Premature switch failure

[Crwilliams]

I’m seeing several switches, installed new, failing after only 5 or 6 years. They are only used one on-off cycle every 24 h. These are the typical designer (rocker) style 120 vac 15 or 20 amp devices. Installed in climate controlled indoor environment. The common factor seems to be that they are driving LED drivers, which in turn are driving some 100 or so LEDs. The drivers are soft start, and therefore the switches are seeing virtually zero current on ‘make’ and maybe 500 mA on ‘break’ — almost dry switching. So here’s the question: are others seeing this failure mode, and are there any switches labelled for this application?

 

[synchro]

When you say the drivers are soft start, is that because the LEDs ramp up in brightness when they are turned on? Or have you actually measured the transient current waveform as the the switch is turned on?

The first stage of a driver is often a rectifier that feeds a capacitor to smooth out the voltage waveform. It can take a surge of current to charge this capacitance when the supply to the driver is turned ON. Perhaps that could be causing the switches to degrade.

 

[Crwilliams]

The ac input to the drivers is via an active power factor correction circuit. This is a FW rectifier with no significant capacitor followed by transistors which chop the incoming 60 Hz at several kHz. It controls inrush at startup (i.e. there is none) as well as presenting a near unity power factor to the ac supply. Fairly tricky stuff but it’s all handled by some special purpose chips.
Tricky to check in the field but I have put a scope on a driver on my bench and seen it ramp up Since the current through the failing switches is so low I’m wondering if there’s an oxide or similar build up and there’s no contact arcing to clear it away.

 

[synchro]

I understand what you are saying about rectifiers with active power factor correction. Such circuitry is used in "active front ends" of variable frequency drives for motors.

Since the current through the failing switches is so low I’m wondering if there’s an oxide or similar build up and there’s no contact arcing to clear it away.

Yes, the rocker switches may have contacts like silver cadmium oxide that need some minor arcing to break through oxidation, etc.that builds up over time. I don't have much expertise in this area, but the article at the link below discusses some of the issues involved in using various materials for switch and relay contacts.

Relay Contact Materials, Ratings, and Styles

Learn about the types of metals and alloys, sizes, and styles available and how to choose the right one for your application requirements.

www.te.com

 

[ptonsparky]

Interesting.
I've seen a similar issue with small interposing relays. One brand worked consistently, another didn't.
Adding a small load in addition to the drivers should help.

 

[Crwilliams]

I understand what you are saying about rectifiers with active power factor correction. Such circuitry is used in "active front ends" of variable frequency drives for motors.



Yes, the rocker switches may have contacts like silver cadmium oxide that need some minor arcing to break through oxidation, etc.that builds up over time. I don't have much expertise in this area, but the article at the link below discusses some of the issues involved in using various materials for switch and relay contacts.

Relay Contact Materials, Ratings, and Styles

Learn about the types of metals and alloys, sizes, and styles available and how to choose the right one for your application requirements.

www.te.com

Great article - thanks for the reference.

 

[tthh]

I had installed an under cabinet LED strip in a kitchen. Used a Lutron wireless Caseta lamp dimmer with Pico remote. The dimmer was on top the cabinets where there was an outlet and thats where I put the LED transformer. The Pico remote was mounted on the wall in the kitchen. At some point it stopped working, but it was a long time like a year. I called Lutron and they said I needed to program the Caseta lamp dimmer to On/Off mode rather than soft On/Off mode. It was a long time ago, so that may not be the exact language they used, but you get the idea.

 

[Crwilliams]

tthh - that's also really interesting. It looks like there can be a definite problem with traditional 240 v, 15 A switches when used to switch full voltage but at mA level currents ('dry' switching). The article quoted by Synchro, above, is also interesting and very detailed. As more and more electronic controls, especially with power factor correction and soft start, are installed this may be an emerging problem.

At my latest switch failure the LED driver was plugged in to a dedicated outlet so I was able to replace it temporarily with an electric tea kettle. That switched on and off reliably after the first mechanical switch cycle. I then replaced the LED driver back in circuit: the switch worked again, suggesting the heavy load had a cleaning effect on the switch. Regardless, I replaced the switch with a new one. We'll see how long it lasts! (Both the failed and the replacement switches were high quality devices from well known American manufacturers.)

 

[Crwilliams]

When you say the drivers are soft start, is that because the LEDs ramp up in brightness when they are turned on? Or have you actually measured the transient current waveform as the the switch is turned on?

The first stage of a driver is often a rectifier that feeds a capacitor to smooth out the voltage waveform. It can take a surge of current to charge this capacitance when the supply to the driver is turned ON. Perhaps that could be causing the switches to degrade.

Measured an apparently identical unit driver on my bench using a 'scope. Not easy to do on site. The driver has a well defined startup behavior that takes about 3 seconds to run through after voltage is applied and before the LEDs light. In fact this start-up delay has led some users to turn it on, see no light, turn it off again, and try another switch .... leading to an "It's not working" call.