OtO Power Connector Re-Design

The Problem

I identified two problems with the current connector after analyzing all of the returned units.

1. Fragility. The connector housing is held together by a small plastic piece and it is too weak. It can easily break from the tension of pulling the connector apart. The size of the power pins is also way too small and can be easily bent or snapped.
2. Corrosion. Since this connector is intended for outdoor use near a sprinkler system it will inevitably get wet. This connector does not seal effectively and due to the current flowing through the connector, it corrodes and breaks quickly.

Testing

After discovering the problem, I researched more robust connectors with better seal designs. I ordered many samples of various designs and would use testing to validate a new connector. If the tests did not yield results that provided confidence in any of the connectors, then designing our own would be the next course of action. To solve problem #1, cycle, force, and durability testing was done by hand. These tests quickly narrowed the field to ones that were far more robust. The more complex aspect of testing was looking at ingress protection. A submersion test was developed that places a connector underwater and passes 0.2-0.6A at 12VDC over an extended period of time. A power supply and resistors were used to dissipate the power as heat. The connectors were periodically opened to look for ingress but also the current was used as a failure indicator. If a connector corrodes, the contact resistance increases and so at a fixed voltage, the current should drop. Through this data, connectors could be compared head-to-head.

Connector Selection

There were two connectors that passed the robustness tests and also showed no signs of ingress or current loss after a few weeks. The next step was to evaluate the two from a user experience standpoint. One connector seemed much more user-friendly than the other and was ultimately the option that got pursued. This connector is a standard 2.1mmx5.5mm barrel jack that is commonly found on laptop chargers. The design can be bent, twisted, pulled, etc... without being damaged. On top of that, it can be inserted in any orientation. To seal, this connector has an o-ring that acts as a face seal when compressed using an external thread.

Design Limitations

When you implement a particular design releasing it to thousands of consumers, it is important to know the conditions in which it performs best, and when it can fail. At this point, I knew the connector would seal when submerged a few centimeters. I also wanted to know if the sealing capabilities would be affected by temperature as these units run in cold and hot weather. So I did a similar submersion test but would cycle hot and cold water every 30 minutes into the chamber. I built this test using solenoid valves connected to hot and cold water lines, a Raspberry Pi, a relay, and a Python script. We found that under extremne temperature fluctuations such as a 30min cycle time, the O-ring fails to seal possibly due to compression set. However, this test is far more rigorous than the field and is still good information to know about your product.

Product Attachment

The connector next had to be attached to the product in a way that was easy to assemble using the current production processes. Changing the connector itself was easy and just a matter of soldering a new component. The interesting design came from what happens when the connector is not being used for example when the device is running on solar. The jack side of the connector needs to seal so that it isn't wet when it gets connected. It also needs to be securely attached to the product. To resolve these problems, I used SolidWorks to design a rubber plug (seen in orange) that fills the jack side of the connector that can be easily removed by the user. I also worked on a snapping mechanism that holds the connector in place on the bottom of the device. I iterated using 3D printing and eventually finalized a design.

I also got to conduct FEA studies to ensure the injection moulded snap on the base plate would not yield before it deflects the amount that we require. This was a great way to validate the deisgn before sending it out to be 3D Printed. My study including applying the known deflection required for the connector to snap in, and ensuring the stress levels were below yielding.