Water Power


Run your car or bakkie on water with a HHO or HHOO generator.

Home Contents Information Archive A Basic Cell

The Beginning

You may e-mail me at john@hydroxypower.co.za with questions, suggestions or if you can help in any way.

Latest update   5 January 2012

Latest Update
The Beginning

New! How to build a basic hydrogen-on-demand wet cell.

Keep it green with organic Septic Tank and Pit Latrine Treatments



January 2008 to July 2009:


Out of the hundreds of different designs, configurations and combinations that I have tried, I have eventually settled for a hybrid of the Bryan Martin cell and Eddie Batista's functional principles.

I have stayed with Bryan's original cell of 9 plates which I have more-or-less set out as per his configuration (I have not kept to his spacing between plates or the sets), except that the end sets are configured for HHOO as per Eddie's idea (i.e. I have swapped the positive and negative around so that it produces HHOO instead of HHO).

                                                                    See photos of what all I have experimented with!


The wiring schematic is also just below.

I have not staggered the plates as per Eddie's design as the dimensions of the container, which just fits into the small space that is available under the bonnet, does not allow it.

The dimensions of the plate sizes are roughly ala Eddie, 9cm x 19cm. They are 1,5mm (or is it 1,6mm?) thick. I cut the plates out of scrap non-magnetic 304 grade stainless steel with an angle grinder. The plates are far from evenly sized and the (crude!) rounding-off was on a small bench top grinder.

I use 8mm rod or bolts, 13mm nuts, and 8mm washers & spring washers (all ss). For my next project I want to try 6mm bolts. Since the nuts are an integral part of the spacing of the plates (in my designs), I think that the smaller nuts for the 6mm bolts and rods might give me more flexibility as far as experimenting with spacing is concerned.

I could not find any threaded nylon rod anywhere so I used the plastic nuts and screws that are used to fasten on car number plates! They are just long enough to hold 3 plates together with spacing of 2mm between each of the 3 plates. I punch my own spacer washers out of the plastic strips that you find hanging in front of cold rooms, such as the ones which supermarkets use for storing their fruit & veg.

I set the start up amps at 8,5 and it produces 1,25 litres + of HHOO per minute when it gets to about 9,7amps (after about 5 minutes). I have a problem with the fuse (rated 30A!) overheating big time (why? anghasi!). So I keep the operating amperage under 9A and the gas production in the region of 800 - 1 000ml/ minute. I must mention that this amperage includes the approximate 1A which is drawn by the fuel pump (the electrolyser and pump runs on the same circuit).

At the moment I am using a plastic container which is sold as a microwavable and freezer friendly spaghetti canister, capacity 2,1 liters.

I keep the electrolyte at a level just below the ss bolts and nuts connecting the plates, and the connecting strips from the electrolyser to the bolts which pass through the sides of the container.  Therefore I only fill it with about 1,5 litrers of electrolyte and as a result the heat (and amps) climb very quickly.

COOLING THE HEAT                                                       

I have incorporated a very simple cooling system which works fairly well. I bought a copper core radiator, the under the dashboard type for the interior heating and cooling of a car (Toyota I think?), from a scrap yard. I also bought the cheapest petrol pump, make unknown (I just asked for the cheapest that the motor spares shop had). On the side of the electrolyser container I drilled 2 holes for brass fittings. Inside the container I connected plastic tubing: from the one brass fitting the tube goes right to the bottom of the container (for the return flow); the tube from the other brass fitting reaches about halfway down (for the out flow). 

When this is connected to the fuel pump it circulates the electrolyte through the radiator and via a reservoir back to the electrolyser. I have fixed a large fan (from a very old computer power supply) against the radiator which in the circumstances does a marvelous job of keeping the electrolyte cool. 

This does help to control the heat and amperage so that it does not run away. Unfortunately because the volume of electrolyser is still too little, I do have to limit the amperage as it will run away (notwithstanding my cooling system) if it starts off at 9A or more.

The real answer would be a PWM controller. The good (?) ones that I have seen on the net: one type costs $55 + $9 shipping to South Africa. The other one is $199 + $19,95 shipping to SA! I don't know of anyone making them here in SA (specifically for this purpose).

There are many good free PWM controller printed circuit designs available on the net, unfortunately I have no electronic aptitude to put something like that together.


