E = -L(dI/dt)
Hi, iso9001. Is the relay socketed? If so, a good place to locate the diode is across the coil socket contacts. It's hard to believe you can't fit a 1N400X diode across the relay coil, or somewhere close to it, given the size of a typical engine compartment.
This is old material, but it's worth repeating if it helps anyone understand why a diode should be placed across the coil. The equation in the subject line describes an inductors' electrical behavior (similar to Ohms' law, E = IR, with resistance being replaced by inductance, and current with the change in current). The d/dt in the equation means change over time. When you open the circuit, the current will try to maintain itself at its present level until a path is found to close the circuit, then current decays exponentially*. How can this happen when the circuit is opened? According to the equation, E (voltage) goes to infinity. In the real world, without an alternate path (diode or equivalent), an air gap breaks down (with a spark) or, when present, a semiconductor junction does. Current always follows the easiest path.
The minus sign in the equation explains why a reverse-biased diode across the coil shunts the current. The voltage generated by opening the circuit has a polarity opposite the closed-circuit current flow. It's called back emf for a reason. An integrated diode helps protect chips like the ULN2003A from damage, but doesn't help reduce EMI (electro magnetic interference). Nigel pointed out the downside of placing the diode back on the circuit board. To minimize EMI, reduce the size of the current loop. The loop is much smaller with the diode placed at the relay coil contacts than 3' away on a circuit board.
correct me if wrong, but the original ULN2003 did not have protection diodes, I think it appeared with the ULN2003A series.
* the equation doesn't model series resistance, which is where the exponential decay comes in.
btw elfvenlord, welcome to the forum. Nicely done first post!
Hi, iso9001. Is the relay socketed? If so, a good place to locate the diode is across the coil socket contacts. It's hard to believe you can't fit a 1N400X diode across the relay coil, or somewhere close to it, given the size of a typical engine compartment.
This is old material, but it's worth repeating if it helps anyone understand why a diode should be placed across the coil. The equation in the subject line describes an inductors' electrical behavior (similar to Ohms' law, E = IR, with resistance being replaced by inductance, and current with the change in current). The d/dt in the equation means change over time. When you open the circuit, the current will try to maintain itself at its present level until a path is found to close the circuit, then current decays exponentially*. How can this happen when the circuit is opened? According to the equation, E (voltage) goes to infinity. In the real world, without an alternate path (diode or equivalent), an air gap breaks down (with a spark) or, when present, a semiconductor junction does. Current always follows the easiest path.
The minus sign in the equation explains why a reverse-biased diode across the coil shunts the current. The voltage generated by opening the circuit has a polarity opposite the closed-circuit current flow. It's called back emf for a reason. An integrated diode helps protect chips like the ULN2003A from damage, but doesn't help reduce EMI (electro magnetic interference). Nigel pointed out the downside of placing the diode back on the circuit board. To minimize EMI, reduce the size of the current loop. The loop is much smaller with the diode placed at the relay coil contacts than 3' away on a circuit board.
correct me if wrong, but the original ULN2003 did not have protection diodes, I think it appeared with the ULN2003A series.
* the equation doesn't model series resistance, which is where the exponential decay comes in.
btw elfvenlord, welcome to the forum. Nicely done first post!
