Almost Building An Engine From Hardware Store Parts


You can build a surprising amount of stuff from parts you can pick up at a hardware store. Sometimes, though, getting a project built from sections of pipe is very, very difficult. That’s the case with [Lou]’s hardware store engine: despite an inordinate amount of cleverness, he just can’t seem to get an engine made from pipe fitting to work and is now asking for some ideas from other ingenious makers.

The engine uses regular oxygen and propane tanks you can pick up at Home Depot with torch heads soldered onto half inch pipe. The fuel and oxygen are mixed in a T fitting until a grill igniter sets the gas mixture ablaze pushing a cylinder down the length of a copper pipe. The cylinder is attached to an aluminum flywheel that also controls the opening and closing of the oxygen and propane valves as well as switching the grill igniter on and off.

Right now, [Lou] can get the engine running, but only for one stroke of the cylinder. He’s having a bit of a problem turning this into a working motor. If you’ve got any idea on how to make [Lou]’s engine work, drop a line in the comments. We’ll throw our two cents in and say he needs a valve on the exhaust, but other suggestions are always welcome.

82 thoughts on “Almost Building An Engine From Hardware Store Parts

      1. Yes and no, it’s the low tolerances approach. When you don’t have the highest precision you can make up for it a little with overkill.

        Yes, Americans used the overkill approach (350 ci engines making only 220 HP) but it was a precision problem they were compensating for, which is likely a big part of the problem here.

        A 5 HP Briggs and Stratton is over 200cc where a 125cc dirtbike makes at least double that horsepower. The Briggs has sloppy tolerances and uses brass bushings and a plastic camshaft. It’s cheap, sloppy overkill, but very effective.

        1. The topic needs some definition.

          Most of those engines aren’t overkill.

          For example Truck V8s (also used in sports cars) had 170-200hp but those were output sometimes down to as low as 3000rpm with Torque in Ft-Lbs up to 50% more than the HP figure,when those engines were allowed to revv higher using different carburetor setup,valves,etc,they were twice as powerful but maybe at twice the RPM,let’s say 200hp at 3000rpm or 400hp at 6000rpm just like a modern Corvette V8 of that displacement.

          The 5HP Briggs doesn’t have anything to do with a Dirt Bike engine as the Dirt Bike will have a Two Stroke that wont sweat revving nearly as much as a 4 Stroke and have a higher Horsepower output from less displacement but less Torque which can get you better Torque through gearing,while the Briggs should have the highest output possible from slow moving parts (higher displacement,more HP at lower RPM) and more torque.

          1. On the other hand, your average Briggs is pretty old fashioned. Flat head, typically lower compression, no timing advance/retard, and a very basic carburetor. Good for idling and running at a set speed under load.

            Motorcycle engines, on the other hand, are usually overhead valve, variable ignition timing, better tuned carbs (with a variety of mixture compensations methods), etc. A 125cc 4 stroke (to compare apples to apples) can still produce more horsepower over a wide torque band.

            Then again, the B&S can also put up with maintenance that would put most motorcycles in for a rebuild – which is really the key to its success. A long-running, low maintenance engine.

  1. From my point of view (note: i am no mechanic or engineer, i just know – i think, i do – a lot about gasoline engines, as in cars), there are some problems.

    One is friction. I didn’t have to watch/listen closely to notice how all the parts are scratching against each other.

    The other one, he already pointed it out, is the mixture. If he adds more oxygen manually, the system fires fine(tm). Obviously afterwards it doesn’t get enough oxygen again, so it stops firing. It should help opening the Oxygen valve shortly before opening the propane and closing it a time after the propane is closed.

    Next comes the exhaust. He said there is a brass plug that fits perfectly inside the copper cylinder. So, essentially there is no exhaust at all. Nevertheless one can see the exhaust gas coming out of the cylinder, so the brass plug does not fit perfectly, but i would say it is too snug. Despite of this, if i would build an engine like that i would add a dedicated valve for exhaust.

