See the attached schematic. As I test this out, I will occasionally see the K1-B contact remain closed after I open the "External System Enable" switch. The relay contact will "stick" even if the motor is not running as I open the "External System Enable" switch. So the only current through the contact is that of the power supply. Note that I had run the motors before opening the "External System Enable", I am simply stating they are not running at the time I open the "External System Enable" and see the N.O. remain closed.

My thinking is this relay cannot handle the inductive load, period. As I look at the spec for it and it only lists "resistive" load. But I'm not sure about that. Also note, these motors are NEVER running at the same time.

Can anyone shed some light on what might be causing this to happen? Could it be too low a current from the power supply? I've see other threads stating you need to properly "wet" the relay contacts? Regarding ample load as relay closes, as I understand it.

Relay Data Sheet <--- http://www.mouser.com/ds/2/307/omron_en-g5la-548280.pdf

enter image description here

  • \$\begingroup\$ Could be a deffective switch, or something wrong with the wiring - for example when K1-B closes it feeds the coil, self sustaining. Try to disconnect the 24VDC and you will see that it releaes. \$\endgroup\$ – Marko Buršič Sep 10 '15 at 16:58
  • \$\begingroup\$ Ive tried multiple relays, same result. thanks though! \$\endgroup\$ – becjasl Sep 10 '15 at 17:07
  • \$\begingroup\$ Ok,as said the wiring is incorrect then. \$\endgroup\$ – Marko Buršič Sep 10 '15 at 17:17
  • \$\begingroup\$ I've checked for voltage on the coil while "stuck, there is none. It is not a wiring error \$\endgroup\$ – becjasl Sep 10 '15 at 17:25

When the relay closes, the charging of the 1000-uF/50V capacitor can create a large inrush current -- assuming that the 24-VDC power supply has a large output capacitor that can supply the inrush current. This inrush current could exceed the relay ratings and weld the contacts.

Some possible solutions are the following:

(1) Use a much smaller capacitor value in place of the 1000-uF cap if you don't need the extra holdup time. The LM340T-5 datasheet only calls for a 0.22-uF capacitor.

(2) Add some series resistance before the 1000-uF capacitor to limit the inrush current. Since the contacts have a 10-amp rating, you only need a few ohms to limit the inrush at 24 volts. You may need to check the pulse-energy rating of the resister because now the resistor will absorb the inrush energy: (24V^2)(1000 uF) / 2 = 0.29 watt-seconds

(3) Use a relay rated for a tungsten load, which can handle a high inrush. The inrush current should still be checked against the relay limits (either measure the inrush or calculate the inrush from the capacitor ESR value).

  • \$\begingroup\$ Great advice. This is where I will focus my attention. I will also contact an app engineer from the MFG to discuss. \$\endgroup\$ – becjasl Sep 11 '15 at 12:31

OK, then your relays have reached end of their life. The contacts are gone and when you turn on the device, the tiny contact that is still present heats up and sticks together.
The possible cause probably is that you have switched the relay while motors were turning, this causes arcing, because it is impossible for inductive load to cut off the current instantly therefore the motor winding induces high voltage in order to push the current trough contacts with arcing.
The remedy exists, its called supressor. Simple metal oxide varistor MOV will do the job. You have to solder it in between the relay output contact. Once the contact will open the voltage from motors will rise, at that point the MOV will conduct and it will take the spike off preventing the arcing.

You can cut the protection case of relay and see the weared contacts.

  • \$\begingroup\$ Would this result in intermittent operation? Once "stuck", if I reset power, it doesn't necessarily "stick" again. thanks for your feedback \$\endgroup\$ – becjasl Sep 10 '15 at 18:27
  • \$\begingroup\$ Then I dont get it, you said that when stuck there is no voltage on the coil, therefore no power. \$\endgroup\$ – Marko Buršič Sep 10 '15 at 18:32
  • \$\begingroup\$ Correct, it does "stick". Once in this condition, I then shut all power off (everything drops out and contact opens). Re-apply power, and it works as designed. I can cycle external enable switch and the contact breaks like it is supposed to. It's intermittent. \$\endgroup\$ – becjasl Sep 10 '15 at 18:39
  • \$\begingroup\$ Post a picture of the PCB, measure again the voltage at the coil, make sure your instrument is set to DC voltage. \$\endgroup\$ – Marko Buršič Sep 10 '15 at 18:45
  • \$\begingroup\$ I am 100% certain there is no voltage (DC) present across the coil (or coil to common for that matter) when it is this "stuck" condition. I checked multiple times on multiple boards. It's as though de-energizing the coil is not enough to break the contacts, intermittently. coil is 1600Ω. \$\endgroup\$ – becjasl Sep 10 '15 at 19:01

Are you sure the relay contact is closed, After you deenergize the coil. How do you test? I have seen relay contacts stick intermittently.It happens. Is the relay installed in a socket. If it is, pull relay out of socket to verify. I like the inrush suggestion. You can connect a PTC thermistor to limit inrush. Also there are various snubber circuits to connect around the contacts. I would call the application engineers for manufacturer of relay... Not mouser. Always call app engineers for suggestions. Then your not liable. Set something up on the bench.. power supply ..relay and large cap..

  • \$\begingroup\$ I physically remove the "External System Enable" wires from the board and the PWR LED stays on. Also I still have V+ and +5. Relay is soldered to PCB. Great tip on application engineer. Thank you \$\endgroup\$ – becjasl Sep 11 '15 at 12:25

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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