Projects


Bill's Toys

Genie Man Lift is further down the page


So;     Is it vintage, antique, or junque?

A while back, I acquired a ‘73 model Avion 31' LeGrand.  The purchase was originally intended  to serve as a field office / work space with sanitary facilities and a place to crash close to the job.  Hotel bills were eating up my margin on extended projects.  When I got into the walls, I discovered that I had found something special and couldn't bring myself to take a "sawz-all" to it.

I'm not into restoration, as such, but I do figure to rebuild the trailer into a servicable, pleasant living space.  Since I'm not doing a restoration, appearance is secondary to reliability.  I want the wheels and bearings in like new condition, and the lights up to and better than any state DOT requirement.  I don't care beans about their silly laws, I want this thing well marked at night for my own safety.  I'm allergic to pain, you see.  New brakes, new wiring, and a place to work and sleep.

So, why retrofit with LEDs?

From a standpoint of reliability, LED lighting is far superior to incandescent lamps.  And, less electrical load on the truck.  But, as with any retrofit, there ain't no free lunch, as Lazarus Long is so fond of saying.  It's  ‘spensive, big time, in cost or in man hours.  And, there isn't a retrofit available for this particular model, anyway.  My other equipment has Bargman 4x6 lights and were plug in replacements.

The Avion has 4 point something inch square lights. About four and a quarter, give or take.  Rebuild parts are available from  www.vintagetrailersupply.com   Both lenses and buckets  are available at very reasonable prices.  But, it's still lamps, not LEDs.  So, being the stubborn type, I live up to the title "Artificer" and figure out a way to make it (them) work anyway.

Outlined here is my approach, two different solutions actually, one based on parts on hand and another based solely on least cost.  The latter comes out at little more than the cost of a couple of  1157 lamps, per light.  But, more man-hours, and some little finagling.

First, the technical reasoning..... Yes, I'm an engineer.  And yes, I write like one.  Read it anyway, it'll cast some light on an obscure subject. (grooaaann)   Look at the inside of a stock tail light lens.  In addition to the red plastic cover, with lots of ridges, there is usually a reflector of some sort.  What light is projected from the back side of the lamp is reflected through the lens, with a different focal length than that straight from the lamp.  The lamp sits back from the lens a controlled distance.  Just as the lamp in a slide projector is calibrated to the lens.  The ridges on the lens project the light in a controlled pattern, with a wide angle of dispersion.  This effect causes the lens to be visible at a very wide angle, almost 90 degrees off center on some newer designs.  This, of course, depends on the hot filiment being near center behind the lens.

Now, think of a light house.  The lamp in a "first order" light (the largest) is only 1000 watts, equivilent to five big shop light bulbs.  Yet, it can be seen for 30 miles and more at sea.  With a controlled pattern.  Similar to Morse Code, actually.  These lenses are  over six feet across and as much high.  I can lie flat on the bottom of such a lens.(And have, but not for long.  The sun was hot!)  Hundreds of tiny slivers of glass set "just so" in a frame.  Before electrification, there was no more than a kerosene lantern to project that light for dozens of miles.

To my knowledge, all the first order lights in the U.S. date from the mid 19th century or later, with the lenses fabricated in France.  They are so precise that no optics lab here at the time, could do the work. Check 'em out at http://www.lighthousefriends.com/light.asp?ID=347.  This link jumps directly into St Augustine light.  The one where I tried to nap inside the lens. And, for what it's worth, I'm the electrician that added the sun relay.  Though, I like to think the painting on my home page speaks clearly enough.


A tail light lens is but a miniature version of the same technique.  Using a (relatively) small lamp to create a cone of light of far greater intensity than the lamp alone could generate.

Now, look at a universal LED trailer light kit.  Forty bux, most anywhere.  One with a dozen or so LEDs.  There will be a miniature lens around each LED.  Remove the cover and hot the brown wire (+ tail light circuit) and white wire (- ground).  Use reduced voltage, if you like.  Makes it easier to see what is happening.  My bench supply is 8 VDC. Just about right.

