I have an older car with a known issue with the window switches: they carry all the motor current, and after a while the contacts suffer carbon buildup presumably from flyback arcing. Commonly, owners will add relays (one for each direction of each window) to spare the switches, but wouldn't it be simpler to add flyback diodes?

If so – since the motors can be driven in either direction, so a regular diode can't be added at the motor – is it better to use two regular diodes at the switch (addressing the flyback arcing where it's occurring), or two Zener diodes (or similar?) at the motor? And in each case, what would be an appropriate diode choice? (The circuit is ~13V, protected by a 30A fuse, and the resistance measured at the switch connector is 1.8Ω for driver side and 2.4Ω for passenger side, for a calculated 7.2A and 5.4A respectively.)


simulate this circuit – Schematic created using CircuitLab

I have found references to adding flyback diodes in this application, but only to the switching coil of the added relays (which I know some relays even have built in), not the actual window motors.

Thanks in advance for any and all answers!

  • \$\begingroup\$ Bottom antiparallel zeners looks good. \$\endgroup\$
    – winny
    Oct 15, 2017 at 20:50

1 Answer 1



simulate this circuit – Schematic created using CircuitLab

Figure 1. (a) Snubber with four diodes. (b) Snubber with bridge rectifier. Note that the two solutions are electrically identical.

The Zener diode solution looks fine. Diodes can be used as shown in Figure 1.

How it works:


simulate this circuit

Figure 2. (a) and (b) show the window being powered down and then the down switch goes open circuit while the contacts transition. (c) and (d) are perhaps less likely (depending on the switch construction) but shows the down contact breaking while the window moves up.

  • When studying inductor circuits I find it easiest to think that the inductor doesn't like change in the status quo. If the current is zero it takes a while for current to build up when voltage is applied. Similarly, once current is flowing it tries to keep it flowing - even if the voltage is removed. This is the essential concept to grasp: if the current is interrupted the inductor will do all in its power to raise the voltage across the gap high enough to allow the current to continue. (The act of doing this uses up a lot of energy so it doesn't last long.)
  • (a) shows the current flowing from left to right when the down button is pressed.
  • (b) shows the situation the instant the down contact is released; it is open circuit for an instant as the contact moves through the air back to the rest position. The current has only one way to go - through D11 - and this closes the loop neatly back to the source of the current.
  • (c) shows the reverse current direction when the up switch is pressed.
  • (d) shows a slightly contrived situation that the down switch is pressed at the same time. It goes open circuit. Now the only way to keep current flowing is for D20 to conduct. In fact this might not be as contrived as it first appears if, for example, there are up and down limit switches in the circuit.

What is the function of D7/D8 and their connection to the ground rail?

See (c) and (d) above.

Would it be problematic to implement D5/D6 at the switch, given its ~8ft distance from the motor?

No problem. That just adds a little resistance to the wiring. It is the switches we're protecting anyway.

  • \$\begingroup\$ I like the solution with the diodes/bridge! The zener diode solution would require a quite large peak power rating. One could also use a series of C-R in parallel to the load. \$\endgroup\$
    – next-hack
    Oct 15, 2017 at 20:58
  • \$\begingroup\$ @transistor Thank you for the reply and diagrams; it's taken me a minute to process! If I understand (a) correctly – when the "Up" circuit is broken, D5 takes the flyback from the left side of the motor to the right side via the + rail, and when "Down" is broken, D6 does the reverse. Is that right? What is the function of D7/D8 and their connection to the ground rail? Would it be problematic to implement D5/D6 at the switch, given its ~8ft distance from the motor? \$\endgroup\$
    – Luke M
    Oct 17, 2017 at 22:11
  • \$\begingroup\$ See the update. \$\endgroup\$
    – Transistor
    Oct 17, 2017 at 23:13
  • \$\begingroup\$ @transistor Thank you again, your explanations are very clear! Re (d), I hadn't considered limit switches, but I don't believe I have any (none are shown in the wiring diagram p.44 and the motors draw current even after the window hits its limit), and up/down can't both be pressed at once, so perhaps the second pair of diodes are not needed? That said, the released switches connect to the negative ground, not the positive, so I think it's the latter two diodes I would need (like so), right? \$\endgroup\$
    – Luke M
    Oct 18, 2017 at 2:57
  • \$\begingroup\$ I think you've got it. \$\endgroup\$
    – Transistor
    Oct 18, 2017 at 5:06

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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