HHO generator

The mayonnaise jar has two pair of stainless steel strips separated with nylon screws and spacers. The green color is from the addition of 20% anti-freeze.
Ethylene glycol is a bad idea! See below...

The first test was with the jar only and the vacuum tried to collapse the jar, so the check valve and aquarium filter (blue) were added to provide a vent. At idle, the engine ran a little rough but applying power to the H2 generator smoothed the idle out perfectly, and the vacuum increased by 1/2" of Hg. I found this very encouraging, especially since the engine is so large... 7.5 liters.

This test was with 1/2 teaspoon of sodium hydroxide (NaOH) as an electrolyte. Current was 10 amps at 12 volts (120 watts).

At this point, the jar was tested inside the car by running a vacuum line to the engine. Power was via the cigarette lighter socket which is fused at 30 amps.

The red bottle is a bubble trap with 1:3 white vinegar and water solution. A drop of food coloring was added to make the level easier to see. The HHO then goes to the smaller orange bubbler that has baking soda solution, then on to the intake manifold.

SAFETY
If you are using NaOH (lye) or KOH (potassium hydroxide), keep some vinegar handy to neutralize it if it gets on your skin. Don't get it in your eyes! Hydrogen is extremely flammable (remember the Hindenburg?). Here, I am experimenting with a hydrogen/oxygen mixture. Guaranteed explosive.

Some of the bench testing before building the above cell.
With a variable DC power supply, two common nails were tried in a cup of water. Baking soda and NaOH were tried separately as electrolytes. The idea was to see if one electrode could be smaller than the other. It appeared that the negative terminal could be smaller, but not significantly so with any repeatability. It was decided to keep the electrodes the same size.

Materials tried
I tried iron, copper, nickel, aluminum, several stainless steel alloys, and titanium. All but stainless and titanium were ruled out as viable options.

Supply voltage
I wanted to test the claim that a number of small cells were required to keep the voltage at 2 volts. At 2 volts there was no gas production (at least with this cell). Production went up proportionally with the voltage. Six volts seemed best, but I am judging gas production by eye. I will stick to a single cell at least for the moment.

Prepping the car
The vacuum connections under under the hood were inspected. Some hoses needed to be replaced, including the PCV hose, and the gasket between the carburetor base and EGR valve.

Jan 31, 08
After the addition of 20% antifreeze, the electrolyte became brownish and cloudy in a short time in use. The SS strips were becoming pitted at an alarming rate.

This site has some data on the benefits of hydrogen: http://www.hydrogen-boost.com/September%202007.html
Very encouraging.

Feb 1, 08
The hose connections on the plastic jar lid are unreliable due to adhesive failure. I built a new cell from a Mason jar (1 quart) with brass tubing soldered to the lid for a solid hose connection. I suggest when soldering, do it with the lid on the jar so that the jar acts as a heat sink to prevent the rubber seal from getting too hot.

I was curious about the claims that temperature reduces efficiency with 12 volts/cell (regulated).
time amp temp °F   ~bubbles./minute bubbles/min/amp
6:25 8.9   88.3    140 bubbles 15.7
6:27 9.8   96.8      160+ 16.3+
6:32 10.1 100       180 17.8
6:37 10.9 110         190+ 17.3+

With a rise of only 22 degrees, the production went up 37%! Temperature is a major consideration.
Adjusted for current draw the output is10% higher than it would have been.

Update: I left the cells on the shop floor overnight so they reached a temperature of 40 degrees. Output was dismal until they warmed up. I really can't imagine an under the hood installation now.

At 5 volts regulated:
1.8 amps 105 deg. 61 bubbles/min (accurate)

This makes it look like 2 jars in series would give 122 bubbles @ 3.6 amps versus 140 bubbles @ 8.9 amps for a single jar.

Adjusted for the difference in voltage, two jars would give 144 bubbles at 3.6 amps. or something like 27% better efficiency at least with these plates and electrolyte.

Feb 2, 08
Two Mason jars in series, 12 volt regulated, 1/2 teaspoon of lye to 24 ounces deionized water.
1.76 amps 104 bubbles/min 59.1 bubbles/min/amp
(About 76% more bubbles for the same amperage as the previous test)

Ethylene glycol antifreeze
I dumped one jar out and poured half of the other one in it. Carefully measured 3 ounces of ethylene glycol into each along with 1/2 teaspoon more of lye. The (deionized) water level was brought back to the original 24 ounce mark on the jar. This means each cell now has 3/4 teaspoon total lye and 12% ethylene glycol.