When I first started experimenting I tried various electrolytes: salt, bicarb, sodium hydroxide (drain cleaner: do not use the type with aluminium pellets!!) and finally potassium hydroxide (KOH). The KOH is a very corrosive chemical, as is drain cleaner! After running my electrolyser for about 2 weeks I stripped the assembly to see how it looked inside. The ss plates had started corroding badly on the edges. I can't remember whether it was the anodes (+) or the cathodes (-). I have never seen this before so I figured that the KOH was reacting with the copper in the radiator and this was causing the ss plates to corrode. I eliminated drain cleaner and salt as alternate options for many reasons and tried bicarb. Instant happiness!

Choices of electrolyte:

KOH (which is somewhat better than drain cleaner) not much is needed but it is corrosive and without any controls (electronic or otherwise) it causes the amperage and temperature to run away very quickly.

Salt produces a chlorine by-product which is corrosive and poisonous.

Bicarb: irrespective of the negative by-products (if any), it is still the best electrolyte for the setup that I run now. I have found it to produce much more gas for much less hassle.  


Often different cells will give off a lot of gunk. I found that out of all the electrolytes (drain cleaner, potassium hydroxide, salt and bicarb) using bicarb generally resulted in far less gunk than any of the other electrolytes!


I use a good quality pine gel to clean the plates, containers, etc to get rid of unwanted gunk & dirt, grease, oil, etc. Always rinse thoroughly afterwards!


I admit that I have never been big on cell conditioning. Probably for two reasons (1) the stainless steel that I use is of poor quality which results in a lot of gunk in the water and on the plates, and (2) I have never had a suitable power supply which can run for long periods without overheating or the amps in the cell running away

I take note of what the guys who have successful cells have to say about this important aspect. Until I have cells made out of grade 316 ss and a good pwm to run it, I won't be able to run this important process.


This is how I wired the electrolyser up. I also incorporated an LED which shows when power to the electrolyser is on. This is very handy as it not only tells me whether the system is operational, but it also shows me if the circuit breaker trips (the LED goes off).


Here are some photos of my experiments. Unfortunately I did not keep a dedicated or sequential record of everything.

I have never had the money to buy "new" stainless steel so I first experimented with anything I could get my hands on at the scrap yards.

Unfortunately my initial research led me to web sites where they where generating hydrogen by reacting aluminium with drain cleaner (Sodium Hydroxide) and here I wasted a lot of time. Yes, it does produce hydrogen, and even more so when you apply some volts and amps.


But as you can see the aluminium plates corrode and then you find in the bottom of the container...

...some nasty stuff! When you switch the power off you will find that the reaction does not stop. The production of hydrogen when aluminium is introduced into a Sodium Hydroxide (or KOH) solution gives an ongoing continuous reaction which only stops when all the aluminium is consumed!


Here are some of the earlier designs that I tried. Coils...I tried them in every configuration and size imaginable. + -, + n -, and + - +. Not bad producers for their size, but they suck amps!

I tried them in parallel, in series (combinations of 2, 3, 4, etc), but to produce anything noteworthy the amperage had to be high. And they gave off a lot of gunk because of the low-grade stainless steel.

To the left of the table are some of the other containers and cell designs that I have tried. My biggest problem always was to get the containers to seal properly. It may not leak any electrolyte, but gas has a knack of leaking from anywhere!

In the background (on the table) you can see some ss discs that I cut which I tried in the cell below.

The discs fitted on a ss threaded rod, with nuts as spacers, which fits down the centre of the tube. The discs and rod were the negative and the ss tube in the pipe the positive. Did not produce that much gas for the amount of amps that it pulled. Amps and heat ran away VERY quickly.

Remember... amps = heat and heat = amps! We already start off with relatively high temperatures here on the Natal lower south coast - even in winter the minimum seldom falls below 12 degrees centigrade! At 9 - 12c we start hunting polar bears!

I bought this 1,2m x 1m roll of low grade perforated ss sheeting for about R20. I had high expectations as far as it's production capabilities were concerned, because it was perforated. But it gave off a large amount of gunk and the amount of gas that it generated was not too wonderful. That puts paid to the theory of "cross hatch" sanding plates to generate more gas, as far as I am concerned. Eddie Batista tells me that in his opinion sanding only contributes to the wear of the plates!  Although his cells do not produce as much gas as some of the other boffins' designs, he is one of the very few people who can show 100% plus increases in economy and therefore his opinion counts in my books.

I tried using this sheeting in many different designs & configurations...


I hammered the inside drum of a washing machine flat for the next coil cell... a lot of effort!

All these coils were a pain in the gat to roll, fit together and space so that the plates do not touch.

