diff --git a/README.md b/README.md
index 9b74c9f5ff89c4f61e47baadd506194dcda7eaca..f6c5e3f932161b142ad661ba41805176db863aa4 100644
--- a/README.md
+++ b/README.md
@@ -127,4 +127,5 @@ Next task is designing the AS5013 board and its connection to the main board on
 I've also been looking at a similar mechanism for a non-contact torque cell.  Most such torque measurement devices use a slip ring or radio telemetry to transmit the measured values to the stationary frame.  We can design a flexure (similar to a helical shaft coupling) with a portion that translates in the axial direction when subjected to torque.  Depending on requirements, we can dial up and down the stiffness and range.  We can mount an axially magnetized ring magnet (or better yet, an opposite pair of ring magnets) and measure the axial displacement using a pair of differential hall elements in the stationary frame.  This allows us to subtract out small deviations in the gap between the magnet and sensor.  Below are some quick simulations of a candidate flexure geometry:
 
 <img src='torque-cell/tc-sim.mp4' width=300px>
+<img src='torque-cell/tc-test-small.mp4' width=300px>