2.20 amps 168 bubbles/min 76.4 bubbles/min/amp
This is 4.8 times the hydrogen production per amp of the original jar
I have full range litmus paper on order so the pH can be measured.

After about one hour, the electrolyte has become quite brown, and the gas production has fallen to 144 bubbles a minute with the amperage stabilized at 2.45 amps, for 58 bubbles/min/amp. This is still 3.64 times what the original jar produced. I feared the SS strips were being eaten away, so after 4 hours I removed them. The cell has the odor of your car radiator when you open it hot. I could see no deterioration of the plates. The solution still fluoresces under ultraviolet light as it does with fresh antifreeze.

I found this from Old Cars that explains the problem.
"Antifreeze also contains additives to help prevent “electrolysis,” which occurs when a conductive solution such as water, passes over the dissimilar metals that make up a cooling system."

" In conventional antifreeze, as the additives wear out, the pH of the water begins to turn acidic."

Testing with cold solution made me realize several things that need to be considered in a practical application.

1. When a warm cell cools, it will try to draw the vinegar solution into the NaOH cell. Several methods could be used to prevent this.
   a. Each cell could be vented with a check valve to atmosphere (photo at the top of this page)
    More plumbing would be required.
   b. A check valve added between the cells and the vinegar bottle.
This is the simplest, but not as fool proof as (a.). This failed the first time I tried it.
   c. A small orifice could be added to allow the air to enter the jar manifold.

2. I expect there will be a summer recipe, and a winter recipe for electrolyte solutions.

Feb 3, 08
Alcohol as antifreeze (this may be my winter recipe)
4 ounces denatured alcohol, 1 3/4 tsp NaOH, 20oz. deionized water
2.85 amps steady stream of bubbles.

I used denatured alcohol from Ace Hardware that claims it can be used as a fuel without residue. That is highly desirable. My experience with alcohol lamps for chemistry experiments left a gummy film that is difficult to remove. I sure don't want that in my engine.

Originally, I planned on either methyl or ethyl alcohol intended as a fuel tank additive, but all say they contain corrosion inhibitors to prevent electrolysis. The low amperage surprised me, but since the amount of gas is sufficient, I am not going to question it. After a few hours running, the solution was still crystal clear. Bubbles are for the first time a steady stream which makes bubble counting impossible. Time to mount everything in the car.

Feb 5, 08
Installation (again)
I find it ironic that such a large car can have so little room under the hood. For a moment I considered the trunk as a home for the hydrogen generator. The thought of accumulated hydrogen and oxygen gas in a closed area scared me. Besides, I want to be able to monitor the operation at least until I am sure the bugs are out. So it will remain on the transmission hump..

First try. The engine was started and allowed to run until it was idling fairly smoothly. It took maybe 30 seconds for the hydrogen to build up against the vacuum in the jar, then the idle speed increased noticeably. The vacuum went from 15 to 16.5 inches. But once the hydrogen was used, the engine slowed back down until the generator caught up and the cycle repeated. This indicated the need for greater hydrogen production.

Second try. This is the first road test. The road is not as flat as I thought. The plan was to use the cruise control and check the vacuum with the generator on, and again with it off. Not certain what happened, but the car did seem to run smoother with it on. At 2.6 amps I am sure it could use a LOT more hydrogen, but I am sure now that hydrogen is the way to go. I purchased 12 smaller mason jars (don't laugh, they are only 8oz.), SS strips, and assorted hardware to make first 6 jars and perhaps later 12. I know now I need to shut the generator off automatically. It got left plugged in twice so far. Old age setting in.

Feb 7, 08
The new jars are finished. A coat of spar varnish was applied to the inside of the lids as an added precaution against corrosion. Care was taken to make sure no varnish got on the rubber seals.

I tested one of the smaller jars with a saturated baking soda (NaHCO3) solution
3.3 volts 1 amp 100 bubbles/min (4 in series = 400 BPM)
4.4 volts 2 amps 150 bubbles/min (3 in series = 450 BPM)
After a few hours there was no deterioration of the plates, and the solution was still clear.