    The flywheel design is a bit faulty. Because of the rod being attached to the outside of it, its center of mass is heavily off. One could explain it this way: the more you move the asymmetry to the outside, the more off is the center of mass. This is bad as it is, but then you start applying power in form of linear movement (the rod only can go out or in) which makes it even worse. Take a look at a crankshaft in gasoline engines. You will see the crankpin journals, where you attach the crankpin and the piston, are off the center of the crankshaft, but then take a look at the counterweights. They are at the exact opposite end of the crankshaft as the crankpin journals. The counterweights are used to balance out the weight difference caused by the crankshafts journals being off center and the linear motion applied by the pistons.
    You can look at this image here, then you will understand what i am talking about:

      1. If you cut a hole in the cylinder wall near the bottom of the piston stroke the piston will act as a valve and release the exhaust as it passes by the hole. Exactly where to make the hole might be tricky. Once the piston passes by the hole you won’t make any more power, since the expanding gasses will have a way to escape.

        If the port opens too early, you won’t make enough power to keep running, to late and you won’t exhaust enough gasses for the piston to move back to top dead center.

        I would start with a small hole, very low down, then add small holes higher up and/or widen the holes a little at a time.

        Also, it will probably be pretty darn loud with an “open exhaust”.

        1. This was my thought exactly. The holes need to be positioned near the top of the power stroke in such a way the flywheel still has enough momentum to re compress the piston. In other words, the holes need to be -x degrees before TDC. Just a small amount.

    1. Steam engine trains use a flywheel / drive wheel design that compensates for the mass offset by adding a counter weight to the side opposite where the piston arm connects to the wheel. Something like this would be effective in countering the abrupt force of the down stroke and would require a larger diameter flywheel and longer and wider stroke and piston sleeve that will allow for this. Where this may not work easily is that in the case of a steam powered train the engine is self lubricating by way of the steam which may be the better option in this case for the tinkerer to go after as an oxy propane combustion source may end up being to much for the store bought pipe under a sustained run after a time and rupture from repeated periods of heating and cooling. The best thing to do is go with material that is rated for a wide range of heating and cooling for safeties sake and the propane and oxy cylinders should be located much further away from the combustion source and secured.

  2. It looks like an attempt at a 2 stroke design. There needs to be an exhaust port cut into the side of the cylinder. Also there should be an expansion chamber to help vent the gasses. As for the flywheel, go back to the hardware store and get some heavy solid rubber tires (or maybe a tire/rim for a trailer). Lawnmowers use the blade as a flywheel, but I wouldn’t recommend using one in this case ;-)

  3. gotta have exhaust. And if the propane O2 mixture is going in without any timing involved, then all you have is one blast and the gas pressure keeping the piston in the extended position.

    Already mentioned is that the flywheel needs to have a certain amount of momentum to keep it going. When you start your pull string yard products, you’re providing the initial momentum to start the sequence. If that flywheel isn’t large enough to sustain momentum, it won’t work.

  4. Here I was hoping for a Hardware Store STEAM engine…
    That said, he should maybe move away from oxygen and use an air intake, as well as an exhaust. I just don’t think this machine will be working in time for the Zombie Apocalypse.

  5. Exhaust is one issue…could just drill a hole in the “cylinder” near the end of the piston’s travel, somewhat like a steam engine design. Not perfect, but with some exhaust being able to escape, it might allow the piston to travel back more easily after the power stroke. Also, the remaining uncombusted fuel in the exhaust right now could be leaving a mixture that is too rich in the combustion chamber, which could be why it will fire again if you manually add more oxygen. Again, this may be solved with exhaust by getting rid of some of the leftover products of combustion from the previous cycle.

    If you do get successive combustion cycles, the copper pipe “cylinder” is going to heat up and expand, meaning a loss of power the longer the engine runs

    1. That is no 4-Stroke design, my fellow friend.

      Intake (piston moves down and sucks in air or mixture through open valves, depending on engine design; directly injected/airpipe injected),
      compress (piston moves up while valves closed, compressing the air or mixture),
      fire (injection of fuel on directly injected engines and generation of a spark with the sparkplug which makes the piston move down; note: diesel engines don’t have sparkplugs and mostly are Directly injected),
      exhaust (exhaust valves open and piston moves up, pushing out the exhaust gas)
      4 different steps for one logical cycle, but actually two revelations of the camshaft.

      The build here is far away from 4-stroke.
      There is no compression stage and also no dedicated exhaust stage.
      It is more comparable to a directly injected 2-stroke.