The LEDs are tiny, little more than a pinprick of light.  Place the lens over the LEDs and the apparent size is as big as your thumb nail.  And lastly, put a standard tail light lens over the LEDs.  You can barely see them.  Close up.  At 100 feet, they're hardly visible.  On a dark night.  Like a firefly... And, when you come right down to it, that's where they are most important.  Be seen before the driver behind gets too close.  If they are only ten feet back, all they need worry about is that 4 inch, schedule 40 steel tube bumper, anyway.

So, my solution; use the LED lens.  That's what this is about, making the proper lens fit and look right on the Avion.  The way it actually works is the LED lens is inside the stock lens.  And it works well; the further back I stand,  the better the light.  At 100 feet, they are stronger than 1157 lamps behind a stock lens.

Note: I do need to point out that this project will disable the back up lights.  In my experience, the stock backup lights are next to useless, anyway.  When the truck back-up lights reflect off the front of the trailer, I need some serious lights aft.  On my trailers (several), I use tractor lights from a Co-Op for work lights.  The "trapezoid" pattern makes the best work site light. I have steering diodes so some of the work lights are also on the truck's back up light circuit.

For this project, I dropped a cable from the port side tail light down to the frame. 
Attached to large spring clamps, the big lights clamp to the bumper rails when I'm pulling this trailer.  That way, they can be relocated for use as exterior work lights when the generator is running. Or, very important, redirected to point low for spotting in a campground late at night.

Be comfortable with what you are doing...

You will want to look things over before starting this conversion. 
First, open up one trailer light and examine the lenses.  Of the two, the back up light is deeper than the tail light, by a significant amount.  Note also, the square clear insert for white light.  The lens ridges are a different shape.  The buckets are the same size on both lights, the clear lens is a greater distance from the lamp filiment.    We will use this to our advantage.    I highly recommend acquisition of two more back up light lenses.  For a number of reasons....  They are "Monarch" number 9075.  (The red lens is 9074)   Check with the link listed above.

Next, make a trip to the local "home center".  Lowes and Home Depot are the two in my area.  You want the plastic used as replacement for "single strength" glass.  About 1/16" thick.  The type you want is a poly-carbonate.  Acrylics such as Plexi-Glas or Acry-Lite are cheaper, but are harder, more brittle, and prone to cracking if they are flexed near a scratch.  The most likely material will be GE's Lexan, although I have seen other brands occasionally.  This material is readily worked with wood working equipment, saving considerable time.  One square foot will do four lights, though I doubt you will find a piece that small.  Get a half sheet, it's great for windows as well.  Don't worry about the clear, it's only a mounting plate.  You could use sheet metal but it's harder to work with.  The whole assembly will be covered by the stock lens.

Tooling: In addition to the #2 Phillips, you will need a hacksaw and a coarse file.  A "Sharpie" will work as a marker. Clean it off with ispropyl alcohol.  For the hacksaw, see the photo above.  This is a freehand blade holder that will be very handy for cutting the plastic.  Grind a notch in one end of a blade, as pictured.  A hand drill or a battery drill with the slow speed screw driving option.  An electric drill is usually too fast for this sort of work.  Drill size will be dependant on screw size.  I used 4-40's, they just looked right.  6-32 will do as well.  For #4 screws, use a 1/8", for #6's use 5/32". A little over sized, in both cases, for wiggle room.  And a 3/16" for the #8 sheet metal screws that hold it all together on the trailer.  For cutting the inside holes, I used a jig saw clamped upside-down in a vise.  Use a metal cutting blade (small teeth) and run it as slow as possible.  And, of course, a good utility knife, with a new blade.

Building the gizmo

If you bought the usual universal tail light assemblies, they will be five inches, more or less, sorta kinda square.  With a round area in the center somewhat over three inches in diameter.  This area will cover the LEDs and have the mini-lenses for each one.  This is the part we are interested in.  Keep the side light lenses and gaskets, they come into play later.

