Peltier-based dehumidifier

This simple apparatus is a particularly interesting way of demonstrating the principle behind the common household dehumidifier:

The big difference however, lies in the way the low temperature (required for the condensation of the water vapour) is achieved.

Ion Motor

Ion motors are very cool stuff. So cool that Nasa decided to invest research effort in this subject since the early 1960's when the physicist Harold R. Kaufman achieved the first design. In spite of the small ammount of thrust (when compared to conventional rockets), ion motors are capable of a very large exhaust speed, being equivalent to ten times the exhaust speed of a high quality rocket (the later can only achieve between 3-4 km/s). This means that ion motors are capable of continuously accellerating an object up to the exhaust speed (considering the space environment, where drag is practically absent). The time for exhaust speed to be reached depend of the mass being moved.

Water Fuel Cell - DC voltage results

Today I performed some tests measuring gas production with DC voltage. The tests were performed in order to obtain aproximately the same consumed power, namely 22 Watts.

DC test performed:


High frequency test (100 KHz) at same voltage:


The DC test shows a clear increase in efficiency (almost 25%) relative to the pulsed electrolysis.
Still to compare these results with low pulse frequencies (it is likely that power losses are quite high at 100 KHz).

Water Fuel Cell - more results

For one measurement taken without the inductor coil, we have the following values:

Water Fuel Cell - measured results

I performed two consecutive runs without changing the conditions. The data obtained is as follows:

Considering that it consistently took 4 minutes to produce 80 ml of H2 + O2, this
gives a rate of 1.2 litres/hour of H2 + O2.

Water Fuel Cell - back to business

After reparing the WFC and adding extra safety measures, here is the entire setup producing hydrogen at 12 Volts DC and 1.94 Amps. Next I will provide a grid with measured values, including gas volume:

Water Fuel Cell - cleaning the mess and repairing the electrolyser

After yesterday's mishap, and of realizing how lucky I was to still be in one piece, it is time to take a deep breath and after looking into what went wrong, follow all the safety measures to prevent one such event from ever happening again.

I started by getting a new lid for the electrolyser and putting a new seal rubber in it, this time with two clamps instead of one. Now only the clamps are used for holding the lid. The original ring around the lid was discarded. By itself this doesn't provide extra protection against an explosion, apart from the slight probability of the lid popping out in one piece instead of fragmenting into several pieces (today I still found fragments of the lid in distant corners of the room).

Water Fuel Cell - HUGE ELECTROLYSER EXPLOSION

Sometimes mistakes can be the death of the artist. In this case I got closer to it than desirable: after building a tiny bubbler for testing sustained combustion I hooked it up to the electrolyser (aka: Water Fuel Cell) - one tube going from the electrolyser to the bubbler and another tube from the bubbler to the gas exit, where combustion would be tested.

The bubbler was working fine, air tight as necessary. After approaching the output from a flame, small explosions along the tube would occur, without affecting the WFC.

Water Fuel Cell - inductive load added

Just as refered by several people doing this type of research, adding and inductive element to the load would create a resonant LC circuit, with the WFC as a capacitor. I used as an inductive load the secondary coil from a 220 V / 12 V transformer which was added in series with the WFC. The most notable effect was the suppression of the high frequency signals at the WFC, along with a voltage dropout of nearly 5 volts. The curious thing however is the fact that the gas production doesn't seem to be affected for the same input voltage/current conditions. As the transformer becomes hot after a couple of minutes of operation, this shows that there is less energy going to the WFC.

This sample video shows the cell in operation under these conditions:

Water Fuel Cell - fixes and more tests

The container is now air tight. This allowed me to start doing some electrolyser efficiency tests. By attaching a small hose between the container and a small flask containing some water I could see how fast the bubbles would form, by varying pulse width and frequencies (both the carrier signal width/frequency and also the gating signal width/frequency).

Water Fuel Cell - more improvements

While the air leak problem have not yet been solved, other problems are now fixed, and some improvements have been made. The control circuit have been verified, and it was found that the 7805 regulator that was being used, didn't had the GND pin connected (oops, minor mishap). In spite of that the two 555 timers would work, as the regulator would let the input voltage pass through. The output signals were not however perfect square waves. As the GND was hooked up, the MOSFET would no longer switch from cutoff (the 4 volts output from the second 555 would not be enough to activate the transistor). So the 7805 was replaced by a 78L09 and the circuit started performing normally, delivering a clean signal to the cell and smaller heat buildup in the MOSFET.

Starting to bust water into its basic elements

The basic electrolyser design is finally finished. After a lot of steel cutting and drilling, I've finally finished the first prototype of the machine which will tell wether or not pulsed electrolysis of the water, through special frequency and pulse duration control, can by itself be the key to making the process several times more efficient than conventional DC electrolysis.

For now there are a few issues to resolve, namely:

• Make the seal on the top cover air tight - even though the container was bought under the premise that it would be air tight, in practice is was verified that it was not;


• Control circuit FET transistor is heating up too much under a 4 Amps load - must check if the gate voltage is being enough to cause it to switch from cutoff to saturation and not somewhere in between. According to the device datasheet, it dissipates up to 150 Watts of power. In this case it has to dissipate around 40-50 Watts of power (10 to 12 Volts at 4 Amps), which is probably too much for the 25 cm^2 heat sink installed.

This is the complete setup (electrolyser + control unit + DC power supply):

Water - alternate fuel source and a panacea for mankind?

When it comes the time to pay after filling up the tank, you dearly wish your car could run on anything cheaper than stinky old gasoline (or diesel). Wouldn't an entire global economy depend on this precious product, and you could establish a parallel with other things like drugs, where the desperate sense of necessity leads to the unavoidable obligation of paying a large sum of money just to obtain a given ammount of it.