Learning to Weld

Over the holidays, I decided to do something I’ve been thinking about for years: I got a welder and I’m learning how to use it. I did a lot of research, and decided that MIG welding was the best choice for my needs, which is general household use and automotive chassis/suspension work. However, I also know that I’m not going to weld more than a few times per year, and MIG welding has the added complication of leasing a gas bottle, knowing what to fill it with, paying a welding supply company to fill it, transporting it, and storing it in my already-crowded garage. With that in mind, I decided to go with a similar welding process, self shielded flux core, also known as FCAW-S welding.

I read a lot online, and there are a lot of people who say that flux core is a waste of time, it produces bad welds, etc. A little more research leads me to believe these people are either bad welders, are using bad tools, or won’t accept that a good weld doesn’t have to look like a perfect “stack of dimes” TIG weld to be structurally sound and get the job done. I think a lot of the frustration comes from people using inexpensive AC output welders like you get from Harbor Freight, instead of a slightly better DC output welders, like the Hobart Handler 100.

The Hobart Handler 100 is the machine I bought. I think a lot of people will say I should have stepped up to the 130 or 140 to get the full MIG capability, but that’s something that I truly don’t need now, or in the near future. If I find that I want to work on thin sheet metal (auto body) and do a lot more welding than I anticipated, I will probably step up to a much nicer machine, and keep the FCAW rig as an outdoor welder, if the need arises.

With decent equipment (including a good helmet), and good surface prep, I was able to produce sort-of-ok welds within the first hour. This is a T-joint with 0.083″ square tubing, which simulates what I’ll be working with when I install subframe connectors on my Mustang.

Not bad for a first try?

Trying an overhead T-joint while laying on the ground, as if to simulate actually working on a car, is another matter. I struggled to make a garbage weld, and if I did this to my car, I would have to grind it off and try again. Oddly enough, an overhead butt joint was relatively easy to do. A lot of success depends on gun angle, wire stick-out, travel speed, and how well you can see, which highlights the need to get a good helmet and good gloves that work for you.


After some more practice, I want to cut through some of these joints and see what kind of penetration I’m getting. I hope to be ready to use this welder on my car with confidence by the end of 2019.

Happy new year.

EPAS Almost Done

After coming to grips with the idea that my new electric assist power steering setup in my 1986 Mustang SVO would lose the collapse functionality of the stock steering column, I created my own collapsible steering shaft in front of the column to make up for it. This isn’t as difficult as you might think. You just need the right combination of off-the-shelf tubing and universal joints, but when you tell people you’re making a custom steering shaft for your car, they tend to look at you like you’re crazy.

Custom steering shaft seen from under the car

I’ve ordered a rubber boot that will hopefully fit nicely between the u-joints. The shaft has grease on the sliding mechanism, so I want to protect it from dust and debris. Under normal circumstances, the shaft shouldn’t slide at all, but I wouldn’t be surprised if chassis flex caused it to slide a bit.

I need to finish and clean up the wiring, put thread locker on all the set screws, and figure out how to tighten a bracket that is nearly impossible to reach. After that, I should (finally!) be on the road again, just in time for winter storage.

I’m going to drive it a while as-is before I add speed sensitive assist. I’ve been having some trouble with an analog input getting interference from the vehicle speed sensor that I might touch on in a future post.


Electric Power Steering Progress Pics

As is tradition, I have an automotive project going much slower than expected. I thought I’d share my progress anyway. This is the installation of electric assist power steering in my 1986 Ford Mustang SVO.

Here is the steering column coming out after removing all the trim pieces and unplugging all the electrical connectors. Unplugging 30+ year old connectors without breaking the tabs off is no easy feat, but I managed to do it. It’s difficult, frustrating work at times.