Remove (screws) the side marker and tag light assemblies and cut the wires as close to the potting of the main circuit as possible.  I recommend they be staggered a little to prevent shorting together.  We will splice on wires and reuse the side lights later. 

The tag light block is too large for the Bargman lamp holder on the trailer.  I made an LED circuit to fit.  If you want, leave the incandescent light as is. 



To prepare the lens, see photos left (2).  Start by cutting off large chunks until it is down to size then file it smooth.  I cheated, using a benchtop 6" belt sander and 80 grit belt.  Take it easy, the plastic will stick to your skin until the curl hardens.  When the lens is nice and round, cut down the edges until it is at about half thickness.  The target is to lay it flat on the Lexan and have the stock back up lens cover it flush.

Next, take the light assembly and cut it down the same way. as in the  photos(2) above.  In "a", the last corner is being cut off.   At "b" is the desired end result.  Note that there is a rib that is part of a circle around the circuit board.  You will want a 1/8" to 3/16" lip outside that rib.  An inch or so at each corner is sufficient, the middle part can be shaved off.  Also take a quarter inch or so off each end of the rib.  This will make the adapter ring wider on the sides; less likely to break while working it.

Using the stock Monarch lens, layout the rough cuts for the Lexan.  Or use one of the buckets.  This piece of plastic will replace the bucket flange.  I made the first one by hand.  It's obvious in the photos.  The rest I cut on Wife's band saw.  She wasn't looking.....  Good thing I got the chips swept up quick.....   If you lack the band saw, use a hack saw with a 24 or 32 tooth blade.  Go easy, the plastic will crack if you get in a hurry. 

For roughing, score the plastic to a straight edge, as pictured here, several passes.  I used a saw blade as a straightedge ‘cause I was in a hurry and it was at hand.   Score half way through and bend at the score.  The handle shown here is a stab saw that uses regular hacksaw blades. A utility knife will cut the plastic well enough, just go for small bites.  Then file to shape.  Target size and shape is flush with the Monarch lens. Believe it or not; for a straight cut this method is much faster than the jigsaw.

This photo shows the jigsaw set up I used to shape the plastic.  To cut the inside opening, I drilled 1/2" at each "corner" with a paddle bit then shaped the inside with a saw blade and a file.  (Numbers 2, 3, and 4 were made with the jigsaw.)  The target shape is to fit around the rib on the light assembly.  Drill the mounting holes 3/16" located by the lens or bucket flange.

After thoughts: The lens is a hard plastic, probably an acrylic.  Treat it gently if you clamp it up in a vise.  If it cracks, it is still usable, provided there isn't a piece gone.  The light assembly reminds me of a self lubricating plastic called Delrin.  I see it often in model building.  It works easily with a knife and a hacksaw is like cutting soft wood.  The LEDs, on the other hand, are extremely hard and brittle.  I broke one bumping it with a metal tool.  It went dark...  The LEDs are in series circuits, 3 in each.  I shaped a 5mm "lunar" LED to fit flat and soldered it in.  That one is dim, the other two work correctly.  The Lexan is harder than the Delrin, but far less than the lens.

The photo left portrays the three pieces individually and as an assembly.  There are some notes so you will want to see it full size.  Looking at the LED assembly, there are a number of holes in the circuit board.  I don't know what they are for but presumably this assembly is used in a number of applications.  The four I used are shaded in the photo.  Drill 1/8", centered in the hole.  The potting compound is not a hard set epoxy, it cuts readily with a knife or sharp drill.  Take care not to cut the circuit board.   Clamp the lens to the front of the LED assembly.  There are registration tabs to align the lens to the LEDs.  When you have it set, file a notch in each piece so you can reassemble it quickly.  Use the circuit board holes as a guide to drill the lens.  Test fit with screws to hold the assembly together, sandwiching the Lexan between the lens and the LEDs.  Then place the Monarch back-up lens over the outside to assure the screws aren't hitting. 

If things don't close up properly, you may need to take a little more off the lens back.  See right. Don't take a lot.  If the lens is too close to the LEDs, we lose part of the magnification.