  6. everthing said here seems to be good advice.
    i think i can add some ideas that haven’t been mentioned yet.

    first, there might not be enough time to refill the combustion chamber.
    the wheel pushes the valves open for just fraction of the time it’s opened when
    he does it manually.

    and second, as mentioned the exhaust gases are pushing the rod even after the ignition, but additionally, the valves open at the same time the piston slides back in the tube.
    at that moment, the piston has to work against the full pressure of both fuel tanks, since they are filling the chamber while the piston is compressing the gas.

    i think he should do some math before he tries to fix it.
    at the moment, it looks more like trying until it works.
    while that might have a slim chance of success somtime it would be faster
    to look at some numbers first.

    and maybe it’s time to think about an arduino as an ecu, for ignition / injection timing ;-)

  7. My thought is that there isn’t enough time during the intake stroke to fill the cylinder with fuel. When performed manually, the valves are opened for several seconds – which is why it works. Upon ignition, the flywheel appears to have enough intertia to complete several revolutions, but with the intake stroke being so quick the cylinder never fills enough to effect a combustion.

  8. This is unconventional in that the intake and compression stroke are combined. The only reason this is working at all is because the fuel and air supplies are pressurized. As others have stated, the exhaust is the real issue. There is no mention in the video about any sort of exhaust valve and unless this is addressed, this will never work.

  9. Definitely need an exhaust port and/or valving. And it will be a 2 stroke as long as there is no 2:1 timing arrangement to drive the valving as in a cammed engine. (There would be some exceptions to this such as the atkinson cycle engine.)

    Does the hardware store where you got all this stuff sell lawnmowers? There’s your 100% hardware store engine. Perhaps even cheaper and lower tech than the pipe fitting engine.

  10. you could try basic mechanical engineering and correct the timing..

    I’d do one tank and a transmission. Transmission increases torque or rpm while lowering fuel consumption. Then again it’s a propane engine and pointless to make…

  11. I agree with the comments above, with a few additional comments.

    As others have said, this is essentially an attempt at a two-stoke engine. For this to work, he needs to open exhaust well before the piston reaches the bottom of it’s stroke (BDC) to allow time for all the combustion products to exhaust. Then, he needs to open the intake valves around BDC, and ideally close exhaust shortly after. This will waste some amount of unburned fuel, but it’s kind of inherent to the two stroke design. This sequence is more representative of how most two-strokes operate, but will require a more complicated valve armature.

    Secondly, he needs to install either a gas regulator or a needle valve on the propane side to regulate the fuel to air ratio. I don’t know what pressure the oxygen and propane cylinders are at, but if they are at the same pressure, then his mixture is way too rich. This is why he can add oxygen after a combustion and get another combustion; there’s plenty of unburned propane still in the chamber. Per wikipedia, the ideal combustion ratio for propane and o2 is around 1:24, so the oxygen pressure needs to be 24 times higher than the propane pressure!

    Also, a bigger flywheel would help a ton.

  12. Let’s not forget the 3 key components to a combustion engine: air, fuel, spark.

    I think other posters are correct: the problem here is air (in this case, the exhaust.) I think of the 2-stroke head I recently replaced on an old dirtbike, and the cylinder had some very strategically placed ports on either side, in order to allow the piston to pull air in for the air/gas mix (intake), and push air out the exhaust port.

    I think it will be challenging to determine positioning of your exhaust port(s), and the size, since you have to compensate a little for PUSHING compressed gas into the cylinder (where most 2-stroke engines typically PULL air/gas mix.)

    Fun project, and I like the community involvement on this one.

  13. I would like to point out, that in this designe, the difference between pipe Engine and pipe Bomb is whether or not the piston jams.

    Please exercise the appropriate caution.

    I noticed several posters suggested using a flywheel. In this single cylinder designe using a spring to return the piston might be easier.

    1. Then it would probably work twice. Using a spring means overcoming the full resistance of the spring on EACH stroke. Using a flywheel the resistance gets less as momentum builds over a few strokes. That’s why steam engine trains don’t have big old springs on their pistons.

  14. It need an exhaust at full stoke, a hole on the cylinder at the right place would work.
    Also start refilling sooner at full strike too even with some overalp, try to stop filling before you are far into compression.

    Since filling seems slow, reduce your target RPM by using a bigger flywheel. 60 – 100RPM would be reasonable.

  15. Diagnosing via video, I would suggest perhaps not filling quite a much propane for the oxygen provided. If you can reignite the mixture w/o the addition of more propane, the engine is running far too rich. The other commenters have a great idea w.r.t. the simple exhaust port as well. Should be near the end of the stroke and for reference, just adding a simple tube will help scavenge the exhaust if you can get continual combustion. It just conserves the momentum of the gas in the exhaust and will draw more combustion product into open air. Best of luck.

