February 2nd, 2008
Ran a constant amperage test during today's prep cycle. Started off cold at 27.9C and used the variac to adjust the voltage during the cycle to keep the amperage at 2A. As the test progressed over the next 1.5hrs, the unit gradually gained heat (up to 40C) and the voltage adjusted down from 197V through to 176V pic -> graph. Did an output test using a 2 litre inverted bottle and found the output to remain constant 2 lpm during test. Bit disappointed as that only worked out to be around 84% of Faraday. I guestimate the output has at least halved even though now using more watts, from what I first observed when the unit first fired up (using just under 300 watts - 1.6A at 185V) on Jan 25th.
February 1st, 2008
Tried three extra strengths of KOH today. Went to 100g/4ltrs, then 150 and finally 200g/4ltrs. The amp draw around 165V hasn't changed much at all, although the increase draw to over 2A with 200g strength is now down to 180V (2.2A). The 180V point has been gradually increasing as the KOH strength changed. Output is still nowhere near what it was with the original s/h mix, although it's getting better.
There is maybe another reason things changed during unit modifications - the plates were stored in distilled water and then installed straight back into the unit after the mods. Originally I soaked the newly ground plates in a high strength NaOH mix overnight and when assembled the next day, already had a slight yellow tinge to them all. That's maybe another thing I can test next disassembly or maybe I can mix a bit of NaOH into the KOH mix and see what happens.
January 31st, 2008
Drained and re-mixed electrolyte to 100g/4ltrs (was 50). Refilled and restarted testing - very interesting....even though the electrolyte is supposed to be the same strength as the old one (which was old s/h stuff from 61 plate unit conditioning process), the output gas volume is nowhere near what it was originally!!!. Three things that I know of have changed 1) now using distilled water (from solar still) instead of RO water bought from store 2) the insulated 1/8" steel rod that was inside the locating plastic tube is no longer there. 3) the four threaded bolts that were on the outside of the tube bolting the two end plates together are no longer there
Looks like there's several things I'll have to try over the next few days, HAVE to find out what the contributing factor(s) to the really really good initial performance was.
1) mix a new batch of electrolyte (same KOH strength) with RO water
2) next disassembly, replace the internal steel rod and seal the plastic tube with it inside
3) temporarily tape the four threaded rods against the side to see if that has any effect
January 30th, 2008
Finally re-assembled unit after changing a few internals. Charged it initially with new electrolyte 1/2 strength of last KOH mix - now well into cleansing/prep procedure once again.
pic -> unit back together being prepped again
pic -> closeup of new recessed end cap
January 29th, 2008
Well that was interesting!!! Just had a hydroxy explosion/implosion inside my 100 cell unit just after I re-assembled it with the new screw on end caps. On starting it up, thankfully I filled the unit right up to the top knowing full well that it would not perform properly until the level dropped enough to make the cells sealed again. As a result, the whole thing wasn't totally full of hydroxy when something shorted internally. There was a loud crack inside the unit, the bubbler water seemed to lift up violently, then the unit's plastic output tube flattened slightly as the bubbler was half-evacuated in an instant - lost about a litre of tap water into the unit in the blink of an eye - unbelievable to witness.
As a result, I've had to fully disassemble the unit to a) clean it out properly, b) find the source of ignition.
I think the short was a result of me forgetting to seal the ends of the index rod - a polyethelene irrigation riser with a steel rod centre. This is used to make sure the discs dont rotate inside the cell, keeps it all aligned. Having the electrolyte in contact with both ends of the steel rod probably provided a direct path for 150+ volts from one end to the other, resulting in an arc to one or both of end plates. I knew there was something wrong when I started it up as it was using more volts and putting out hardly any gas at all. Couldn't have predicted what was going to happen though.
January 28th, 2008
Looks like I overfilled the cell last night and didn't check on it for 1.5hrs. As a result the temp got up to 63C, the end plates and Sikaflex softened enough to allow the unit to leak slightly from each end plate. Disassembled unit today and have almost completed screw-on cap modifications to replace bolted end caps. Will seal with some large 'O' rings I had left over from earlier D18 experiments.
