Difference between revisions of "3T Toroidal Chokes"
| Line 20: | Line 20: | ||
The chokes were modeled in FEKO 2019 to predict their behavior. Models differ from the measured behavior in SRF by around 100 MHz. We believe this is due to inaccurate calculation or modeling of the physical capacitance, which we measure to be on order of 1 pF. | The chokes were modeled in FEKO 2019 to predict their behavior. Models differ from the measured behavior in SRF by around 100 MHz. We believe this is due to inaccurate calculation or modeling of the physical capacitance, which we measure to be on order of 1 pF. | ||
| − | [[File:choke_feko. | + | [[File:choke_feko.png|300px]] |
Revision as of 12:18, 4 August 2020
Overview
A toroidal choke design with a self-resonant frequency to enable shimming through DC on 3T RF receive loops. Able to be made by hand with a 3D printer and magnet wire.
Design
The choke consists of a 3D-printed toroidal core wrapped with 22 AWG magnet wire. For our models, we used cores printed on the Dimension sst 1200es with ABS filament and BELDEN 8051 22 AWG magnet wire. Cores measure 8mm ID, 16 mm OD, and the wire wraps as many times as can fit on the core. In addition, each end of the wire wraps once more around the core (external to the first winding) to add capacitance and inductance.
Specs
SRF
The chokes show a consistent self-resonant frequency of 150 MHz. (NOTE: This is worth confirming when hand-winding, as we have an unsolved issue where certain chokes have an SRF of 2/3 the SRF of others with the same design. Having measured similar inductances across all chokes, we expect that this is a capacitive effect -- standard chokes have a capacitance on order of 1 pF)
Heating
The choke can handle 3 amps of steady state current without heating dangerously. Pictured below is an IR image measuring the peak temperature immediately after 10 minutes of 3 A of current in a closed container.
Simulation
The chokes were modeled in FEKO 2019 to predict their behavior. Models differ from the measured behavior in SRF by around 100 MHz. We believe this is due to inaccurate calculation or modeling of the physical capacitance, which we measure to be on order of 1 pF.
