Incompetent Wiring unwound

The weather is nice today…. so I decided to do the unthinkable and start tracing wires under the dash.  Ugh, uber-cluster-fuck.  Several iterations of mis-wiring are present.

My front turn signals did not work when I started.  Oh so many fun things found.

  1. 12 v flasher in a 24v system…. yes Dorothy, voltage mismatch causes bad juju and a fast flash.  Cured via eBay for $8 from a local company.
  2. The bulb was missing from the driver side.  That was easy.
  3. Someone has taken the signal/wiper switch off the steering column before, screws are stripped.  Off to Bolt Depot for replacements…..  Ordered several options, I’m sure something will fit
  4. Still no idea why there is no power getting to the passenger side turn signal.  Opened the dash a bit and discovered a wonderful AP4600 paging car alarm.  Deposited that in the trash.  The first of many items that will be coming out from under the dash as I untangle the wiring.

I’m debating a few things:

  • Convert to 12v turn signal system.  It’s all fed from one place, so this is feasible and would make it easier to get LED lights.
  • Replace front turn signals and markers.  The existing ones are Hella and hella hard to find…. lol.

Unobtanium production – making plastic sockets

2016-01-31 14.17.122016-01-31 18.30.56

I completed testing the prototypes yesterday and worked the kinks out of production.  I ran 6 units in about 4 hours this morning and now I’m running 20 units over about 12 hours.  Material cost is around $5 which makes it irrelevant considering that I don’t know who made the ball joints or where to get the plastic part that has failed.  This should give me enough for now.

I do have 12 larger 19mm ball joints that have the same design.  Regrettably I’ll need to fabricate  the same part for them as well.  Even when I reflect on the idea of spending a week making a part it’s still about as fast as ordering from Europe and I know exactly where to get the next replacement parts.



In the spirit of Open Source Hardware, here is the STL file to print the 16mm ball sockets.  A couple of caveats.  I redesigned the joint to have a ring that holds the ball captive.  While the ABS can handle the insertion, it tears the socket apart when it is removed.  The sockets cost me 15 cents to print, so I shrugged my shoulders.  These should almost never be taken apart.

16mm-balljoint-r35 – for ABS

16mm-balljoint-r14-FINAL – for PLA

Yes, there are two different files.  The PLA one is optimized for that plastic, the ABS one for that plastic.  ABS shrinks when it’s printed which gummed up my entire design.  PLA actually looked better and seemed to be fine with insertion/removal.  It has a lower plasticity temperature and will go soft at 140 degrees Fahrenheit.  I was concerned that the compartment that houses the generator could see that temperature.  Sigh, PLA would probably have been fine, but ABS is a tougher plastic.

3D replacement parts – tuning the unobtanium

2016-01-30 14.50.57

I’m so close…. I have 3D printed ABS Plastic sockets that fit the ball like a glove and fit the tie rod end so tightly I have to use pliers to push them out.  As you can see from the lower picture it has taken me many iterations to get here.

2016-01-30 14.50.43

I’m working on tuning the retaining ring right now.  The original part was a semi-flexible rubber that expanded when you inserted the ball and then the hoop locked it in place.  ABS is a great fit for this as it is slightly flexible.  However, I’ve chosen to just print a ring right above the bulge of the ball stud.  This captures the ball stud effectively and achieves the same effect.  The socket will expand slightly when the stud is inserted and the tie rod end will “lock” the stud in place preventing easy withdrawal.  A photo of the CAD design is below.


2016-01-30 14.50.48


3d printing for unobtanium parts

2016-01-28 19.55.05 2016-01-28 18.23.02

A couple of weeks ago I decided to buy a 3d Printer Kit from Prusa Research and put it together.  You can read more about that on where I focus on electronics.  One of the great uses for a 3d printer is printing hard to find parts.  These plastic/rubber sockets are a perfect application.

I’m using a $600 printer, $18/lb plastic and free software to model them.  I’m currently printing version 19 trying to get the dimensions just right.  In the photos above I show version 13 in PLA a starch based biodegradeable plastic.  I want the final part to be in ABS, and didn’t realize it shrinks a few percent when printed.  🙁  So now I’m working on adjusting that out.  It takes about 45 minutes and costs around 15 cents to print one.

I have more than a few parts that are hard to get or out of production on this bus.

Here are some photos of the lead up to the photo above:
2016-01-29 10.19.19

An ABS version being printed

2016-01-29 10.19.12

The assembled printer


A drawing from Autodesk Fusion 360, which is free for hobbyist and startup use.


A rendering from the same program

2016-01-28 18.39.18

versions 1 through 12…. they don’t always turn out so great…..unobtanium-gcode

A rendering from Simplify 3d which is the software that transforms the drawing into “g-code” that the printer understands.unobtanium-v10

Another rendering.  There is a torus or donut shaped ring visible.  I moved it further down to retain the ball.

2016-01-26 20.36.49

Another picture of the 3d printer, fully assembled.

Tie-Rod Mystery

So, my bus has these nifty self-levitating luggage doors.  Well, they would be nifty if they didn’t have some worn out hardware.  These ball joints are on the doors.  I have 24 of these and 12 larger cousins installed.  These are 16mm balls, which I can get sockets for $4/ea.

Sad thing is that the nylon “cup” is all that is bad.  It’s probably not nylon, it’s a semi-flexible translucent rubber.

I have half a mind to try and print the damn things on my 3-d printer when I finally get it later this week.  Hmm, let’s see….. assuming that all the sockets are the same price that is $144 worth of plastic bits….. probably worthwhile.  They are relatively simple as plastic bits go, so it shouldn’t take too much trial and error to figure out the exact size.

I may just do that.

In the meanwhile, does anyone know what brand these are?2016-01-19 19.10.11 2016-01-19 19.09.50 2016-01-19 19.09.34 2016-01-19 19.09.29 2016-01-19 19.09.26

A replacement candidate arrived yesterday in the mail….

2016-01-19 18.49.48