January 25th, 2008
Rushed to get cell assembled and forgot to align the discs perfectly and as such, they have a slight twist along the length. Not enough to impact the function of the cell and after I cleanse and prep the discs, I'll disassemble to clean any muck out and then re-assemble properly next time.
pic -> 100 cell cleanse and prep
Having said that, I've had a bit of a realisation while playing with the cell to find the best amperage/voltage to prep it at. Somewhere along the way while researching electrolysis on the internet, I came across a chart that depicted the increasing amp flow as the voltage increased. I remember there being almost a logarithmic curve that began when the voltage reached around 1.75V, being almost vertical when it was just over 2V. While playing with the cell, I cranked it up to 2V per cell to try and get 2 amps to flow through, but only managed to increase the amps to 1.5-1.6 amps from 1 amp at 1.85V, BUT the amount of gas spewing forth increased at such an alarming rate. Only a few seconds later, the cell was evacuating itself via foam being pumped out. The amount of output formed a constant pressured violent stream, enough to start splashing the water in the bubbler so violently it started overflowing. I've never seen anything like this with ANY other experiment that I've done. This reaction was at the point around 300 watts so is well within a vehicle's capability. The discs should be able to handle between 3 and 6 amps. Don't know that I'd ever be able to put that much through it and keep any electrolyte in it.
Anyway, I can't keep running the cell at 2V because it gains heat so quickly (anything above 1.75 seems to gain heat). Have started wondering if I could keep the cell just 'idling' at 1.5VDC (at 0.2 A) and somehow inflict a second pulsed power source at 2V with 3-6 amps (with diode to protect first source). This seems to be something along the lines of what Bob Boyce is doing, but I was wondering whether just second source pulsing over first source DC would be enough to continue with, without having to go the full hog with PWM3 boards and toroids.
Also been wondering if a Tesla bi-filar coil wound around the cylinder and pulsed with the second source surge would affect any magnetic field created within the radial water cells, after all they have a current flowing through them so most probably are generating some sort of magnetic field as well. Will have to see if there's any way of measuring this. Just tried a compass needle. It is really attracted to something inside the cell all along the outer length, although when powered up, the deflection doesn't seem any stronger. Will have to try this again when the cell is being drained and disassembled.
Another thing that has been made obvious by these tests, when a great amount of gas is produced, the use of bubblers has to be looked at closely. Because of the constant stream from the output pipe up through the water, there is NO way that a flashback would not go right through the bubbler back to the cell. Will have to look at this closely before installing in vehicle. May have to make smaller 7 cell test unit to test different options on the bench before applying to the large cell.
January 23rd, 2008
Have soaked discs in a high concentration NaOH bath overnight, then washed off with distilled water. Most of the plates show some initial signs of passivation (yellow tinge). Have now set end plates up with bolt through countersunk s/s bolts, sealed with Sikaflex 291 - just waiting for another night for the 291 to set properly. Have loaded the main cylinder with most of the plates and spacer rings. Once end plates are cured and bolted together properly, will mount the top plate with Sikaflex-291 and turn up other way to complete loading last few plates, rings and other end plate. Hopefully all will work out right with the spacing. In hindsight, I may have been better off using some sort of screw end caps so the length isn't so critical. Oh well, will see how this one goes first before I worry about improving anything. Waterfill is NOT addressed at this stage. Once cleansing and plate prep is complete and cell is functioning correctly, will worry about waterfill issues.
pic -> cylinder loaded almost ready to cap off
pic -> end plates waiting for Sikaflex to cure
January 21st, 2008
Preparing discs now. Have decided to spin in lathe and lightly grind in two directions with small angle grinder. If disc does not become hot there doesn't seem to be much of a warpage problem.
January 17th, 2008
After two solid days of trimming and turning the 101 discs to size, I finally had a chance to load the PVC tube with the spacer rings and s/s dics as a dry run to see what the overall dimension is. Calculated it should end up around 434mm and it measures about 435mm and reasonably flat at the top. I'm really pleased with that as it means I don't have dags anywhere, all discs mount flat. Tomorrow will have a go at machining the end plates. Still considering waterfill issues so I might put off drilling the gas escape holes through the discs until I settle on what waterfill method to use.
pic -> PVC tube filled with spacers and discs
January 13th, 2008
Started preparing 4" tubular version of a 100 cell series unit this weekend. Am using 1/8" PVC spacers cut from 4" PVC pipe being used for outside container. If heat causes the PVC to soften later, will just add a steel sleeve over 4" tube to reinforce. Machined 2 X 4" X 1/16" thick end plates to start with. If I can source some 0.015" stainless 316 shim, I will try using that for centre 99 plates, otherwise will have to cut down and turn to size another 99 plates from the 6"X6" plates I have already.
pic -> jig to mount tubing length to cut spacers from