, not sure if it technically works that way!)
B/L - the load (2 loads in parallel)
8(?) - goes direct to ground

I understand how the relay works at the moment. However, what I plan is to cut (L) and insert a device which will prevent the relay energising unless some further conditions are met. This is achieved with some logic gates and an output transistor, which will 'join' L when my device allows it.

So far so simple?

What I can't tell is if I need a protection diode across the relay to protect my transistor? And if I do, would this interfere with the logic provided by the combination of L, L/B and B/G?

If this is a problem, would I be better using a small relay? Are these more immune to the relay closing?

\$\endgroup\$

2 Answers 2

4
\$\begingroup\$

The resistor across the coil is a snubber, which will dissipate the extra energy and make sure the coil voltage never exceeds I*R, where I is the relay's "on" current.

Depending on R and I, you are completely safe, or you might see a controlled low-going transient at relay switch off. Notice high-going spikes won't be seen outside of the relay -- the series diode prevents it.

I wouldn't worry about it. You're not going to see the kind of uncontrolled spike that you get when a bare coil is looking for a place to dump its energy.

\$\endgroup\$
5
  • \$\begingroup\$ Thanks for this, that will make my life much much easier! I figured that resistor must be some kind of protection but never heard of a snubber...every day is a day at school. :o) \$\endgroup\$
    – Jon
    Jan 4, 2013 at 15:54
  • \$\begingroup\$ Snubbers are used almost everywhere there's an inductive circuit. Get used to them :) \$\endgroup\$ Jan 4, 2013 at 17:52
  • \$\begingroup\$ If the resistor value is N times the DC resistance of the coil, it will clamp the voltage negative voltage to N times the voltage seen by the coil. Further, if the DC current through the coil is I, the resistor will waste additional current I*/N. Unless *I is so small that doubling the relay's current requirement would be no big deal, N is going to be bigger than 1. I would expect the resistor to prevent thousand-volt flyback, but I would not rely upon it to keep flyback below 100V unless I measured it. \$\endgroup\$
    – supercat
    Jan 4, 2013 at 17:54
  • \$\begingroup\$ The issue I have is that I can't see a way to add a flyback diode across that particular relay, without connecting B/G to L (B/G can be high when L is low). I am new to relays (and most things) so I'm sure I might have missed something. \$\endgroup\$
    – Jon
    Jan 5, 2013 at 0:14
  • \$\begingroup\$ Is there anything in the fact that there's already a diode between L and the relay coil? Excuse my ignorance, but wouldn't that give flyback protection? \$\endgroup\$
    – Jon
    Jan 5, 2013 at 0:16
1
\$\begingroup\$

For a one-off project or prototype, markrages' answer is probably correct. According to the schematics, the specified relay has a snubber resistor that should typically be enough protection, although you'd have to measure to be sure the voltage stays within the maximum ratings of your transistor.

Cars and their electronics are often subjected to some pretty extreme electrical and mechanical stresses. Those stresses could cause the snubber resistor to fail before the coil does, which would cause your circuit to go down along with it if it has zero protection of its own.

Also consider that automotive relays are installed in sockets and are easily replaceable, even by end users. It could be replaced with an incorrect type (or cheap knock-off) that lacks the built-in snubber resistor.

With all that in mind, the right answer really depends on how rugged and reliable you need your circuit to be. If you're designing a one-off gizmo for your own car, by all means leave out the diode if feel so inclined. If you are designing for a mass produced device, adding the protection diode would be a sensible choice. It would make your product a lot more reliable at a pretty insignificant extra cost.

\$\endgroup\$

Not the answer you're looking for? Browse other questions tagged or ask your own question.