Tag Archives: DC Electrical

I let the magic smoke out of a wire *or* the Magic Door (Linear Actuator, Part 2)

So you might remember that when I put in the linear actuator for the door it had a handy remote that could easily both open and close the door.  Now, as much as the remote controller unit was a black box, its exterior functions were really simple – the two wires for one channel either are pos/neg to open or neg/pos to close, thus sending the 12 VDC through the linear actuator’s motor one way or the other to get the desired result.

And, in understanding that, I had thought that I could just put a DTDP switch in place and be able to electronically open and close the door with the switch.  My thought was that I would use the positive power and ground that the controller was hooked to, and that way, the circuit would connect in parallel to the linear actuator’s wires, but bypass the controller entirely, and all would be good.

It was easy enough to hook up, and after double checking the circuit, I tested it.  The door was closed, and I flipped the switch, and the actuator whirred and the door opened, leaving me happy, until I flipped the switch back and as the door started to close, one of the wires from the controller to the linear actuator let out all its magic smoke as the insulation melted away.  I quickly flipped the switch off and examined the system.  All was as it should have been, the polarity to the linear actuator just being put to the opposite wires.

A quick check on the remote showed that the controller unit was still clicking along with the use of the remote’s buttons, but the wire was fried, and the linear actuator wasn’t working.  As I needed to clean up the bus to move it, I quickly disconnected the switch, cut out the damaged wire and spliced the ends.  The controller clicked but again the actuator didn’t twitch.

I tested the actuator wires by making a circuit to the positive and ground, and the actuator slide the door closed, much to my relief.  And then I decided to test out the second channel of the remote.  I quickly wired it to the actuator (with no additional switch in the circuit), and tested it out with the remote.  Again to my relief, the door opened, and then closed again with the remote.  So the system worked, and I buttoned it all up.

So now I’m left with a DTDP switch, which I think may end up running some lights, a controller board that (I discovered after the fact) needs a new wire soldered to it so the first channel will be usable again, and a quandary about why one wire of the controller shorted the circuit while the other didn’t.  But, the door still works with the remote, which is the important part.

A Linear Actuator *or* The Magic Door!

Being that things warmed up and I was trying to get some stuff put away in my shop, I decided to tackle the linear actuator and get it in place.

The linear actuator, extended.
The linear actuator, extended.

Now some of you may be pausing (or going to the search engine of your choice) and asking ‘what the heck IS a linear actuator?’  Well, it is a screw/worm gear drive that pushes/pulls a shaft along a straight line, into and out of the housing.  These are the things that move lots of slide-outs on modern RVs and trailers.

But, you might also ask, if you don’t have a slide-out, why would you need a linear actuator?  Well, one of the problems with our bus, when we got it, was that there was no way to lock it.  The emergency doors had handles, but they had no key-locks though, they could be locked with a padlock on the interior handle or by installing a keyed household deadbolt through the door to interfere with the existent deadbolt inside, which is what we did to the back door.   The front door locked like a dream, as it was one of the swing-lever accordion doors.

The front door, with the swing-latch in the closed and locked position.
The front door, with the swing-latch in the closed and locked position.

As you can see, this is really secure.  Even if someone broke the glass to get in, they’d have to reach up and unlatch the handle by the gear shift in order to swing the door in.  You might also note the cloth by the latch – one of the problems with the door is that the latch has worn and it rattles while you drive.

The swing arm in the open position.
The swing arm in the open position.

Anyhow, you might note the long rod that connects the swing arm in the center to the door.   When the swing arm is moved to open the door, the rod gets pulled in and pulls the door accordingly.  In looking at it in the right way, if the swing arm was a static unit, in order for the door to open, the bar would have to shrink, optimally about 13″ to open the front door fully.  It just so worked out that Firgelli Automations sold not only a linear actuator with the motor integral to the unit (rather than mounted alongside), but also a remote kit.

Now, I chose the higher-powered 12″ model rather than the 14″ one, as I didn’t want to stress the door by pushing it too far, and the bar just wasn’t long enough to take the length of the fully opened 14″ one.  But the high-powered one has a push/pull strength of 150#, and a stopped strength of 300#, which seemed pretty strong as a replacement for the 5/8″ steel bar.

Threaded bar end and the hinge at the door.
Threaded bar end and the hinge at the door.

In looking to replace the bar, I needed to save the threaded end that attached to the hinge at the door, and the collar end that attached to the swing-arm.  These were easily enough cut away with the angle-grinder, and leveled (as best I could).

This is a picture of my 12" linear actuator, open, against the cut bar.
This is a picture of my 12″ linear actuator, open, against the cut bar.

I laid these out against the actuator and things looked pretty good.  But where I had planned to weld right to the actuator, it turned out that the fittings on the actuator were aluminum, so I had to create some pockets of angle-iron that would be the way to attach it.

Parts of angle iron for the connector pockets.
Parts of angle iron for the connector pockets.

These had to be cut down, welded, ground a bit more, welded again, then ended with some flat plate and drilled so that they would attach to the actuator.  As the bar-end of the actuator was not a nice, square piece but rounded as the bar was, and then rounded to allow for a swiveling action that I really didn’t want,  I had to make the pocket that would connect to the threaded rod not only a tight fit, but also longer so that it wouldn’t swivel or flex at all.

The connector cups, fitted and drilled to be attached to the actuator.
The connector cups, fitted and drilled to be attached to the actuator.

