Before we break into the data, we'd like to establish a bit of nomenclature for clarification. This experiment is being carried out in V 1.3 cells, however it is using V 2.0 protocol. This may be an important distinction in the future, so we'd like to be clear now. Remember that in V 1.0 the first two versions of replication cells were in H2 atmosphere whereas the key distinction in protocol V 2.0 is the ~1 mBar dynamic vacuum. 

We've power-cycled both active (A) and control (B) cell NiCr wire with independent power supplies using the same control script. Also remember that neither of these cells have Celani wire installed yet. Luckily the power steps only rise to a 25W max in 5W steps with 1hr time constant, so we can finish a complete cycle in ~5 hours and it doesn't take too long to really narrow our confidence interval. 

That being said, running 6 calibration cycles has cut the confidence margin roughly in half as we doubled from 3 cycles. The environment is relatively static but not perfect, so we're unsure of how much this margin will improve with more cycles. We mentioned in an earlier post that the direct flow of air from inlet window to vacuum hood easily tainted our validity, but this offset has been fixed with a polycarbonate spoiler on the inlet window to disperse the current. It wasn't really exciting enough to warrant a photograph. 

Below is the table of calibration data. The values in bold represent the power margins that we must exceed to suggest anomalous energy in our active cell. Cell A's calibration results are displayed above Cell B's at 95% confidence to the left and 99% to the right. It's also important to note that T_Ext Rise values are the difference of the external glass temperature above the ambient. This is all just a refresher  (plus we're trying to make blogs a little more lay-person friendly). 

We can reasonably operate with the 95% CI and call anything above a third of a Watt excess energy, but there will always be elusive contributors to error that reduce this interval even further. Bearing this in mind, a 99% CI is our preferred standard, and we have a ~0.5W resolution in its widest margin. We think we could do better. 

The best answer right now is to make changes to the hood and surrounding environment to make conditions a little more constant. Our first step is to pull down the hood sash to eliminate sporadic eddies in air current. We are certainly open to more practical suggestions. What do you think?

This is the exact reason we're making moves to improve our calorimetry! The explicit goal is to better insulate the cell from ambient. 

Our next step will be to gather SEM images of Celani wires before installing them in the V 1.3 cells. 

Thanks for reading!