Pulling the steering column

With everything laid out on the bench, it’s time to go past the point of no return and cut the original steering column. The outer part of the column is just a hollow steel tube. There’s nothing too special about it, but when you cut enough of it to allow room for the electric motor, you also have to cut off a fairly beefy support bracket that helps locate the steering column. This reduction in support is addressed by a new firewall support bearing and a bracket that holds the electric motor at the correct angle. It’s acceptable, but maybe not as stout as the factory setup.

Ready to cut the steering column

Inside the hollow part of the column, there is a standard 3/4″ DD-shaped steering shaft. It has to be shortened, and then notched. The DD shaft inserts into a coupler that is pre-welded to the shaft of the electric motor. Two set screws go through the coupler and fit into the notched section of the steering shaft. The set screws have to screw down almost flush with the coupler, because the outer steering column tube slides over this whole assembly.

I had some trouble fitting everything together, and I also managed to strip the set screw threads on the coupler. I ended up tapping the hole to a larger size, and everything worked out. In general, when you strip the threads in a hole, if it was a metric size to begin with, you can go to an SAE size thread that is just a bit larger, and if it was an SAE hole to begin with, switch to a metric size thread that is a bit larger. This lets you tap out a larger hole without increasing the size too much.

Column reassembled with the motor installed

Time to test fit the assembly in the car. I was worried the motor might be too close to the brake or clutch pedal, but it’s not. Your foot has no reason to be that high on the pedal assembly, and the assembly itself doesn’t interfere with the motor either. This is going to come together just fine.

Installing the column in the car

The biggest problem I have with this kit is that this assembly takes the place of what used to be the collapsible part of the steering column. I am going to address this by building a custom collapsible steering shaft on the engine side of the firewall. This is pretty easy to do with 1″ DD tubing that slides over the standard 3/4″ DD shafts. The collapsible steering shaft in combination with extra couplers and joints between the steering rack and steering wheel makes me reasonably satisfied that a hard crash won’t send the steering wheel through the driver’s seat. This is something I’m going to put more thought into.

I’m waiting for more steering shaft parts to arrive, and I also have all the electrical work to do, and I will post again when that is all done.

Using CAN bus on an older car

There’s a lot of information online about OBD-II scanners and “car hacking” in general, but what if you have a car built before 1996 that doesn’t use OBD-II? I’ve found the underlying technology of OBD-II, CAN bus, can be a useful addition to an older car. I’m going to detail a few things I’m working on that might give you some ideas for your own car.

The car I’m working with is a 1986 Ford Mustang SVO. This car is a great starting point because it is has computer controlled fuel injection and ignition timing, but not much else in the way of computerized controls. The factory engine control unit (ECU) leaves a lot to be desired in terms of how airflow into the engine is measured. With that said, the SVO becomes a great candidate for MegaSquirt, an aftermarket DIY-oriented “stand alone” engine management computer. My SVO currently uses a MegaSquirt 2 that I built myself, and I have a MegaSquirt 3 on my test bench that will eventually go into the car. The MegaSquirt ECU will become the heart of the CAN bus network in the car.

Newer versions of the MegaSquirt firmware have CAN bus functionality built in. The easiest and most common way to use the CAN bus is to use the MegaSquirt tuning software to enable CAN realtime data broadcasting. The data format is described in this PDF. Digital dashboard systems from major brands such as AEM and Autometer can utilize MegaSquirt’s CAN bus data format to replace the factory gauges in your car.

This is about as far as most people will want to go, but I wanted to do some custom CAN bus programming. Here is the plan for my car.


Meguinauge is my Arduino project (source code on GitHub) that uses CAN bus to display engine parameters (i.e. gauges) from MegaSquirt on a 20×4 character LCD. I deliberately chose to use a character display to preserve a “period correct” look for the car. There are buttons to select how many and which gauges you want to see, and a warning light when an engine parameter goes out of range.

My engine simulator is running lean under boost, triggering the warning light on the breadboard


Carfuino is my Arduino project (source code on GitHub) that uses CAN bus to retrieve vehicle speed and a custom trip odometer reading as the basis for an automotive performance computer. Performance computers that measured things like 0-60 time became popular aftermarket accessories in the 1980s and are built in to many performance cars today. Carfuino will display data on a 16×2 character LCD.