So after a little extra work, I had everithing ready to put together.  The connector pockets needed to be painted to keep down rust on the new grinds, and I decided to go with the brass paint, for the fun of it.

Here's the new linear actuator in place!
Here’s the new linear actuator in place!

The linear actuator bar wend in place great, however I discovered that if the round, extendable actuator bar with the screws on it moves and allows you to screw it into the door fitting, you’re actually unscrewing it from the motor and it’s a bad thing ™.  Luckily, the bar screwed right back into the actuator, and all I had to do was detach the other end from the swing arm and then screw the whole bar in and it worked out fine.

The remote, inside the door switch area.  The wires for the linear actuator go out through an existing hole in the bottom.
The remote, inside the door switch area. The wires for the linear actuator go out through an existing hole in the bottom.

I decided to use the space inside the door switch area of the swing arm to mount the remote, which worked well, as it already had a power line and ground screw in place.  The best thing about this placement is that the wires from the actuator don’t hang up on the swing arm when I use that.  With the actuator fully extended, the door operates normally.  With the swing arm locked, it works like this:

True it doesn’t open quite all the way, but it’s enough to get into and out of the bus, and the full extension of the linear actuator closes the door up snug enough that the door handle doesn’t chatter because it’s loose and worn anymore.

Now, all I need to do is get a DPDT swtich to run the door from inside, and it will be perfect.  But for right now, I have a keychain remote door lock for the bus!

Testing out the Backup Camera(s) (Part II, The Testing)

(Continued from Part I)

So with the cameras and the screen, I was ready to test them out. The screen was easy, as it had its own AC adapter.  I was able to just plug it in and it came right on, gave a nice blue screen indicating that it was set for the VGA input, and after 10-15 seconds of finding no signal, it went to sleep.

I could wake it easily by either selecting a new input (VGA->Cam 1->Cam-2->VGA cycle), or by just hitting the power button, but with no input, it just went back to sleep again.

The screen with a real image!
The screen with a real image!

I have a mini-HDMI->VGA adapter for my tablet, but that didn’t work to give a testable signal, so I had to go hook it up to an old XP box.  The booklet manual said that the optimal resolution was 800×400, but the computer’s resolution wouldn’t go down that low.  At the lowest setting though, it was pretty easy to see.

Tiny little icons at higher resolution on the graphics card.
Tiny little icons at higher resolution on the graphics card.

Putting it back up to some 1100×800 dpi (the computer’s regular output setting), the image was still pretty clear, but the text and icons got really small.  I’ll have to play with the setting once I get the bus’ computer up and running.

But with proof positive that the screen was in good shape, I went to checking out the cameras.  The little, cheap camera just had the red and black wires for power, so rigged a plug using a female four-pin power connector from an old computer fan, and hooking it up to an adapter that was meant to power a hard drive. With the VGA connected to the screen, and a button push to cycle camera 1’s input I got … a black screen.

A little clarification.
A little clarification.

At first I checked all my connections, then realized that if I cycled the input again to camera 2, I got a picture. One Sharpie later, I had that system all worked out, and went about trying to get a nice image I could photograph easily, but found that it wasn’t easy to get what I wanted because the picture was, indeed, mirrored (just like I knew it would be but my hands still wanted to turn it the other way).

The little, cheap camera`s eye view.
The little, cheap camera`s eye view.

One of the other issues that people complained about was the guidance lines.  In looking at the view, I don’t see them being very intrusive.  I can, however, see how they form a great fixed reference point for backing up.  For this picture, the camera was at couch height, and the distance to the far wall is ~21 feet.

I hooked the other camera up to the same power source, and put it in a similar placement (just next to the first camera).

The bigger back-up camera's view.
The bigger back-up camera’s view.

The lines are more pronounced on this camera, but they are more colorful.  The back wall here looks closer, but you can see less of the walls, even though the two cameras were at the same distance.

In recognizing this, the cheaper camera has a much greater field of view, but with much more distortion.  At the time, I just noted it, but in thinking about it since then, it seems that this difference will actually help me with placing the cameras on the bus.

But there was another thing to test with the more expensive camera, and that was the IR LEDs.  For that, it was an easy thing, as the photosensor that turns the LEDs on doesn’t need complete darkness – even a good shadow would do it.  So for that, I just held the camera up off the couch (looking at the couch), and lowered it until the LEDs kicked on.

No IR LEDs on ...
No IR LEDs on …
... and with the IR LEDs on.
… and with the IR LEDs on.

The ‘night vision’ works fine, though the bright ‘light’ of the LEDs washes out the colors.  But really, when I’m backing up at night, I think the color of the object I’m getting too close to is less important than if I can see it clearly.

So my plan is now to mount the small, cheap camera up high on the bus (there’s a bevel right above the back dome window and below the clearance lamps that should put it at a great angle), and use it as a regular rear view mirror, displayed on the screen during normal driving operation.  I’ll get a nice view of what traffic is right behind the bus, and a nice wide angle on the sides.  The other camera looks like it will need to be pretty close to the ground for the LEDs to have a good effect, so I’m going to play around with mounting it under or just above the back bumper.  Since I can change between the two views with just a click of the cycling button on the screen, I should be able to get both a ‘big picture’ and then a more detailed view when backing in somewhere.

I’ll post again for on-the bus placement testing and installation …