best way to reverse polarity on solenoid

HI "Iamanidiot",  Do you have a "Complete" schematic for this. I am looking for the lowest parts count and hopefully 1 or PULSE activation (from a momentary switch perhaps) .   Again we are locked into a single coil latching solenoid @12 volts on control side.     My understanding of the DPDT (double pole double throw) switchs is that they connect either A and B terminals or they connect C and D ,  (thus you can get two separate circuits (Either AB or CD but not both at same time).    If you tie A and C together  you get a SPDT switch..  By using this as SPDT you can apply + for example to two different places (but only one at a time).   

None of this actually reverses BOTH NEG AND POS polarity (both neg and pos must be reversed to activate the solenoid from one position to the other since it is a single coil type. )

Wonder if what is really needed is a   4P2T switch with an at rest position so that when control power is dropped it doesn't go back to the other position? (or use 2 DPDT's ganged together, higher parts count again). 

Guess we were hoping that someone had already figured a way to get it to one pulse with some neat gimmick method with low parts counts..  one guy mentioned using diodes to reverse the polarity?  any info?

I took the liberty of completing the drawing - I am an idiot I hope you don't mind me adding to your initial drawing;

Thanks.. I am getting the idea..  THANKS!   Let me add an additional twist to the equation for more fun !!

Ideally we would like to use our microcontroller to use only ONE port to send one pulse the  large single coil latching solenoid XX (100amp72volt) to make it go from position A to position B.      Of course this is very hard to do with existing microcontrollers.  So we then fall back to using intermediate relays.. BUT we would like to have no power whatsoever on anything once the big XX solenoid latches up (either A or B position).   Do you have a way to eliminate all power on control side .. 

It looks like with the DPDT system you named, it would still be feeding current to the big XX solenoid which because it is a latching big solenoid doesn't need control power any more. 

In other words a DPDT relay with a center "rest" position would be nice .. so that no control power flow  except when pulsed from microcontroller (MCU)  (its a 5 volt pulse which we can very easily with a mosfet boost up in voltage and amps from micrrocontroller). 

Failing a one pulse system ( using no control power except when pulse come thru), a 2 wire system would be next best (uses 2 ports (wires)  from MCU but probably only 1 would pulse at any one time).  

Just so we all know.. the insde the relay schematic for a single coil DPDT relay is below.   1 and 8 are the control power .   3 and 6 are the +  and - for the main battery supply respectively .   Connect 2 motor leads to 2 and 7 ;  and  connect with external wires terminal 2 to 5  and also  connect 4 to 7;  that gives polarity reversing when relay is actuated.  

  The at rest position is shown;  so motor terminal 2 is + and motor Terminal at 7  is  - .    

When the power at 1,8 is reversed. the contacts are then 3 to 4 which is also connect to 7 making the motor lead opposite of what it was (now it becomes +),  same for other side  6 connects to 5 which is also connected to 2 (making it -) which is opposite polarity of previous.

Thanks to Iamanidiot's comments. 

A momentary switch just applies voltage momentarily.  This means when you release the switch, it goes to the off position.   This new twist you bring into the equation would have really been helpful in the initial post.  Since you are not intending on using a manually operated switch, I have wasted tons of tme trying to help you understand something that is totally useless.

Yes I do know how to eliminate all power drain on the control side.  There is none.