My car did not come with a vehicle speed sensor (VSS), so getting this value is a custom job for me. I thought about different ways to handle the VSS signal, including using an Arduino analog output to send a variable voltage that maps to speed. For example, a 0.5v signal could represent a dead stop, and 4.5v could be 200mph. But when you consider the amount of electrical interference and grounding problems in a typical automotive environment, and the possibility of using the VSS signal for multiple subsystems, CAN bus made the most sense.

EPAS Middleware

Speaking of VSS signals, I am installing an electric power steering motor and ECU into my car to replace the hydraulic power steering, and I am going to modify it to create speed sensitive power steering. Most aftermarket electric power steering kits use a knob to adjust how much power assist to deliver, and I am going to replace the knob with a digital potentiometer that moves based on vehicle speed. This is yet another Arduino project, and it is really two projects in one. There is the aforementioned feature of changing a potentiometer value based on speed, and it will also be the controller that accepts and raw input from the VSS. It will take the signal, convert it to a more usable format, and put it on the CAN bus. This way, the VSS signal can be used by the Carfuino project, and MegaSquirt can use its “CAN receiving” setting to take the value from the CAN bus and use it internally.

OBD-II Emulation

I currently have no plans to do this, but it seems like it would be a fairly straightforward task to use the CAN data from MegaSquirt, emulate the OBD-II format, and send the signal to a standard OBD-II diagnostic port. This way you could plug in any off-the-shelf OBD-II scan tool or gauge and it would work. I don’t really need this because I find the raw MegaSquirt logs to be more valuable, but it’s an interesting idea.

I hope this has given you a few ideas about how you can use CAN bus to bring your older car into the modern era. It is important to check local laws and regulations regarding vehicle modifications. Do not use an Arduino in a way that can cause you to lose control of your vehicle.

Mustang 2018 Update 2

I got my ’86 Mustang SVO all put together, but it’s already time to take it apart again. Here’s what’s going on lately.


I don’t even want to know how many hours it took to trim and fit the silicone hoses required to fit my new air box, but it’s done! If you know about SVOs, you can immediately tell it’s not a stock airbox, but to the casual observer, it looks like it could be stock, and that was the idea. I’m happy to have less noise from the engine compartment, a fully functional charcoal canister, and it still flows plenty of air. My intake temps will likely go down a bit, as well.

Never mind the power steering pump bracket, with no power steering pump.

The quieter intake really underscores just how loud the blowoff valve is. I recently ditched my Forge blowoff valve in favor of an APR bypass valve set up in a blowoff configuration. That quieted things down a bit, but I wanted more. I decided to use a valve cover breather on the “exhaust” side of the valve to act as a muffler. It doesn’t make it quiet, but it helps.

If it looks stupid, but it works, it’s not stupid.

Short throw shifter

The MGW shifter feels great. The longest shift handle they sell (“comfort”) is borderline too short for this car. I guess people are looking for the shortest throw possible, ergonomics be damned. The new shifter is transmitting more noise from the transmission than I’m used to. I might add a layer of Dynamat under the boot to help with this. Overall, a good purchase.

Steering rack

With the Flaming River 15:1 quick ratio manual steering rack coupled with the aluminum steering rack bushings, I have the best steering feel and feedback I’ve ever experienced in a fox body Mustang. Turning the wheel at a dead stop is a real chore, though, and it’s difficult to make quick transitions at low speeds. I am going to move forward with an electric assist power steering project. More on that below.

What’s Next

  • I am building an Arudino-based performance computer to fit in the lower console, replacing the factory clock. It will measure things like g-forces, 0-60 time, etc. It might use the same microcontroller and codebase as my Meguinauge project.
  • The electronic assist power steering kit is on its way.
  • I acquired parts to add a speed sensor output from the transmission. I will use the speed data for the performance computer mentioned above, and also a “middleware” controller to add speed sensitive assist to the power steering ECU. It will essentially take the VSS input from the transmission and spit out an output through a digital potentiometer. I’m still working out the details.
  • My budget is tapped out at the moment, but at some point I want to start acquiring 8.8″ rear end parts to swap the rear before doing some suspension upgrades.