Sealing probably isn't necessary but; I've been an electrical man for over 40 years.  It gets sealed whether it needs or not.  The actual electrical assembly is potted and rated as submersible.  Our concern is moisture in the lens distorting the light and leaving ring around the collar.  Did I just date myself?  Yeah, well......  You might do well to go out to the trailer and dry fit in the housing.  Once everything is sealed, it's a lot of work to back track.

When you are happy with the dry fit, break out the clear silicon goop.  RTV wasn't tested, I don't know what the solvent might do to the plastic.  Goop up the back of the Lexan plate and place it over the circuit assembly.  You did make the notches I suggested???  Now put a bead around the back of the lens.  As you place the lens, twist it a little back and forth to smear the sealant and fill any gaps.  It looks messy now, but will be covered.  Place a dab on each screw and snug them up.  They needn't be tight, just enough to hold everything together.  Let it sit for a day or so until the goop sets up completely.  I gave it 36 hours.  Note too that the assembly needn't be perfectly square.  So long as everything closes up tight, and fits the housing, that's all that matters.

Side lights:

This was the most traumatic part of the project.  Here I have this 35 year old classic, albiet well worn, but with near perfect skins.  Only a few stone dings and needs polishing.  Hey, in a steel town, everything gets nasty, quickly.  Steel mills, you know.  Now, I have to cut holes in this elegant form.  Well, it comes back to the legal aspects. But there is, once in a while, though not very often, rational thought behind them.  One more side light is maybe one less 1d10t I have to worry about.
 

I happened to have a piece of aluminum with rolled corners.  See photo left for how perfect a fit it made.  A small piece of flashing bent in a vise is as good. No weight to speak of.   I did have to cut down the mounting feet by about half.  Wanted to keep the rivets under the lens gasket.

In this view, I had marked the outline of the lens and cut the opening with a drill and hand  nibbler.  It obviously isn't square, doesn't matter.  Clecos are God's gift to the sheet metal man.  I prefer this screw type over the spring type.  Sometimes, fits are a little iffey and I can get heavy handed with these things.  

Here is of the inside with the lens in place.  The photo says all that needs to be said here.  But, I'm still upset over cutting that hole



Lastly, this shot shows a completed tail light alongside a stock light for comparison.  The extra back up lenses are ordered and will be used with the budget version that follows. If you look closely, you can see the thickness of the plastic mounting plate as well as the LEDs through the clear portion of the stock lens.





Budget LED Lighting
(Doing it on the cheap)


Since I posted this, I have found the same light, with both tail and brake circuits, at the Big Blue Box for around $12.00.  Construction would be the same, but without the electrical modifications.  If you have the electronics parts on hand, there is money to be saved.  If you need to buy them ,  there would be no gain.  It's still half the price of the large version.

Photo 1 shows the part I used as a starting point.  There are numerous sources for these things.  Price is the only factor. When you have acquired the receiver brake light package from that well known tool distributor at a discount price (it was on sale for $7.00),  take it apart.  The two pieces of interest are shown in the photo below. 

But, you think, only two wires.....  Stay with me, that's the finagling I was referring to earlier.  Let's make the fixture first, then we'll tackle the electrics.  For me, the easier part.... Take note of the square notch on the left side of each piece.  It will be a reference point when you start grinding on the lens.


Using the same techniques as the larger version, we need a smaller hole this time.  Maximum size is determined by the LEDs on the circuit board.  But, this time, we will be using standoffs, so accuracy isn't crucial.   Both plates were clamped together and worked as one.

The fit is shown here.  This is the first pass, I had to adjust the spacing for the LEDs to the lens with standoffs. 



Here is the completed assembly. The stand-offs are 1/4" long.  A piece of plastic would do as well.  These things happen to be a stock item in my shop.
  Note that the lens is cut down further than the first version.  The notch I mentioned earlier is the depth guage for how much to take off.  It was a matter of matching the lens position to the LEDs.   Just grind to the bottom of the notch and use 1/4" spacers.  The screws are 4-40, 1/8" holes, drilled through the lens just as before.  Sealing is the same, except this assembly is open on the back.  Not submersible, so we won't want drips on the circuit board.

