In this episode, Shahriar explores a large Stirling cryocooler and its custom controller from a medical-grade ultra-low-temperature freezer:

https://www.stirlingultracold.com/refrigerators-freezers/

The internal components were kindly provided by a viewer. The original LCD interface is designed for freezer operation and offers limited control over the engine. The goal is to build a new controller and explore temperatures far below the freezer’s intended operating range.

Reverse engineering begins with the Debian-based onboard computer and its executable software. Analysis of these programs, combined with captured serial communications, reveals the protocol between the main controller and the engine control board.

New interface hardware is designed to read the temperature sensors and collect telemetry from the engine controller. Custom firmware provides control over the engine’s operating parameters, with real-time safeguards to enforce operating limits. It also incorporates a PID temperature-control loop and a web server for visualizing data and adjusting settings.

With the new controller, the cold head reaches approximately −160°C which is far below its temperature during normal freezer operation.

Please consider supporting The Signal Path:
https://www.TheSignalPath.com/Donate
https://www.Patreon.com/TheSignalPath

Other links:
https://www.TheSignalPath.com
https://www.Twitter.com/TheSignalPath

In this episode Shahriar investigates a faulty Agilent E8257D analog signal generator. These high-performance synthesizers are in wide use & high demand. The instrument fails self test procedure. The error code "704 Filter Test" generally points to the Frac-N module which also houses the modulus divider circuits. This portion is responsible for generating frequencies below 3.2GHz and uses an integer divider, filter bank and variable gain amplifiers.

Using a 12GHz active probe, the signal path (pun intended) is verified step by step from the input to the suspect sub-system all the way to the output of the Frac-N board. The fault is ultimately traces to a custom MMIC module with an Agilent GaAs FET amplifier circuit. A new replacement IC is sourced from eBay which clears the fault. The phase noise performance of the unit is also verified at 2GHz using an Agilent E5052B signal source analyzer.

Lastly, the faulty MMIC module is de-capped and the die is analyzed under a high power microscope with various imaging techniques.

Please consider supporting The Signal Path:
https://www.TheSignalPath.com/Donate
https://www.Patreon.com/TheSignalPath

Other links:
https://www.TheSignalPath.com
https://www.Twitter.com/TheSignalPath

In this episode Shahriar combined two Newport 1936-R optical power meters into a dual-channel model 2936-R. The teardown of the unit as well as the sensitive acquisition boards are presented.

The model conversion is rather simple once the core firmware is analyzed. After a successful conversion, the unit's performance is verified using two identical photodiodes and a HeNe Class IIIb laser. Lastly, trichromatic prism optical combiner is used to measure both the transmission & isolation performance from one of its colored ports.

Please consider supporting The Signal Path:
https://www.TheSignalPath.com/Donate
https://www.Patreon.com/TheSignalPath

Other links:
https://www.TheSignalPath.com
https://www.Twitter.com/TheSignalPath