    1. I second this idea. Perhaps add two more pistons, although that adds a bit of complexity. With only two pistons, if timing gets off, the second piston might start pushing the flywheel backward. With three pistons, you will have to get your timing WAY off to have that problem.

  16. Possible issues to investigate:
    – An exhaust port. Probably timed to close just after opening the fuel air valves. Remove any back pressure that might keep fresh gasses from filling the chamber
    – Reduce as much friction as possible for all moving parts.
    – flow from tanks filling chamber before vales may not fill fast enough.

  17. Thanks to everyone that made suggestions so far. Here is what you are saying, and here are my plans so far. I am working on it more this weekend.

    “Add an exhaust port” Great idea. Yes I plan to drill an exhaust port at the bottom of the stroke, when the piston is farthest from the gas inlet and spark

    “Use a bigger flywheel” I see your point. The wheel did not have enough momentum to come around and open the valves. Instead, I found some MUCH easier-to-open valves in the air compressor aisle. I think the current wheel will be able to open those, just fine.

    “Your timing is off” I plan to use a cam system to change where, in the cycle, I let the gas in.

    See Pictures of the planned changes at

  18. Wouldn’t it make more sense to have an opposing engine to counter the flywheels response to the “piston” counter the fuel induction by increasing the gas inlet, run with an audino for timing gas release to 2 actuators and firing the ignition on a return rather than against? An exhaust would have to be pressured with the system though so having a release at the end of the stroke. This would effectively work out as straight 2 :P

  19. What about adding another piston and cylinder, using this drive to push the exhaust gasses out, maybe not use a hole at the end of the cylinder you may loose alot of your power can’t you use another pulley like your fuel simular to how a head works with tapits etc, all a head is doing on an engine is controlling the gas movement u need to suck fuel in and puff the exhast gasses out, SUCK SQUEEZ PUFF BANG is all an engine does. Also creating a cam will fix timing for gasses going in and out you need gass to exit just after and when cylinder or returning however at 180 degree the reverse will happen without flywheel and u loose all energy just before critical point.

  20. A faster ignighter may also help once it’s running. The grill igniter’s turn-on delay and slow sparking could delay ignition significantly, and limit the maximum speed. Looks like it’s one of those epoxy-filled single-AAA ones, try giving it double the voltage and it should go 4 times as fast. It shouldn’t burn out or anything – theyre pretty rugged. More oxygen could also help.

  21. perhaps a nylon or chopping board hdpe bush at the flywheel end of the cylinder to keep the piston rod centred and reduce friction and clanging.
    definitely needs exhaust ports at the flywheel end of the cylinder a la uniflow steam engine ports.
    perhaps a lubricator activated by the valve rod, putting a drop of oil into one if the exhaust ports each cycle would be good.
    a hit and miss engine governor would be interesting to implement to control rpm and fuel inflow.
    might be easier to buy a pressure cooker, copper pipe, and a camp stove and run it as a uniflow steam engine, with all the excitement of potential boiler explosions!

  22. Thanks to everyone that made suggestions so far. Here is what you are saying, and here are my plans so far. I am working on it more this weekend. I will post pictures as I go, at -Lou

    Add Exhaust Port 25 A LOT of you pointed out that there was no exhaust port. I was hoping that enough would slip by the piston, but I think you are right. I will drill a hole in the cylinder at the very bottom of the power stroke.

    Make Flywheel Bigger 13 Many of you noticed that the flywheel spun partially around and then stopped, when it was trying to open the valves. Very good idea, but I am going to try to solve the problem with new valves that are at least 10 times easier to turn. I think the current flywheel will be able to handle them. If not, I will try a bigger one.

    Change Fuel Mixture 12 The valves were actually set on the first engine to have slightly more oxygen than propane, but several of you said it needs to be 10-1, and I will change it to that. Thanks.

    Add more gasses 10 A few of you thought that there were not enough gasses being let into the cylinder, with each cycle. I am worried about this too. I am changing to a cam system that will hold the valves open for more of the cycle, but worry that the torch heads will not allow enough flow to support continous explosions.

    Change Timing 5 Some noticed that the timing was not optimal. The spark should happen at top dead center and the fuel should be let in right after exhaust. I added a cam system to allow optimum timing.

