LocostUSA.com

Great progress!

Finished installing fuel lines and brake lines. Also welded up a side exhaust. This was all routine work, but I thought I’d mention how helpful it has been to have a 3d printer. First, instead of modeling the exhaust with wire or swimming pool hose, as some have done, I printed a plate that mimicked the exhaust flange on the engine and then added round spacers. I glued and taped these to a stiff wire so that I could more accurately measure the path and clearance of the exhaust.

Second, I built a wire-straightening tool for the brake and fuel tubing using rollers for a sliding screen door. This badly gouged the soft nicopp lines ,so I salvaged the bearings from the screen door rollers and added 3d printed outer rings that exactly matched the fuel line diameter. The soft plastic rollers did a good job straightening the lines and did not scratch them up.

Lastly, I printed some custom hold down brackets to hold the lines in place. I can print whatever size and spacing I need to hold the lines. See the photos below. The filament was petg at 100 fill.

[img]composite.png[/img]

I'm impressed! Would you consider doing additional 3D printing for a fee? I am trying to figure out how to make a filler for my taillights to the rear fender curve.

_________________
Thom

In order to have someone else make your print you have to design it in CAD. A friend of mine uses TinkerCAD (easy) but I use the free version of Fusion 360. There is also a free program called FreeCAD. It's a good skill for car guys to learn. I took a class at our local makerspace and they also let me use their 3D printers. You can make your parts there but eventually you will get the bug and buy your own. You can get one for a little more than $200.

You can also have things printed online, but the problem is that quite often you make a part and find out it is not quite right. I had to make two or three versions of my hold down clips before I was happy. A test part only costs a few cents so it's no problem. But if you farm out the printing it will cost too much to redesign and reprint parts. Thus its best to buy a printer or use one at a makerspace (they only charge for filament and its cheap).

One problem though. My makerspace only allows printing with PLA filament. That's fine, but PLA can't take heat. So for car parts, if they are going to be exposed to heat, you need to use ABS or PETG.

I hope this helps. Once you buy a printer you'll be using it for lots of things.

As I reworked the wiring harness and figured out my dashboard I found that my fuel level sensor is shot. I could not find a 95 sensor, new or used, and ones I did find for other NAs were over $100 used. Why should an over 25 year old sensor work better than the one I have?

Luckily, or unluckily, for me, the previous owner threw away the instrument panel and associated harness when the car was his project. He intended to use custom gauges, and so did I until I realized that the gauges, sensors and/or adaptors needed to make things work were going to cost a lot of money — over 7$00 for high quality. So I decided to build my own electronic instrument panel using a Raspberry Pi, 7 inch screen, and data acquisition board (I already had a Raspberry Pi waiting for a mission). The other parts cost about $100. This way I can use the present sensors and write code to evaluate them correctly.

This also means that I can build my own fuel level sensor and not worry about matching the resistance range of the original. Anything will work for the Pi so long as I tell it what the values mean. So I bent up some aluminum, added a potentiometer (1 dollar) and reused the existing float and float arm. I had to 3D print an adapter to attach the pot to the float arm. The geometry of the new one is slightly off by a few millimeters but I could correct that and make it exact if I wished. It should be close enough. Empty is 1000 ohms and full is 3000 ohms. If I had my old instrument panel I could have matched the original range (110 ohm is empty and 3 is full) by selecting a different pot and adding some resistors in parallel and series. I don’t see why this should not work.
[img][attachment=0]fuel%20sensor%20V2.png[/img]

I'd be sure to check that the pot doesn't dissolve in fuel vapor.

Good idea. I'll seal one in a jar with gas for a few days and see how it does. I know the 3D print will be OK because it is printed with PETG filament -- known to be resistant to gasoline. If it fails I have an ultrasonic sensor that will work but I'll have to put a small hole in the tank.

The wiring harness is back in and the car starts and runs pretty well. At first it had very rough idle and I spend a lot of time diagnosing it. But suddenly it cured itself, something sticking must have gotten unstuck. No trouble codes thrown by ECU. Today it ran round the block with no problems.

So now for the body. I decided to use a skin-on-frame approach, very similar to the way kayaks are made. Some folks are 3-D printing the frames for kayaks as well. For car examples, see the BMW GINA, https://www.youtube.com/watch?v=Blqn0Om ... CarsGarage and the EGAD Light Cocoon, https://www.youtube.com/watch?v=Tmw8dxP ... =EDAGGroup.

I don’t have a design yet, but I made a test panel to experiment with (see photo). It took only four prints on a Creality CR10S 3D printer to make this panel, that is, it’s made of four prints that are then plastic welded together. Each quarter took 12 hours to print. As you can see from the photo, the test print is almost big enough to be the hood (but I will not try to reuse it). Next, I will cover this panel with polyester cloth, heat shrink the cloth, and spray it with several coats of polyurethane. If I need more strength, I can add a layer of fiberglass on the back side.

