Page 1 of 1
Circuit theory to reverse polarity of 12V, <1 amp
Posted: Sun Aug 10, 2008 6:21 am
I am in the thinking stages for a circuit that will reverse the polarity of a 12 volt signal that will draw less than 1 amp. This circuit is for my dad's garden railway and will be used to control a railway "frog" (as show here:http://www.trains.com/grw/objects/images/switches_4.jpg
). This switch is not like the one in the Aug '08 N&V issue, it has only two inputs, positive and negative. The frog moves in either direction depending on the polarity of the inputs (it runs on DC, not AC).
What I was thinking, and I may be way off on this, is to use a 7812 and a 7912 voltage regulator to produce both a + and - output using a common ground (similar to the circuit on pg 36 of the July '08 issue of N&V). The common ground would be connected to one of the terminals on the frog, and the other terminal would be connected to some sort of switching device that would be used to select either the +12V or the -12V line from the regulators. Does this make sense, and would it work?
I want to design the circuit myself as a learning tool, but just want to make sure I'm on the right "track", hehe.
Posted: Sun Aug 10, 2008 8:21 am
It can be done with positive and negative supplies, but a better way is single supply and an "H" bridge. See http://en.wikipedia.org/wiki/H_bridge
for description with pictures. For an electronic H, switches S1 thru S4 are replaced by bipolar transistors or FETs. For 1 amp range IC H-bridges are available, and save a lot of hassle.
For a "frog" you could use a DPDT switch for "H". Does the frog need power to stay in position? If not, add a monentary contact switch for a pulse of power.
Posted: Tue Aug 12, 2008 8:53 am
Hi, I've been thinking about this for a few days, and have looked at some H-bridges. The 7912 is cheaper than any H bridge I've found, although I did pick up a couple to try them out.
The frog just required a pulse of electricity, like what would come from a momentary switch. Why is an H bridge a better design? All my research on them talks about using them for motor control, which really isn't my case. I'm not disagreeing with you, just trying to learn!
Electronic Frogs. Wow.
Posted: Tue Aug 12, 2008 11:40 am
A few questions to help me understand what's going on here:
When the frog is in a "resting" position -- that is, it is not moving -- what is the voltage between the + and - inputs? And the voltage when it's switching? Does the voltage remain the same after the motion is complete, or does it go to 0?
Does the voltage need to be applied for the entire time that the frog is in motion? That is, what happens if the voltage is removed in the middle of the switching?
Does the frog need power when it is NOT switching?
Why do you want to use 5V.D.C. for a control signal?
Posted: Tue Aug 12, 2008 1:16 pm
When the frog is in the resting position, there is 0 volts between the + and -.
The 12V is supplied to the frog for only 500ms or so, maybe as low as 300ms. From the outside it looks/acts like a solenoid, it just needs a quick shot of juice to move it. After the frog moves, the voltage goes back to 0. Think of it almost like a motor that goes in one direction for 400ms to move the frog one way, the motor goes in the other direction for 400ms to move the frog the other way.
The frog does not need, and should not have power when it is not switching.
I made a mistake in my first post, the frog actually requires 12V, not 5V, however, the circuit may still use 5V.
Posted: Tue Aug 12, 2008 1:52 pm
Would not a simple SPDT (center off) momentary switch be the easiest
and most reliable solution? Common ground to the frog ground, positive
and negative to the 2 outside switch terminals and the center terminal to
the frog power terminal. I believe this is the way my old HO switches
were wired. Just make sure the switch is rated for the voltage/amperage.
Posted: Tue Aug 12, 2008 2:59 pm
Yes, that would be the easiest, but it would be no fun. I want to design this for the possible use of a microcontroller or something cool like that.
Posted: Tue Aug 12, 2008 4:23 pm
Okay, how about a dual 555 solution where each is configured as a 1 shot
at 500 ms duration feeding the base (source or sink) of either a NPN or
PNP transistor rated at 1+ amps. That way, all you have to do is feed
each 555 with a pulse to trigger it. There are many 555 designs out there,
and a "Google" (I prefer MSN or Yahoo over Google) reveals page after
These actually feed the rail switch (the moving parts of the rails), not the
frog. The frog is the spot where the intersecting rails form a flattened "X"
that resembles a frog in mid-leap.
Posted: Tue Aug 12, 2008 6:58 pm
An H bridge allows voltage applied to the load in either direction with a single polarity power supply. 5V (or 3.3V) logic signals an control load voltages of 12 or more. Two logic signals are needed, one for direction, the other for on/off.
When I get my domain name renewed, I can post a schematic I did last year.
Posted: Wed Aug 13, 2008 11:21 am
I would use a "electronically-activated bipolar switching device with audio feedback" -- a DPDT relay for polarity selection, and an SPST relay for power control. Set the relay to the polarity you want, then apply power for as long as needed. Use something like a ULN-2003 for the drivers, and stick a diode across the coil of the relays.
We used this method to control vending machines with micros, which is why I prefer it: tried (for me) and true. BTW, the micro ran on 5VDC, the relay on 12VDC.
Posted: Tue Aug 19, 2008 4:07 pm
Ok, I came up with two designs, although I'm not really sure how practical they will be in a real world scenerio.
The first one uses a 5 volt source to trigger some darlington transistors with will then flip the track.
The second design uses just the 7912 coltage regulator to do it.
In both cases I would also have another switch that turns the circuit completly off since I only need power applied for a half second or so.
Let me know what you think.