Portfolio

Simplexity Projects

Below is a collection of some of the projects I have had the opportunity to work on while at Simplexity Product Development. The descriptions below are simplified and generalized to remove client sensitive details, and potential intellectual property has been removed.

Molecular Diagnostics R&D Tool Control PCB

A client needed an R&D tool that performed various microfluidic, thermal, and motion operations as a Library Preparation Breadboard. Full details on the product are available on the Simplexity Website. The electronics were controlled through a PCB stack that had a STMicroelectronics Nucleo board as the main microcontroller to allow rapid development. I designed the schematics and layout for the main shield PCB to interface with electronic peripherals. Some notable interfaces on this shield were:

• Various Valve Drivers (inductive load) to employ Spike-and-Hold capabilities on valves
• An H-Bridge based motor driver
• 4-20 mA DAC and ADC for precision vacuum control
• Various optical/digital sensors

Additionally, I designed the system electrical schematics, cable designs, and contributed heavily to the fluidics, pneumatics, and vacuum design and routing.

Thermal Control Subsystem

A molecular diagnostics client required precision temperature control of a flowcell inside a larger R&D tool. The application required rapid temperature changes, minimal temperature overshoot, high temperature precision, and high thermal uniformity across the flowcell, all while maintaining fluidic connections with the flowcell. I designed the TEC system that could achieve these goals including design of the mechanical parts and assembly, electrical system design, coolant system, and control hardware/tuning.

I discuss this project as well as some more technical background on Thermoelectric Coolers (TECs) in a previous blog post I wrote: Understanding Thermoelectric Coolers: Cooling Tech Simplified.

Satisfying the precision requirements took significant analysis and testing. I created a custom peltier element model and designed the thermal system using Modelica to get a preliminary idea of the thermal performance when ramping. Thermal resistance between the TECs and the plates was minimized by lapping the plate surfaces and testing various high performance thermal interface materials.

To speed up testing, I wrote a python script that interfaced directly with the thermal controller and measured temperature using external RTD sensors. This allowed verification testing to be accelerated, more repeatable, and autonomous.

AI Robotic Arm

Aivot develops AI enhanced robotic solutions to intelligently automate labor. The robotic system consists of numerous sensors and electronics paired with robotic arms to learn to how to perform tasks from human demonstrations. A full case study of this project is available on the Simplexity Website, and I was responsible for designing the vertical lift mechanism, the electronics box wiring and cables, and performed the initial motion bring-up for an early prototype.

Due to the high number of electronics and actuators, this project required thorough design from the parent electrical architecture documentation down to the individual cable and E-Box wiring schematics. The cables and E-Box were specifically designed to reduce the electrical noise generated from the high motor currents while also allowing enough flexibility for the arms to achieve a large region of motion.

The motors also required individual bring up for each axis of motion. This entailed configuring the motor controller with proper safety limits, coupling the motor motion to magnetic encoders, and PID tuning to achieve desirable motion profiles under maximum load conditions.

Ceiling Mounted Asset Tracking Device

I designed the mechanical parts for an original prototype for Radar. Radar develops precision asset tracking systems using RFID and computer vision and needed a mechanical enclosure for ceiling-mounted electronics that were in development.

One of the principal challenges for this project was devising a method to exhaust the heat generated by the control boards in a sealed volume without using fans. To achieve this, the control boards were mounted using thermally conductive standoffs, thermal interface material, and a slight interference fit to ensure good thermal contact to the Aluminum Base plate. I then submitted multiple base plate designs for thermal simulation so we could further ensure that the heat would be exhausted in worst case conditions.

Additionally, there was a desire to reduce the COGS of the overall product. Some mounting brackets were designed as injection molded parts with common features to reduce part count while other mounting brackets were designed in sheet metal to reduce production costs. The Aluminum base was also designed so that it would be cheap to prototype using CNC machining, but the design could be easily modified at a later date when volumes demand the part to be die cast. The overall part count was reduced by merging distinct parts and cable routing features to simplify assembly/production.

I also designed the exterior of the enclosure so that it would be installable in a number of different environments, and the cosmetic cover was designed with a high degree of collaboration with an Industrial Design team.

Pogo Actuator

A Molecular Diagnostics client was developing a microfluidics cartridge, and I worked as a mechanical engineer to develop part of the instrument that forms an electrical interface with the cartridge. A simplified Front View of the “Pogo Subsystem” is shown below in the “Unclamped State”

and the image below shows the “Clamped State”.

Due to severe space constraints, the Pogo Subsystem needed to be roughly 2D as depicted above. The workflow was:

1. Load cartridge (into the screen) like a tape into a VHS player (Yes, I still remember those) so that it sits on the reaction block (pink).
2. The instrument motor drives a series of cluster gears that then rotates a lead screw.
3. The rotating lead screw pulls down the Yoke (brown) which compresses both the springs above the pogo block (purple) and the pogo probes against the cartridge.
4. The control board (green) now has electrical contact to the cartridge via the pogo probes.

Design of this system faced a number of challenges. Firstly, due to the number of pogo probes needed, the output force of this subsystem needed to be on the order of 150-200 lbf to guarantee electrical contact. Don’t get your hand caught in there! This required Finite Element Analysis to verify the Yoke could support the forces without significant bowing. Additionally, there were initially challenges in getting the desired force without the motor stalling. A detailed load path analysis in the drive train highlighted unforeseen friction, and the issue was eventually resolved using a thrust bearing at the bottom surface of the lead screw.

Another aspect of the design was that the reaction block essentially set the cartridge reference plane or “datum” for other subsystems. As a result, the reaction block was designed to maintain a tight tolerance between the floor and reaction surface, and it also had other precision locating features that other subsystems could use.

Finally, the wide aspect ratio of the yoke was prone to binding, so careful consideration was needed so that it could slide over the guide rails with low friction and without over-defining the assembly. The chosen design path was to use custom designed hole and slot bushings that were pressed into the yoke and detailed assembly documentation to avoid over-defining the lead screw axis.

Personal Projects

In Progress

neilfoxman · GitHub