Mustang 2018 Update 1

I’m finally getting to the point where I can get a lot of work done on my 1986 Ford Mustang SVO. I have a few long term projects in the planning stages, but I’m finishing up some smaller projects to be done this summer.

First, I am replacing the failing power steering pump and rack with a quick ratio manual steering rack, along with upgraded aluminum rack bushings. Mustangs had the option for manual steering as recently as 1984, and my 1986 SVO is just about the same weight as the factory manual steering cars. I think that steering effort will be just fine with the manual rack, but I have another upgrade planned for 2019 – electric assist power steering! There are kits available to add electric assist to fox body Mustangs, and some people also use parts from Saturns to fabricate their own electric assist units. I will probably be getting the kit, but with a twist: I’m going to use an Arduino (or similar board) to make the power steering speed-sensitive. It’s going to take some work just to get a speed sensor in my car and get it talking to the main ECU. I’ll have more updates on this when the project is in full swing.

Another project that I’m finishing up is the intake and airbox. When SVO owners remove the VAM (vane air meter), they typically put a conical air filter on the end of the intake tube. That’s what I had done, but there’s a problem with that: the charcoal canister vent line goes to the airbox, and without the airbox, that vent line has nowhere to terminate. March Performance makes a “Ram Air” kit for fox body Mustangs, and you can buy the airbox as a standalone item. I bought one of these and added a fitting that would accept the charcoal canister vent line. I also fabricated a bracket out of sheet aluminum to physically mount the charcoal canister in a new location. From there, I am finishing up a custom silicone intake tube that goes from the airbox to the turbo inlet. This is more complicated than you’d think because the airbox outlet is several inches higher than the turbo inlet, and space to maneuver is tight.

Custom bracketry for the charcoal canister

These projects are turning out to be more complicated than anticipated, so I did something else I’ve been meaning to do that is much more straightforward. I removed the factory Hurst shifter and replaced it with a MGW shifter. This is said to be the best shifter on the market for the T-5 transmission, and the build quality is excellent. I used their “comfort” handle, which is the longest one they offer. No regrets on that decision, because it is still shorter than a factory shifter and remains in easy reach while driving.

Factory Hurst shifter

MGW shifter with MGW gripper ball

Other than that, I’ve been working on minor maintenance and restoration tasks. I’m not done with this year’s projects yet, but I’m already planning my next steps. I’m going to upgrade my MegaSquirt ECU and finalize my Meguinauge project, and move on to another Arduino project for this car. It involves CAN-BUS communication and a 3-axis accelerometer. Stay tuned.

Motorcycle 2018 Update

Here are some pictures of my motorcycle project I completed in Spring 2018. This is a 2007 Kawasaki EX250F7F (a.k.a. Ninja 250R). It was a great bike before, but I was never fond of how it looked, and it needed some maintenance work and upgrades. I am happy with how it turned out.

The cafe racer look comes from the kit I bought from Blue Collar Bobbers. I upgraded the suspension with Sonic Springs and new fork oil in the front, and a stiffer 2008-2012 Ninja 250R spring/shock in the rear. The brakes were overhauled with improved pads, new fluid, and stainless steel lines. Adjustable hand levers from the Kawasaki EX500 complete the package.

What I started with. Tribal flames aren’t really my jam.

The second picture is the finished product, as is tradition.

Sawed off the back of the frame. No going back now!

Cleaning the carbs. I used carb cleaner spray, a single copper strand from a lamp wire, and a Pine Sol soak.

Adjusting the valve clearances. I’m glad I won’t have to do this again for a while.

Frame reassembly and wiring are coming together.

Rebuilding the front suspension. Bless this mess.