I tested the pH with some strips I purchased from THIS fellow. It read 10 and a fresh solution of baking soda read the expected 7. This is scary, because there was no indication from the current, gas production, or liquid color to indicate that anything had changed! I really don't want NaOH getting into my new engine.

All 6 jars in three banks, 2 jars, each now produce a calculated 2,000+ BPM at 3 amps, or ~670 BPM/amp. This was with the regulated 12 volt supply. In the car the voltage will be about 13.2 volts with the engine running. update The voltage across two cells reads 11 volts in the car with the engine running due to resistance of the long wires. I will use heavier, shorter wires in the final version.

Next step is to package it with the vacuum switch and three relays (one for each bank).

relay circuit

Note Part of this has been replaced with an automatic temperature control HERE.
Vacuum switch is set to close at just a few inches of vacuum.
This will allow power to the H2 generator only if the engine is running.

Feb 11, 08
I found that 6 cells wired as three banks of two give me a decent amount of hydrogen. They got a bit warmer than I wanted, but I had planned an old computer fan for the console box to help dissipate any gasses that may accumulate. The vacuum switch will close 3 relays, each with an LED power indicator.

Feb 14, 08
First real road test of the 6 pack H2 gen. I am delighted! Just touch of the throttle at 30 MPH.
30 MPH 17 inches Hg
45 MPH 15 inches Hg

I am still having some problems with the bubble jar, because the bubbles are now big enough to let water into the lines due to splash when they pop. In the works is a larger jar, blue aquarium air bubbler stone to make the bubbles smaller, and a pair of needle valves. The latter was $4.37 at Walmart.

Feb 16, 08

Cover off.

The top is covered with some scrap Naugahyde. The colors match the car interior.

Left front: are the three relays super glued to the vacuum switch. Each is rated at 15 amps, but carry only about 4 amps here, so they should last a long time.

Center: The aquarium needle valves

Center bottom: Old 12 volt computer fan to disperse any gas buildup and help keep the jar temperature constant.

Right front: Anti-flashback jar. Plastic, with an aquarium bubbler. I consider this jar an absolute necessity. The bubbler solved the problem of water being drawn into the engine. I may add a warning light for low water level.

Note: The cigarette lighter cord is not heavy enough.

I realize now that there are two factors that cause the water in the bubbler to get drawn back into the gen jars...

1. When you first shut the engine off, there is still a vacuum from the manifold in the gen jars. The lower the electrolyte level, the more vacuum.

2. When the jars cool, the H2 and O2 in them contract and draw water in.

The three LEDs are installed on the console lid now. They are a little bright at night, but I wanted them bright enough to see in the daytime.

Tuning I have not really seen this addressed by others much except for retarding the ignition timing. Since the whole idea of this is to make the cylinder charge burn more completely, doesn't it make sense that the engine needs less fuel? Since my huge engine is barely above idle when cruising in town (and town is where the FE is the worst), the idle mixture is the first priority. With the engine at operating temperature, the two idle mixture screws are turned inward until the engine slows slightly, then turned back a little.

Feb 21, 08 Another road test only with hydrogen entering the engine through the air horn instead of the manifold. This eliminated surges of hydrogen when the manifold vacuum changes abruptly. Very smooth, and made final tweaks to the timing and carburetor easier. The needle valve has been removed since it is no longer needed.

Feb 23, 08

Expansion trap

This solved the problem of contraction of the gas in the cells drawing solution after the engine is off, and ambient temperature drops. As the gas cools in the cells, solution in the backflash trap (center) are pulled into the backdraw trap (right). The next time the cells produce hydrogen, the solution it pushed back where it belongs. At no time can solution contaminate the cells, and backflash suppression is maintained. So far it is working just fine.

Blue Solution consists of water with ~1 teaspoon of white vinegar. This is intended to neutralize any caustic properties of the gas mixture. In winter it will have alcohol added.

Green (left) A plastic pill bottle packed with cotton that was saturated with a baking soda (NaHCO3) solution and allowed the dry. Hose nipples were glued in the end holes. This is intended to neutralize any acid in the gas.

Feb. 24, 08 Finished bench testing NaOH, NaHCO3, and Rochelle salt as electrolytes in a cell with titanium used as the negative plate. Not a trace of sludge in the bottom of the jar and no deterioration of either plate. May replace the negative SS plate in the six jars with titanium strip.

