Though the airwaves have been quiet, we've been scrambling to keep our data stream up and running. The newest software version hasn't behaved lately, and the past week has been devoted to sorting the bugs out. The best candidate to test software fixes belonged to the RF test and it was temporarily comandeered for the good of the experimental network. But great news! The fix is in and we can again start where we left off. 

Since our last post we've gotten a bit more detailed with the loosely-written protocol, assigning specific wavelengths and duration, and even adding a backup plan in response to recent concerns. Take a look to be better informed with our plans for this test. 

The powder resistance is quite the beast to deal with. It's shifty behavior makes designing a good test of frequency verses resistance tricky. Our initial impedance measurements varied from 15 Ohms to ~20 MOhms due to the varying electrical path through the powder. If the electrons can forge a stream from the passthrough to the cell wall, then we see a workable 15 Ohm impedance. However the slightest physical agitation destroys this path and impedance skyrockets to 20 MOhms, essentially rendering our waveform generator's small amplitudes useless. The natural solution to this problem: increase the voltage across the media and force a path through the powder. 

Here we have the circuit that can tame our wild powder and bring back some control to our RF experiment - a simple xenon flash tube circuit from an old 35mm film camera. This 3V-powered transformer circuit outputs 300+ Volts across the flash tube, a small, useful package to employ on our unrelenting powder media. As needed, we can discharge this high voltage circuit to whip up some good connectivity across the cell. 

Its manual trigger has been hard-wired with an optocoupler to discharge the flash tube every 16 seconds - controlled with the waveform generator. Yes, this is slow. It's limited by the 3V power supply with ~250mA current. The 330V at 200A could be a pretty good combination, especially for the super cheap circuit. 

Currently we're trying to figure out a way to reduce the rise time of the flash. It takes 4 microseconds to get full power (at 200A). That's pretty slow on our operating scale. . .

Who knows, maybe we'll even see interesting enough results to pursue lower frequency but higher voltage AC power across the cell. The possible avenues for future experimentation seem plentiful at this moment in time. 

What we DO know is that the future powder experiments look pretty fun right now. We will be trying this same protocol on a pure 40nm Ni powder cell. We will pull one of the other cells off the reef and monitor how H2 flux affects the lattice before we even apply AC power. The pure powder, baked out to remove all impurities and oxides, reduces the chemical uncertainty we've run into with cell 3's BaTiO3 mixture since it may have been reduced already and rendered useless.

Also keep on the look out for an oil-based powder experiment with hydrogen loading from melted phenanthrene. 

Got any more powder experiment ideas? PLEASE let us know! Diagrams, protocols, theory papers, metallurgical knowledge - all welcome. 

Our next step is to put up a lexan blast shield around the cell and start the RF characterization of unloaded powder. Fingers crossed!