I tried this in different configurations: + -, + n -, and + - + for HHOO. The + - configuration (HHO with no neutral) worked so well I named it the "Skop Gat" cell. Unfortunately, although it produced a lot of gas, it sucked very high amps - even when I had 2 in series. I didn't have a multi meter at that stage but it would flatten a fully charged car battery (heavy duty) chop chop and within a minute the connecting straps would be very hot and start to cook the electrolyte. The poor grade ss also gave off a lot of gunk.

I even used the back of the drum to see what these plates would give me! Result? Not much better than smooth, solid plates! I also could not beat them completely flat so they stayed warped.


I was using power from a car battery that I would frequently recharge until eventually the battery died (more from old age than for any other reason: it was already two years + past the warranty date!).

I then saw on the internet how to make a bench-top power supply from a power supply for a computer. Older computers have better power supply systems in that the 12 volt rail is rated in some cases as high as 20 amps. Look on the side of the casing to find out what it will give you. But be warned - the power supply will only allow you to draw up to about half of the rated amperage (and then for only a very short while) before it shuts down. With luck, if you quickly switch it off and leave it for a minute or so it will reset itself. But if you see smoke.... wave it goodbye!

I have learnt that it is best not to allow the electrolyser to draw more than half the rated amperage of the power supply. So I usually keep it at or below 45% of it's rating. But even at this limit it will not last long. For long, continuous running I try to keep the amperage at or below 30% of the rated capacity. This limits the amount of testing that you can do and also it restricts the amount of gas that you can generate, but at least you will still have a power supply! I understand that you can connect these up in parallel in order to provide more amperage. I will try this sometime and post the results.

As you can see I put in an on/off switch and a led (just visible on the side of the note pad) for visual reference. This particular power supply is a very old one and only rated at 7 amps (12v + rail). This one I can run closer to 75% for fairly long periods of time. I have been told by the technical guys at the computer stores that all the modern power supply units are built to last only for 1 - 2 years!

Connect all the black wires together for the negative 12v rail, and all the yellow wires together for the positive 12v rail.

Bridge the only green wire on this plug with any of the black wires. This supplies power to the fan and puts a load on the system to make it work (as I understand it). All the other black wires have been cut off and joined together for the 12v negative rail. All the yellow wires have likewise also been cut off and joined together for the 12v positive rail.

Put a fuse on the positive line!


When I tried Eddie's HHOO design I just could not get the same results as him. I know that water (tap or otherwise) differs from town-to-town and country-to-country (he is in the USA), so this would be the first variable. The second would be the ss plates that I used. I think Eddie's might be a better grade, but at the least his are uniformly cut! My plate spacing is also about 1mm smaller - theoretically better (? depending on who you listen to!). The cell (+ - +) drew very high amps with reasonable gas production but it got hot quickly. Our temperatures (even in winter) I think are higher than what Eddie experiences in his part of the world during their summer, so this would also be a great factor. I also tried using neutrals between the active plates. I had to increase the amount of electrolyte to get the same amount of gas as before, and then the same problem also arose, i.e. increased amps = heat = increased amps = increased heat etc. Oh well...

As you can see it also produced a lot of gunk which blocks the production of gas. The many holes in the plates are from all the different configurations that I have experimented with!

This is the container in which I do all my initial testing, it holds about 4 liters of electrolyte.

This is a variation of one of Bryan Martin's designs that I have been experimenting with lately. I have added 2 extra neutrals (4th plate from the left & 4th plate from the right). If I had nylon threaded rod instead of having to use the short plastic number plate bolts, then I'd make these extra neutrals true floating neutrals, i.e. they would not be connected to any of the other plates. The design running in my car at present (up to 4 August 2009) is the original 9 plate configuration as described at the top of this page, which is like the one below but without the 2 extra neutrals. It was with this cell that I saw for the first time how important cell design is with regard to temperature and gas production. I regard Bryan as a leader in this regard.


I previously tried using 110mm standard and heavy duty (2mm and 3mm wall) pvc drain pipe for the electrolyser housing. The screw-on end-caps are expensive and it is a pain in the gat to eliminate leaks.

At first I (foolishly) tried everything that others were doing, like fitting aquarium one-way valves for the cell to suck in outside air. This valve fitted into the little pipe on the left. This is only necessary if you place the cell under high vacuum. However, even when I put the gas line right up to the intake manifold it still did not suck enough to give negative pressure in the cell. As a result I found that in my applications this feature was therefore totally unnecessary.

