Convert AC switches to DC switches, was: 12 volt light switch

Posted by Duane C. Johnson on February 22, 2004, 3:22 pm

Hi Douglas;

Douglas Gaulin wrote:

Generally this is true.
You can derate the switch current 3 to 4 times, however
this is not always safe especially if the load is inductive.

I have designed a circuit to add to a standard AC switch
to convert it to a DC switch with the same current rating.

All switches can connect and carry high currents up to the
full current rating whether DC or AC. The problem comes
when the switch is opened.

AC switches have an easy time when opening because the
resultant arc is quenched 120 times per second each time
the AC current passes through zero.

It's a different story for DC switches. The current never
goes to zero. When an arc is established, and it always
arcs, the arc is difficult to quench. Most DC switches
quench the arc by having widely spaced contacts. Some have
a magnet near the arc so the arc is made to curve away
resulting in an increased effective gap between the contacts.

These features tend to increase the cost of DC switches.
Maybe as much as ten times.

There had to be a better way. There is!
Just don't let the arc start in the first place.
With a bit of cheap electronics the switch current carrying
contacts are shunted with a power MOSFET until the contacts
are widely spaced. Then the MOSFET is turned off.

The switches I used were Leviton 15A and Hubbell 20A types.
The Leviton cost \$.25us.
The Hubbell cost \$.59us.
The P-channel MOSFET costs \$.79us at DigiKey.
http://www.digikey.com
Pretty cheap for a high performance DC switch.

The circuit is so simple that I have made a few by just
free form soldering the parts on the back of the switch.
Pretty simple to do.

See my schematic of the circuit at:
http://www.redrok.com/misc1.htm#dcswitch

The circuit can switch nominal DC voltages from 12VDC to 48VDC.
This implies a maximum voltage range of from about 5VDC to
as the AC current rating of the switch. Actually the current
rating can be more because the arc is not established but I
wouldn't recommend this without some testing.

BTW, the AC switch used in the circuit is a 3-way type.
The circuit represents a Single Pole DC switch.

Pictures to come.

Also, is anyone interested in a PC board version?

Try it, you'll like it!

Duane

--
Home of the \$5 Solar Tracker      Receiver
http://www.redrok.com/electron.htm#led3X  [*]
Thermonuclear   Solar Energy from the Sun / |
Energy (the SUN)           \  \     \  /  /  |
Red Rock Energy             \  \     /   /   |
Duane C. Johnson   Designer  \  \  /  \ /    |
1825 Florence St  Heliostat,Control,& Mounts |
White Bear Lake, Minnesota    === \   / \    |
USA      55110-3364                ===   \   |
(651)426-4766        use Courier New Font \  |
redrok@redrok.com     (my email: address)  \ |
http://www.redrok.com   (Web site)          ===

Posted by William P.N. Smith on February 22, 2004, 3:36 pm

That's a neat hack, thanks Duane!

Would a capacitor across the switch contacts also work?

--
William Smith
ComputerSmiths Consulting, Inc.    www.compusmiths.com

Posted by Duane C. Johnson on February 22, 2004, 4:35 pm
Hi William;

"William P.N. Smith" wrote:

Yes, But it has to be tuned correctly for the specific

Kettering's ignition system in old cars is an example.
The way it works is the capacitor absorbs the current.
The voltage across the capacitor rises slower than the
point's gap breakdown voltage increases. Hence no arc.
The capacitor can't be to large or the spark as the
point closes can erode the point.

This is not very suitable for general switches because

Duane

Home of the \$5 Solar Tracker      Receiver
http://www.redrok.com/electron.htm#led3X  [*]
Thermonuclear   Solar Energy from the Sun / |
Energy (the SUN)           \  \     \  /  /  |
Red Rock Energy             \  \     /   /   |
Duane C. Johnson   Designer  \  \  /  \ /    |
1825 Florence St  Heliostat,Control,& Mounts |
White Bear Lake, Minnesota    === \   / \    |
USA      55110-3364                ===   \   |
(651)426-4766        use Courier New Font \  |
redrok@redrok.com     (my email: address)  \ |
http://www.redrok.com   (Web site)          ===

Posted by nicksanspam on February 22, 2004, 5:54 pm

Nick

Posted by William P.N. Smith on February 22, 2004, 7:36 pm
nicksanspam@ece.villanova.edu wrote:

Works for inductive spikes, but not DC current from (say) a resistive

--
William Smith
ComputerSmiths Consulting, Inc.    www.compusmiths.com

•
• Subject
• Author
• Date
 Re: Convert AC switches to DC switches, was: 12 vo... William P.N. Sm... 02-22-2004
 Re: Convert AC switches to DC switches, was: 12 vo... Duane C. Johnso... 02-22-2004
 Re: Convert AC switches to DC switches, was: 12 vo... William P.N. Sm... 02-22-2004
 Re: Convert AC switches to DC switches, was: 12 vo... Duane C. Johnso... 02-22-2004
 Re: Convert AC switches to DC switches, was: 12 vo... Duane C. Johnso... 02-22-2004
 Re: Convert AC switches to DC switches, was: 12 vo... Duane C. Johnso... 02-24-2004
 Re: Convert AC switches to DC switches, was: 12 vo... Bruce in Alask... 03-04-2004