Now, the really interesting part.  Assuming you used the same part number as shown in the first photo, there are twelve(12) LEDs.  Three groups of four in series.  Looking back at the photo of the basic circuit board, there are three diodes and three resistors making up three circuits.  The LEDs are scattered about the board so that loss of one group still leaves a viable light pattern.

I bench tested the load for a 24 hour "burn-in" on a ½ watt resistor.   No noticable heat.  The measured load was 60 mA in reduced voltage mode for tail light use.  I seriously recommend you use a 1 or 2 watt.  The 1/2 watt was to test for "worst case". 

To the left, take note that the brown wire feeds the circuit through the resistor.  The photo shows a 220 ohm 1/2 watt resistor.  You'll need to expand the photo to see the connections clearly.  Each connection is covered with a clear vinyl tube.  The tail light is dim, about the same as the larger lights above.  The yellow wire bypasses the resistor to apply full voltage to the circuit for brake/turn.

This is what you see in the photo above.  Insulate the resistor and splices well, these are active 12V points.  They get shoved around a bit when reassembling the tail light.  I use the clear silicone goop to restrain everything.
 






Man Lift

Yet another "unconventional" design...


The machine had two micro-controllers and a '186 to operate 3 cylinders and a hydraulic motor.  Surely, you jest..... I am a one man operation.  There is no provision in this design for diagnostics.  If it quits, it goes to the shop.  Once you get it retracted and moveable..... The usual practice is to replace parts until it starts working again.  And in my experience, I prefer magnetic controls for a simple application like this.  Have a look; there is also a third (smaller) board from the basket pendant. 

So, OK, here's what happened... I purchased this machine for use when I had overhead cables to deal with.  It gets worked a few times then; I'm on a job and it craps out.  A stuck limit switch, no big deal, in the steel mills..... There, a machine is designed to be repaired in place (RIP).  Troubleshooting is a snap, just look for the missing input when the machine cycles. 

In the field on a job site, there isn't time for such shenanigans.  I lose the job... Or call the service center; "Bring it in, we'll have it a week or so."  Bull, I bought this thing to work, not to sit in a shop.  So, I rent another one just like mine, got the job done, then spent days trying to deal with computer communication.  For what?!?; three cylinders and a motor.   Won't happen again!  All I need to troubleshoot this is a common mechanic's 24 volt test light and a screw driver. 
If you're interested, these boards were functional when they were removed.  I just don't like equipment I have to pay someone else to work on.  I'll listen to any reasonable offer on the computer boards. 




This is the circuit I devised.  If you don't read ladder elementaries, it looks like so much gibberish.  I took what documentation came with the machine and duplicated all the functions, including safeties.  Then came up with a couple of features I thought should be included.  In the interest of simplicity, each function has its' own diode.  Probably twice as many as I needed, but no wires wandering all over the page this way.  Diodes are cheap, on-site time isn't. 








The OEM panels are plastic overlays on an aluminum box.  Membrane keypads.  I cut the center out of the panel and laminated a graphic between two layers of poly-carbonate. It's sealed at the edges with my usual silicone goop, as each layer was installed.  Pilot holes were drilled at each switch location.  The buttons aren't rain tight, so I finagled a drip cover out of the clear plastic.  Hinged and spring loaded against a blowing rain, with a thin rubber seal at the edges.
 
This is for the main (ground) control panel.  There is a smaller one in the basket, on a pendant.  Wiring from these button panels has to go through a diode matrix.  The matrix is fabricated of poly-carbonate (I sure like that stuff) with machine screws to hold the diodes and act as stud terminals for the wires.  The button matrix panel is hinged and folds out of the box to give access to the relays in the back.  I made the outer cover hinged as well, had plenty of the  "piano" hinge stock left over from the matrix panel and rain covers.