My present designs do not fit into these pipes. My best previous design was 13 plates with 2mm spacing between plates, +n-n+n-n+n-n+. This cell just squeezed into this pipe!

A 20 liter bucket is one of the first big containers that I tried, the idea being that a lot of electrolyte (12 - 15 liters) would help with keeping the cell cool. Even with this amount of electrolyte it would still heat up within 20 minutes to an unacceptable level. I had to mount it in the boot and ran the gas pipe inside a hosepipe under the car to the front. The bucket is a "triple lock seal" ...  ja boet! Even with lots of silicon it was very difficult to seal properly. I tried plates and coils in this container, but heat remained a problem. This was before I tried using many neutral plates as in my 13 plate cell.

I also tried the 10 liter bucket in the left foreground, orange lid on the table. It was impossible to seal this one as the pressure from the gas generated by the unit just kept on finding leaks at the lid! This lid had no "lock" and even rubber straps over the lid did not help.


The container that I am presently experimenting with is a 2,1 liter spaghetti canister which is the tallest one in the range (285mm high). I contacted the manufacturers to find out whether they perhaps have anything similar with a greater capacity (say 3 liter or more). The answer was "no", only the lunch-box types have a greater capacity. Unfortunately this is the only shape (spaghetti canister) that will fit in anywhere under the bonnet.

Preparing the inlet and outlet for the circulation of the electrolyte through an outside reservoir and radiator for cooling. The pipe for the return (inlet) is not in place yet. Tubes are 6mm and 8mm (inner diameter).

I drill the holes slightly smaller than the brass fittings. These were the only type I could find which are threaded at one end and barbed at the other. I first bought them for about R12,00 (for a pair) at a local hardware store. They now cost R20,00 each! Gooie genugtig!

I heat the brass fittings with a "hairdryer" and then turn them into the holes.

The anti-splash guard on the inside of the gas outlet. I also silicon the lid along the green rubber "O" ring to give a proper seal.

The gas outlet on the top. The fitting takes a 6mm (id) tube.

The completed electrolyser (well almost; the inside tubes for circulating the electrolyte have not yet been put in). The strap connections from the cell to the terminal bolts are covered in plastic tubing to prevent current leakage.

The electrolyser in place just behind the aircon radiator (to the right). The exhaust manifold is to the left. I fitted a sheet of ss to shield the electrolyser from the engine and exhaust heat. The scrubber is in the foreground. The electrolyser sits on a L shaped wooden bracket that is fastened with a leather strap and cable ties to an engine support member.

The scrubber is made out of 50mm pvc pipe with the end caps silicond on. It is filled with 350ml ordinary tap water (total capacity 500ml).

With everything connected up, part of the air intake channeling in place.

With everything in place you can't see the electrolyser.

Here is a water trap that is fitted between the scrubber and where the gas enters the intake manifold. The plastic container is a 125ml screw-cap jar. The inlet on the inside has a thin tube reaching to the bottom of the jar. This tube goes through 2 foam rubber discs which will hopefully prevent some of the water vapour returning up to the outlet.

This is the 3 litre reservoir on the other side of the engine compartment, filled to about 2 litres. The + and - symbols by the barbed fittings are relics from when a plate or coil cell was previously bolted onto this end cap. In the photo, just below the reservoir (to the right of the silver foil) you can just see the fuel pump to pump the electrolyte. Just to the right of the reservoir (obscured by the power steering fluid reservoir housing cover) is the radiator (for cooling the electrolyte). The car's main radiator is (in the photo) just below the reservoir (you can see the hose from the radiator to the engine clearly). The alternator is to the left just under the radiator hose. This shows how little space I have to work in!  


Leaks do frequently occur and are the bane of any electrolyser system! To seal the containers for electrolyser leaks is easier to accomplish than to seal it from gas leaks.

Never take any seals or joints for granted! I use silicon everywhere after first ensuring the best mechanical seal possible.

If a cell seems to be producing reasonably well but there is no improvement in either power or consumption, then suspect a leak. A good test is to extend the gas pipe from the scrubber (instead of feeding it into the air cleaner/intake manifold/carburettor) into the bottom of a bucket (at least 10 litres) full of water. I like to use a 20 litre bucket which increases the pressure against which the electrolyser must "push" to get the gas out. I then place this bucket on the ground, not at the level of the electrolyser. Start the engine and then switch the electrolyser on. If there is no gas coming out of the end of the pipe at the bottom of the bucket within about 15 seconds then you can be certain that you have a leak somewhere. Try using soapy water on all the sealed joints to find it.

