Engine Start Switches with Solenoid

In the real B737, the two Engine Start switches in the Engine Start Panel are solenoid held (spring loaded to OFF). When they are set to the GRD position, the start valve opens, the selected igniter(s) are armed and the bleed air valve is closed. Once 56% N2 are reached, the solenoid releases the switch and it flips back to the OFF position with an rather audible click.

Stock rotary switches obviously do not flip to any position by themselves. It would of course have been possible to manually return the Engine Start switch to the OFF position once the engine has spooled up (that is actually done in reality in case the DEU fails and does not send the cutout signal). But building a replica of the solenoid was a challenge I did not want to walk away from...

 

After an late evening attempt to define the basics of my construction, I came to the conclusion that the only decent way to build a solenoid held switch is by using a solenoid (!).

A solenoid is an electric magnet which - when voltage is applied - pulls a metal rod into its core with considerable force. I found suitable solenoids with 14 mm travel and a tractive force of 0,4 - 18 N (initial to final) at Conrad Elektronik, Germany.

The difficult part of the design process was to find a way for the solenoid to do it's job when required, but not to interfere with the manual operation of the switch.

The 3-pos. drawing to the left shows the solution I came up with (for files click here: Corel Draw and AutoCAD):

A towing arm is attached to the axle of a rotary. A lever and a small linkage connect the axle and the solenoids rod. By rotating the switch from OFF to GND, the rod is pulled out of the solenoid (Pic 1). When the magnet is temporarily powered, it pulls the rod in and the towing arm rotates the switch exactly 30 degrees to the OFF position (Pic 2). A manual reset to OFF would of course also be possible. Manual rotation to the two remaining positions CONT and FLT is completely independent from and unaffected by the solenoid's position (Pic 3).

A cross section of the assembly and a "parts list". All pieces of the towing arm are glued to the axle of the rotary switch. The lever, on the other hand, must move freely on the axle.

The small linkage between lever and the solenoid rod compensates for differences in travel between the two (the lever's tip travels on a circular path, the rod in a straight line).

The main headache while assembling this is to find the right vertical distances between the towing lever, the rotary axle and the solenoid housing.

Assembly has started! I used black and transparent acrylic to make details come out better on the photos.

The pieces of the towing arm are not only glued together but also fixed with bolts and nuts (overkill, I know!), so the unit will not break when stressed.

The towing arm on the axle, with lever and solenoid rod attached.

The towing arm is just slid over the axle for testing purposes. It will be glued in place at a later stage.

An important note about the rotary: turning the stock switch requires more force than the magnets can muster. Therefore, it is necessary to modify the switch slightly. Inside the switch is a spring that pushes two steel balls against a star-shaped outer ring. Open the switch, shorten the spring by 2.5 turns and reinstall everything. There should still be clear tactile feedback for each position, but less force required to turn the axle.

Be warned: You must be rather nimble fingered to do this. The balls like to bouce away (and disappear forever) the moment the switch comes apart.

If you try to pry the switch open with a screwdriver or a knife , you can easily hurt yourself. Besides, the little black plastic notches that hold the two parts of the switch together will brake if excessive force is applied. So do not use any instruments but your hands and fingernails!

The plate to which the rotaries are fixed. Since high precision is difficult to achieve on my work desk at home, I decided to drill oversized holes in the plate and to attach plywood disks to the rotaries. During final assembly (with the completed switch housing mounted behind the OHP) these disks are simply glued to the plate. This guarantees precise - and more important: tension free positioning of the rotaries.

Basically the same thing, from a different angle.

The towing arm is still not glued to the axle of the rotary! That happens during final assembly with a few drops of Super Glue. Be very careful not to glue the levers - they must rotate freely on the axles.

The side walls of the solenoid switch box, with 3 rails:

The one on the left holds the top plate with the rotaries. The rail in the center holds the plate where the solenoids are attached. The rail on the right connects the whole assembly to the left side OHP section of my cockpit.

The window was necessary because of the diameter of the towing arms. Making the box bigger was not an option - my OHP is a very crowded place.

After final assembly!

In an ideal world, the two solenoids would be perfectly parallel. But small imprecision during the building lead to the miss-alignment visible in this shot.

Finding the correct position for the solenoids was a question of shifting them around a bit. Once they were in a place where the rods would travel freely in and out, I marked the position on the plywood plate and than fixed the solenoids with bolts.

 

Another angle.

When I tested the assembly I found out that at their rated voltage of 12 v the solenoids did not have the power to turn the towing arm. They only made the axle of the rotaries twitch pittifully. This must be due to the fact that the starting force of the solenoid's rod is only 0.4 N.

Bigger solenoids would have meant to modify the whole construction, so an alternative solution was needed. It came in the form of a 15 v power supply. At this higher voltage, the solenoids work flawlessly and pull the switch to the OFF position in a fraction of a second. And they do this with a very satisfying, rather loud "CLACK!"

Since the solenoids will be powered for less than 0.5 seconds during engine start, I do not think the overvolting will significantly reduce their service life.

Note the six bolts that hold the solenoids in place.

Another 4 bolts are used to fix the entire assembly to the back of the OHP. The nuts for these bolts are soldered to the rail. There is no way to reach that place with pliers for holding the nuts while tightening the bolts...

 

 

The switch assembly, attached to the back of the OHP-left.

Note the PCB with spider LED's for backlighting. This is the reason why the rotaries are mounted so far from the panel.

 

 

A short video with sound to show the solenoids in action. They are not yet connected to my I/O cards, so this is just a mechanical demonstration.

Size: 1 MB, duration: 24 s, format: wmv.

If the video does not automatically open in your media player, just download it (right click and save target as...)

Close-up of the engine switches, right before installation of the panel in the cockpit.