Relays are mounted on brackets in the back of the box with the snubber terminals down the middle. The relays look just like what's under the hood of your car, but with 24 volt coils and transfer (form C) contacts.

Things got stalled for a while;  the tilt "switch" turned out to be a true analog two axis inclinometer.  And the pump motor speed control didn't.  Truthfully, as a "Golf Cart" controller, it is a better design for the application.  But, I was expecting it to work like an industrial D-C motor control.  Yeah, well, that chewed up some time too.  Took longer to find documentation on the drive than it did to design around the issue.  Why was it there?  Speed control for the hydraulics is by varying the pump speed.  Simple, just the way I like things.

Speed control for lowering never was adjustable.  But the valve is "proportional".  I'm studying on a control for adjusting that as well.  The reference voltages are available ahead of the "run" contacts, so I should be able to use that to drive a proportional amp.  Maybe, possibly, I think....  A good project for cold weather when the machine isn't working so much.

In the original design, speed reference was based on a switch bypassing some diodes to get the voltage drop.  Since I had to add the circuit boards for the inclinometer, putting high tech stuff in the controller anyway, I added  LM-317  variable voltage regulators for infinite speed control.  Much smoother.... The adjustment for rotate speed is closed up inside the controller box, not accessable to an operator.  Rotation under load with the boom fully extended is sorta spooky.  I trust my own work, but it looks a lot further down from the basket than it does looking up from the ground.

The inclinometer is a work of art.  I had to build a dual comparator to measure the output (0-5 volts) with true plumb centered at 2.500 volts.  And how did I calibrate it?!?  A machinist's "Master Level", accurate to 0.0005" in 10 inches.  One tic on the bubble is 0.001".  With a few clamps and a known 90 degree "angle plate", and a few other basic measuring instruments.  Yeah, I pretend to be a machinist too, sometimes.  Just enough knowledge and skill to make a mess with chips in the floor.

Once I had the machine leveled up properly with the outriggers, the rest was a drop in.  Getting all the twist out of the frame is the biggest issue.   I made the limits a little tighter than the OEM specs.  The original control logic design was a little iffy; I figured the tilt limits might be the same way.  Hey, it's my sorry carcass up in the basket and I'm allergic to pain.

The comparator measures any excursion from center and if it goes too far, drops out a relay.  One for each direction.  Then another iteration for the "Y" axis(fore and aft).  As it happened, I had a box of salvaged circuit boards that had been forgotten.  Stumbled across them out in the barn, looking for something else.  Whatever they were for had just such a comparator circuit.  So, I removed the excess circuitry, added a regulator, and fitted up a cable connector.  Two in service, and two more for backups.  Almost looks as though it was designed for the application.  Only one jumper on the whole board.

Why the computers for something this simple?  Production costs, probably. A "hand wired" magnetic controller such as I built would cost fifty times more than a mass produced plug in computer.  And the copper cable down the boom has only four conductors rather than the sixteen my design requires.  The membrane keypad costs about as much as one of the many weather tight pushbuttons.  All in all, trying to keep the costs down and still have a viable product.   A better approach I would say, than making the boom of lighter guage steel.  Much better.... 

There were operational issues, as well.  As the machine is cycled, the outriggers settle into the ground, throwing the machine out of plumb.  All that's needed is to re-level it and go back to work.  In a "rental" situation, the operator is not all that comfortable with the machine and doesn't know the finer points of the physics involved.  When the machine goes into alarm, it is taken as a failure and the machine is returned as "faulty"; when all it probably needed was re-truing and a power on reset before going back to work.

That may be how I got the machine at the price I paid.  When it is repeatedly returned as "faulty" and the mechanic spends days troubleshooting a non-existant failure, the machine is classified a "lemon" and disposed of.  I don't really know, but it sounds like a good theory. The machine works every time for me... 


My address is:

bhudson@hudsontelcom.com

Bill Hudson
P O Box 101832
Irondale, Alabama, 35210






Web Hosting Companies