Even a little gas will have some effect on the consumption, so check for leaks if you experience no gains!


I wrapped the exhaust manifold sensor in tin foil which I found works better than anything else that I have tried as far as fooling this sensor is concerned.

I have an efie but is designed for another type of sensor, so I can't use it on my present car.

I also made my own simple efie (design compliments of Gary at Chemlec).

But without access to the equipment to test the exhaust emissions I don't know how to calibrate it. I have tried various settings with and without the electrolyser running, but I have not noticed any difference. Perhaps if I persisted with a few tanks of petrol then I might get somewhere, but right now I don't have that kind of money or patience!.

I fitted it as per Gary's instructions and I can switch it on or off from the dashboard (switch on the far left). I am keeping it switched off at the moment and I am relying on the foil wrapping to fool the sensor. To the right of it is the on/off switch for the electrolyser, this has a led to show whether there is power to the unit or not.

Another sensor (MAF?) between the air cleaner and the intake manifold. All the better to improve the onboard computer's ability to control the fuel "economy"? ... ja boet.


My pill container housing for the 30A relay (under the bonnet and just in front of the computer, firewall immediately on the right).

The 30A fuse between the relay and the battery positive terminal, I boppa'd it with a cable tie to improve contact. This did help a little so that the fuse does not get too hot too quickly.


Where the gas goes into the pipe which links the air filter to the intake manifold. Connected to the brass fitting inside the intake manifold pipe is another length of clear tubing which then runs to just before the manifold itself.


I use copper core wire that I got from an auto electrician, rated 50 amps (dis lekker dik!)

The 3 types of fuses that I have tried (the one on the right was only available in 25A at that time, the other 2 are 30A). The first fuse (on the left) is useless as it easily falls apart when it gets warm.

The first type of fuse holder that I tried (using the 2 fuses on the left in the above photo).  The fuse in the photo below is 20A. This fuse holder (with a 30A fuse) would quickly heat up when fitted with the electrolyser in the car, even though the electrolyser was drawing less than 10A! Even after soldering reinforcing copper leads inside the fuse holder to reduce resistance it still heated up.

I felt more confident with this fuse holder as the connecting leads where much thicker. Huh! It still heated up quickly and the fuse became very loose as a result! 

Finally I found this one which takes a ceramic fuse but sadly with the same result, it also heated up quickly! 

However, when I tensioned the fuse holder up with a cable tie to ensure better contact (and I also lengthened the connecting cable to the battery) then it improved the situation somewhat. It still gets hot but not as quickly.


On the Chemlec electronics forum it was suggested that the fuse holders that I have been using were of inferior quality and/or wrong for cars. Well, they were all bought at an auto spares shop... so anghasi.


I have an aluminium radiator (under the dash heater) out of an old Merc which I intend using in the future if I ever get another car which has the space... I must just remember to only use bicarb as the electrolyte with this radiator, with KOH it will quickly corrode away!

I blocked off the centre port with an adapted lid off a pill container (sealed with silicon).

And here is a cooling fan set-up that should do the job, it produces a nice breeze. These 12v fans are from blown computer power supply units. The above radiator is reasonably covered by these 4 fans. I'll use these for now as they are for free! If I find (as in free or very cheap) a suitable pukka radiator electric fan which can fit into whichever car I use this radiator in, then I will rather go with the bigger, single fan.

Here are some cooling ideas which I previously used with limited success on 110mm water pipe electrolyser housings. The cooling fans and heat sinks are from computer processors.

These were wrapped around 110mm water pipe electrolyser housings to try and dissipate the radiant heat.


As at July 2009:

I have not had the money to fill the petrol tank every time so as to properly measure the improvement in consumption. My guestimation is that with the present 9 plate cell I am achieving approximately a 50% increase on the open road, and 35% in town and 40% on the short stretches between towns (depending how long I drive at 80kms/hr). Not too shabby, n?

I can't wait to hit the open road again, I am confident of pushing that 50% mark higher with the tweaks that I have made. Watch this space...

My aim is still 200%...and I can and will attain it!

Nou ja...

Well at least if nothing else goes right I can still cheer myself up by looking at the beautiful view from our house. We are truly blessed!


At sunrise...

At sunset...


Home August 2009


Free website hit counter


The information presented on this website is for you to look at, learn from, laugh at, or whatever. But if you try anything that you see here it is at your own risk. I will not take responsibility for your stupidity should something go wrong.