Ever wondered how electronic products are manufactured?  In this post, we give you a glimpse into the process we use to manufacture the MRServo slow motion turnout controller.  While the techniques are vastly simplified from that used by high-volume electronics manufacturers, the same basic steps still apply.

Design

MRServo is based on a design originally developed by Nathan many years back.  A few years ago, when we had this crazy idea to start Iowa Scaled Engineering, this became our first product and the process of converting it to a manufacturable product began.  It took several revisions to optimize the design but eventually resulted in the current design.  For those interested, the full schematics, layout, and firmware files are available on the Documentation tab of each product’s page.

Since the MRServo-1 has no relays, it’s PCB is smaller.  This was done for cost reasons, to minimize the cost of the basic MRServo-1 version.  However, the -2 and -3 versions share the same circuitry as the -1, but add the components necessary to support the relays.  For that reason, the -2 and -3 PCBs are larger.  The only difference between the -2 and -3 is in which components are installed.

Below, we take you through the steps necessary to make an MRServo-3.

Manufacturing

We start with the bare PCB, made by a third party using the Gerber files we provide.  For larger volume products like MRServo, we typically work directly with a PCB manufacturer and purchase a relatively large quantity of boards at one time.  For small runs, a service such as OSH Park can be used, where the fixed costs are shared by many individuals, making the per board cost much more affordable.  Many of our prototypes are done that way.

Each PCB comes panelized into a strip of 6 boards.  This makes the process of applying the solder paste easier and helps manage the boards while placing components.

Since we are primarily using surface mount components, we use solder paste instead of the traditional wire solder and soldering iron.  Solder paste is a flowable form of solder that can be applied to the PCBs prior to components being placed.  It is made up of powdered solder held together by flux.

The process of applying the solder paste is very similar to that used for screen printing.  We start with a laser cut kapton stencil with holes cut where we want the solder to be placed – which is on all the pads intended for surface mount components.  While we would love to have our own laser cutter, the stencils are made for us by OHARARP.

The next step is to apply the solder paste.  A decent line of paste is applied to a putty knife.

It is then drawn across the stencil, leaving a layer of paste.  Then, the excess is scraped off and the PCB carefully removed from beneath the stencil.

Below you can see the islands of solder paste left on the PCB.

Once the solder paste has been applied, it is time to place the components.  For storage of the small surface mount components, we have found petri dishes to be convenient containers.  With the lid on, it is easy to transport the components between the storage shelves and the workbench with (minimal) risk of spillage.  The lids can be clearly labeled.  Additionally, picking the parts out with tweezers is quick and easy.

Each component is carefully placed on the board.  Although alignment is important, at this stage getting it perfect is not necessary as the surface tension of the solder, once it melts, will tend to pull the components into proper alignment (assuming they are close enough to start).

To melt the solder and form permanent joints, the boards must be heated to >200C.  This is accomplished with the help of an oven.  The temperature profile is closely monitored to make sure not to overheat the boards.  Once the peak temperature has been reached, the oven is turned off and the boards are allowed to cool naturally, creating the final solder joints.

The boards with all surface mount components soldered:

To finish building the MRServo, the through hole components are then soldered using a soldering iron and wire solder.  A finished MRServo-3 is shown below.

Final Steps

Although the MRServo board has been completely assembled, a few additional steps are required to make it a complete product.  First, the individual boards are depanelized by cutting the small tabs between them.  Next, the firmware is loaded into the onboard PIC microcontroller using the 6-pin header holes located in the middle of the board.  We use a PICkit2 programmer with a custom programming cable.

Finally, the MRServo board is tested for correct operation.  If everything checks out, the product is packed with all of the associated pieces – mounting tape, throw wire, servo kit, and instruction sheet – and is then ready to be shipped to a customer.

Do-It-Yourself

While we primarily intend to build and sell MRServo as a complete product, as an open source design, all the design files are available if you want to try building one yourself.  If you do decide to give it a try, please let us know how it goes.

Conclusion

We hope this post has given you an interesting glimpse into the manufacturing process used for our Iowa Scaled Engineering products.  If you have any questions, please don’t hesitate ask.  Enjoy!