Well, after a lot of learning and a lot of testing I think that we are finally solving a lot of questions about types of plates, number of plates and so on.  This will be a rather long post, as I want to be sure that I describe my test setup as well as possible and then the results. 

I used my 4 Cell Plexy-glass container so that I could test all cells with water already filled and sealed without having to take everything apart to setup for each test.  I had a neighbor come over and write down results to allow me to keep things cooking as quickly as possible.  The cell setup is pictured below and I had cell GH-E001 consisting of 2 plates, cell GH-E002 consisting of 3 plates, cell GH-E003 consisting of 5 plates and cell GH-E004 consisting of our 6 tube sets.  Keeping in mind that subtracting 1 from each of the above number of plates tells you how many electrode sets we had in each cell.  With the exception of the GH-E004 cell which actually has 6 sets of electrodes. 

For a brief background, the goal of the tests were to try to determine if the number of plates or use of tubes improved output, or if the output is strictly a measure of current.  If it turned out to be a measure of surface area, we hoped to determine if there was some critical point at which things stopped improving. All tests were performed while hooked to my running F150 and voltage readings were verified constantly at the cell connections.  There were no changes in voltages within each test, however, voltages did change slightly as we increased draw due to larger and larger cell surface areas. Current readings were read through a separate meter to allow constant checking and recording of changes as they occurred. Water temperature as well as outside temperatures were recorded at the beginning and ending of each cell test, however, as mentioned above, the tests were performed rolling from one right to the other so no real temperature changes were recorded. 

Gas measurement was recorded in a 400mL Glass Jar.  This was chosen because it worked a lot better than trying not to squeeze a plastic bottle while performing the capture test.  The Jar was marked from bottom at 100mL increments and all tests were performed between the 100 and 200mL lines.  Line was fed into Jar without getting any water inside it, allowing for immediate and consistent test results.  200mL line was placed at the exact water level as the water in my bucket to allow consistent readings at room pressure.  All capture tests were performed after allowing cell a couple minutes to get started and each was performed twice to insure proper measurements.

Bucket used was a 3 gallon square cat litter bucket, allowing me to pin the jar square in the corner of the bucket to prevent it from moving or becoming unlevel.  200mL line held at water level throughout each capture. I took my time with the setup to be sure everything was consistent between each cell capture and I feel pretty good that these results could easily be duplicated if needed.  Earlier test results I couldn’t say this because I wasn’t paying as much attention to jar placement and such. So, on to the exciting part: We have all but proved that cell output is related to the amount of current running through the cell and nothing to do with plates, spacing or tubes.  In all tests, the 2 plates which gave you 1 anode and 1 cathode performed better than all other cells.  It was nothing to write home about but was consistently around 15% more efficient.  I suspect that this is the case, due to less loss due to the extra joints and surfaces, but will continue to investigate.  Water level above the cell was proven consistently to be more efficient, by around 13%, with a 10mm height above the top of the plates versus a 1mm level above the plates.  As for my last test, I tested my theory about the cell continuing to heat up until it reached a short circuit state.  The test consisted of me boiling water and using it at 212 degrees as my cell water.  The thought was that as the cell ran the water would continue to heat up, pulling more and more current from the system.  This test was also run on the truck and surprisingly, heating the water some 70 degrees Celsius warmer than the previous tests, only increased the current by 22%, but did not increase the cell efficiency at all. After a period of only a few minutes the temperature in the cell began to decrease until it finally stabalized at around 68C.  All test results between all cells were within 5% of each other with the exception of the single cell unit being around 15% better.  It currently appears that adding KOH to a single cell device will result in the most productive output over all the designs that call for a great number of plates.  Now that we’ve ruled out all these complex designs, my next tests are going to be using electrolyte to see what we can get out of the cell and whether it improves efficiency.  I also want to see if using electrolyte with the timing circuits will allow me to increase efficiency, but not expecting much. 

In the end, its currently looking like we can increase the surface area to a huge size in order to split water with no catalyst, or we can use a catalyst with a very small cell design and be slightly more efficient.  So far we have not found the magic formula however. 

Oh, I almost forgot, cross hatching your cell as well as cell conditioning, also do not show noticeable increases in calculated output.  Keep in mind that I’m not comparing physical volumes here but rather mL/W/hr.  Of course in my tests the tubes output more gas in less time than any of the other designs.  However, once you calculate current draw to output, everything leveled out to the same, or pretty darn close.

Here are a couple images and I’m working on a video that I’ll be posting in the next few days.  These were taken with the top of the cell off after the tests were performed.

–glenn hancock