The final body should be much lighter then if it were made with solid panels that were 3D printed and covered with fiberglass. It will also be cheaper to make and could be illuminated from the inside for some cool effects (see the Light Cocoon video above).
[img]test%20panel.JPG[/img]

That's a pretty neat idea.
Reminds me a bit of scale RC airplane construction, stretching vinyl over balsa.

IIRC, Binky (Bad Obsession Motorsport YT series) has 3D printed fender flares, though they're 1-piece & made on a massive printer in France.

_________________
1975 MG Midget
http://www.locostusa.com/forums/viewtopic.php?f=36&t=24382
13B Powered Rotus Geauxkart
http://www.locostusa.com/forums/viewtopic.php?f=35&t=20899

Yes, some kayak makers use clear vinyl. If I had a beautiful, TIG welded, round tubing frame, I might consider it. You are right, model and full-size aircraft makers also use this approach. And, of course, early airplanes used it.

Geodesic structure for domes and aircraft.

_________________
Miata UBJ: ES-2074R('70s maz pickup)
Ford IFS viewtopic.php?f=5&t=13225&p=134742
Simple Spring select viewtopic.php?f=5&t=11815
LxWxHt
360LA 442E: 134.5x46x15
Lotus7:115x39x7.25
Tiger Avon:114x40x13.3-12.6
Champion/Book:114x42x11
Gibbs/Haynes:122x42x14
VoDou:113x44x14
McSorley 442:122x46x14
Collins 241:127x46x12

Thanks, I'll look that up as I do the design.

I only post when I have hit a milestone or when I think I have done sometime unique that others might find helpful. This post is the latter.

I am building the cover for the tunnel and my problem was that I needed to locate and drill five holes for the e-brake system. The holes have to match five M8 nuts that I have welded into the tunnel. But the drive shaft prevented me from reaching all of them from behind in order to mark the location on the aluminum cover.

I know that carpenters use dowel centers, little metal pieces the diameter of a down, to mark where a matching dowel hole should go. Metal workers sometimes sharpen threaded rods or bolts to serve a similar function in order to mark metal.

I did not want to sacrifice five expensive M8 bolts (and spend the time to sharpen them) so I 3D printed what looks like dowel centers to fit the M8 nuts. I put several layers of tape on the aluminum so the dowel centers could make an impression in them. I put the cover on and then tapped it with a rubber mallet. It worked. The holes were located perfectly.

The original owner of the Kart discarded the Miata dashboard and associated wiring harness, as he planned to add custom gauges. I looked into this and it looked like it would cost at least $700 to add a decent set of basic sensors and gauges. Its easy to add an electronic dash if you have OBDII but this is a 95 with OBDI. So I decided to use a Raspberry Pi (already owned), with a 7 inch display ($80), and a Pi-Plates data acquisition board ($34). I also added a $30 GPS module for the speedometer and an Arduino (already owned) for RPM (I will explain below). So the set-up cost me about $150 rather than $700. With this I was able to directly read the signals and sensor values from the Miata without adding any new sensors. I programmed the Pi in Python using the PyGame library to handle graphics. So far as I know I am the only person who has done this. Here are my takeaways
.

• It works great but it takes 38 seconds to boot up each time I start the car. But I guess I don’t need that information for the first 38 seconds
• The Raspberry Pi is really a computer, not a real-time processor. So, while I could read RPM using interrupts, it was not accurate as you can’t be sure that all processes will be interrupted immediately. So I gave up on that and added an Arduino to measure RPM and just send that information to the Pi.
• I had to add a box (the white one behind the dashboard) to condition the noisy 12 - 15 volt signals so they would not blow up the Arduino and Pi. This is mostly a bunch of resistors (voltage dividers) to lower the voltages and Schottky diodes to catch spikes.
• If I knew what I know now, I would have just used a couple of Arduino processors with small OLED displays to create the dashboard. Instead of one 7 inch display the dashboard would have to be made up of several smaller displays. But there would be no delay when starting the car and the set-up would probably be cheaper. Here is an example of using Arduino to make a gauge, but there is plenty of other sources out there. https://www.youtube.com/watch?v=xlb7FGK ... boutDesign

At this point in the project I am ready to build the windshield and body.

April of 2022 I posted that I build a fuel sender using a cheap potentiometer. I can now report that that failed completely. The gasoline fumes destroyed the potentiometer and it stopped working. It seems that fuel senders use special potentiometers that are chemical resistant. I rebuilt the sender again by hacking a $30 Summit Racing sender designed to be added to racing fuel tanks. It works great but I won't bother posting the details because it would be useless to a Miata owner. The Summit unit uses a completely different resistance range and, on top of that, is in the opposite direction of a Miata sender. Thus my hack will not work with a Miata dashboard. Since my dashboard is custom, with custom programming, I can make any sender work by just changing the code.

I'm posting this only to warm people to ignore my hack of 4/22.