Feb. 25, 08 Last night I swapped the negative SS plates on two cells with titanium. With no other changes, the amperage doubled. I feel this is good, but will know more later. It is sure to reduce the cells sludge buildup. The increased H2 helped the vacuum. This is very encouraging, but I will test it in the car before I convert the other four cells.

I need to rethink retardation of the timing that most builders of HHO boosters recommend. With retarded timing, the vacuum drops on my car to the point where idle vacuum is what 30 MPH used to be. NOT good! Here is the plan...

Change things back to factory settings. With the HHO off, adjust carburetion as necessary. With the engine warm, test the HHO under as many road variables as I can. If it seems to ping, I can easily shut off one bank of cells. I will shut off the HHO before shutting the engine off to clear out any H2. If it turns out that H2 production needs to be tailored to the engine load, I have a number of options using switches etc. If nothing else, I will have learned more.

Another option for engine ping is water injection. Whew! This getting complicated. But not yet overwhelming.

I set the timing to factory but left the carb as I had adjusted it for HHO..
30 MPH = 17.5 inches Hg (was 17)
45 MPH = 16 inches Hg (was 15)
55 MPH = 15 inches HG 15 inches Hg was idle at the start of this project (see Feb 14th)

Temperature gage is now in the middle instead of up near max. I rewarded the car with an oil change (synthetic of course).

Feb. 28, 08 Last night the remaining 4 cells were converted to titanium negative plates. Added a splash shield to the backflash bubbler to help prevent mist from entering the supply hose. It is simply a scrap of stiff plastic package material with a X cut off center so it can be pressed over the brass tube inside the jar. The one on the right presses over the electrodes.

Left: disc for bubbler. Right: disc for gen jars. The outlet hole it protects is shown as dotted.

March 2, 08 The car has been started many times now inside the shop. There are no black sooty spots on the concrete floor behind the car from the exhaust. This is the first time I have seen this in the 24 years I have owned the car!

I have KOH (potassium hydroxide) on order. Many claim this is the best electrolyte. I wanted to give it a try, not for the sake of H2 production, but perhaps it will be better at various temperatures.

The syringe above was drilled out at the tip for a snug fit for a piece of wire insulation. The gen jars will have an added short piece of brass tubing soldered on that will be capped with a piece of vacuum hose that has a screw glued in the end to seal it. This may not be the best way to refill the jars, but at least it is no more difficult than adding water to a car battery.

March 4, 08 Looking back over my data, I believe a temperature controlled HHO system is the only way to go. Remembering all the fuel efficiency reports by others makes me wonder at what temperature they were conducted, and how they managed to keep a relatively constant temperature since they were all under the hood installations.

March 7, 08 The potassium hydroxide arrived and a relatively weak solution was tried in two cells in series at 12 volts regulated. In just a few minutes the electrolyte became an alarming brown color. I emptied the cells and was horrified to find dark brown bubble like protrusions from the back side of the SS plate. They rubbed off fairly easily under running water to reveal pitting in the center of the plate. Seems to be no end to the mysteries.

Mist shields (see above) were added to the rest of the cells.
Mist shield for generator

I found that bending at the dashed lines slightly over the edge of a wood block with my thumbnail makes sliding the discs onto the metal strips much easier.

I am back to sodium hydroxide but with a slightly stronger solution and have the best gas production yet at about 7 amps. The engine is running very smoothly, but does not sound right (ping I believe). Normally ping is a very bad thing, but in this case it indicates that enough hydrogen is being injected to make a difference. I will have to retard the ignition a few more degrees to compensate.

March 13, 08 Another page has been started devoted to electronic control HERE

March 16, 08 I have a digital LED panel ammeter on order to mount in the generator "control panel". The car seems to want more hydrogen and it does have the room since it is a large car, so I am going for a 12 cell generator. It makes more sense to do it now before making the final electronic board. These are the changes that will be made...

A larger output tube on each cell to make filling easier.
Insulation in the generator box.
A second fan.
Indicator LEDs for the heater and fans in addition to the exiting cell power indicator.
The 6 cells in the center will be heated, but the 6 new ones will not.
Heavier power cord.
There will be a manual heater switch in case I want to preheat the cells in cold weather.