559
What really happens is that, at the instant the contacts break, the voltage spikes high enough to jump the gap. As the
gap between the contacts continues to widen, the current continues to jump it (helped by the generation of ozone), until
the energy has been dissipated in bridging this relatively large resistor. In other words, the energy stored in the
inductive load has been used to vaporize a few molecules off the surface of your contacts. Meanwhile, the high voltage
spikes haven’t done any good for the insulation in the wiring, and it’ll eventually start to break down.
So, you don’t want too much resistance. If the resistance was about the same as the resistance in the inductive load
itself, then the reverse voltage spike would be about 14.4V -- which the electrical system is clearly capable of handling
without damage. In practice, you can always go with a resistance a couple times larger than that of the inductive load
itself, since the wiring and components will be designed to handle much higher voltage than 14.4V.
Note that the amount of energy dissipated in the resistor itself is exceedingly small and brief. Like the diode itself, the
resistor can be tiny. A 1/4-watt resistor or even smaller will do.
Keep in mind that installing a resistor with the diode only makes sense on inductive loads where it is important to
quench the magnetic field quickly, such as on relays. It shouldn’t be necessary on electric motors, since the momentum
of the rotor will keep it rotating several orders of magnitude longer than the effects of inductance will last. Similarly,
the A/C compressor clutch probably doesn’t need to disengage suddenly, since it will spin for a bit when disengaged
due to momentum anyway, and the coil releasing the engagement a couple of milliseconds more slowly probably won’t
make any difference.
Now, note that if the resistor is high enough ohmage and high enough wattage, you can leave out the diode and simply
wire a resistor across the coil terminals! This will waste electricity, because whenever the power is on to the relay coil
it will also flow through the resistor with no diode to stop it. The resistor now needs to be large enough to handle that
current on a continuous basis rather than just the instantaneous spikes -- but resistors are cheap, so this is plausible.
While a resistor with a low enough resistance to suppress the reverse voltage spikes on a large inductive load would
waste a lot of electricity, a resistor across a relay coil could have high enough resistance that the wasted electricity could
be considered insignificant. The scheme has the advantage of making polarity unimportant, so power can be wired to
the coil and resistor in either direction without blowing out a diode.
Bosch appears to have adopted this strategy on later relays, building a resistor right into the base of the relay. One such
relay, having a gray plastic housing and number 0 332 204 159, shows a device wired across the coil in the schmatic on
the case but doesn’t indicate what the device is; it’s shown as a simple rectangle rather than a resistor or diode symbol.
However, a dissected relay revealed that the device was a 630Ω 1/2-watt resistor. At 14.4V, this resistor would waste
only 0.023 amps. Such a resistor would theoretically permit a reverse voltage spike of about 107 volts as connected
across the 85Ω coil in the relay. Presumably, this spike is considered a reasonable balance; a lower ohmage resistor
would permit a smaller spike but would waste more electricity. Using a diode would permit use of a lower ohmage
resistor for a smaller spike without wasting any electricity, but requires care in connecting to observe correct polarity.
There are other ways to deal with reverse voltage spikes in inductive loads, including installing zener diodes (which
will “clip” the voltage spike at a particular level) and capacitors (which will absorb the energy and then send it back
through the inductive load, resulting in a resonance wave back and forth until the energy is dissipated by resistance in
the winding). Craig Sawyers likes the zener diode: “This might actually be an improvement over the resistor, for the
following reason. The reverse voltage is in direct proportion to the rate of collapse of current. With a resistor, the
voltage drops as the energy is dissipated, giving rise to an exponential fall in current and voltage. With a zener, the
current will collapse faster, because it will attempt to hold a constant voltage of (say) 36.7V until the current is truly
minute (microamps).” Zener diodes are cheap, but a bit more difficult to find than simple diodes and resistors.
One application where the simple diode won’t do is on reversable motors, such as electric windows, electric seat
adjusters, automatic antennas, and electric mirror adjusters; the diode will conduct when the motor is running one
direction. One possibility is to wire in a resistor without a diode, as described for relays above; since most reversable
motor applications are only operated for a few seconds at a time, the wasted electricity can be considered insignificant
even if the current through the resistor is significant. You could actually put the diode back in if you want, just to
eliminate the waste of electricity in one direction.
Perhaps a more esthetically pleasing solution would be to install two zener diodes, each with a zener voltage somewhat
greater than the 14.4V that the motors run at, wired in series but arranged with opposite polarity across the motor