    Use air, not oxygen 3 Air is about 20% oxygen, and that is the only part of air that combustion uses. An oxygen tank was only $8. An air tank would have been $50, and I would have had to pump it up. There are those little spray air cans, but no way to hook pipe to them.

    This is a 2-cycle engine 2 Yes, I thought this would be easier than a 4-cycle, to build

    Replace grill igniter with… 2 A few are concerned that the grill igniter will not be able to spark fast enough. Hmm. It is pretty fast. I will leave it for now and fix other stuff first. I may have to change it later

    Add Cooling 2 A couple people thought that when this engine does start running, it will heat up and leak badly or seize up. I have plans to add a 1/4″ copper pipe water cooling system, but wanted to get it running at all, first :)

    Stabilize Connecting Rod 1 Some of you noticed that the connecting rod was pretty loose in the cylinder. I will add a 1″ to 1/2″ copper pipe adapter to guide the 3/8″ rod.

    Add a kill switch 1 A few of you thought ahead to when this engine actually starts running and worry that I could not stop it, if it started going too fast and parts started flying off! Good point. I actually have one. I can turn the battery off to the spark.

    Make Cylinder Stronger 1 One person is worried the cylinder could explode. So far, so good. I hope not !!
    Publish Email 1 One person wanted my email address to send me a PDF of a CAD drawing. It is

    Use Gasoline 1 One person suggested gasoline to replace the propane. I need pressurized gas, because I am forcing in fuel while the piston is compressing. Gasoline needs to be pulled in with an intake stroke, but I don’t have that in this 2-cycle engine

    Use Check Valves 1 On person suggested using check valves and I thought this was a great idea. In fact, my first prototype used check valves, but since the gasses kept coming in, the first explosion just lit the gases at the source and I had an internal torch that never went out.

    Add piston rings 1 One person suggested piston rings. This is pretty tough with hardware store parts. I hope I don’t need this!!

    Help Wrap Gifts!! 1 Oops. That was my wife.

  23. You don’t want your ignition to be at the piston’s top dead center of travel, you want it before. This ensures you reach optimal cylinder pressure AT top dead center. If you ignite at TDC, you will be burning all your fuel as the piston is already traveling to bottom of its stroke, wasting a ton of power.

    Also, something to consider on the flow rates of your gasses, if you have a single gas flowing into your manifold, it has the pressure of one cylinder working for it. If you have two gasses pressurized in canisters going into a manifold the higher pressure canister will work against the lower pressure canister. I didn’t see the video, but it sounds like you need some way of overcoming that, like a mixture chamber, and a way to accurately calculate your fuel flow.

    1. Very nice, but there are several machined parts on that one. I am trying to come up with a design that anyone can build, with parts from the hardware store. Once I figure it out, I will post complete plans, so others can do it too. Thanks for sharing that one. It gives me hope that this is possible.

  24. How has noone commented on this being so incredibly dangerous? Are there backflow preventers on the canisters? I see the potential here to fill the propane with O2 or vice versa, creating an explosive mixture in the canister. Just too dangerous to mix an oxidizer with a fuel inside a closed system like this. I’d strongly suggest you redesign this to run on air and forget the O2.

    1. The Propane and Oxygen are really never in a closed system. They are released seperately into the cylinder, which has very little compression. There is quite a bit of room in the cylinder, above the piston, and I am just igniting the mixture at roughly atmospheric pressure, much like a potato gun. In fact, this is what gave me the idea for the engine. Thanks for the concern, but I think it is safe.

      I would actually like to use plain air, but I have no way of forcing it into the cylinder, and I need that, since there is no intake stroke in this 2 cycle engine.

      1. I think you could make a variation on how a diesel pile driver works and skip the oxygen

        Around mid stroke make a combined exhaust/air intake port. as the piston passes the port mid stroke you’ll first exhaust spent gasses but as the piston continues down it will suck in air. Use the valve to inject just the right amount of propane after the port is closed

        1. Ok, I have been thinking about replacing oxygen with air, not for safety, but to get more flow. I need much more oxygen than propane, and I am not sure that little tank can keep up, at speed. I may use a compressed air tank. They are $50 versus $8 for oxygen, so that I why I started the way I did, but I think I need to change. With compressed air, I should be able to clear the whole cylinder and not need the side port. I worry that a side port would let some of the explosion power escape. I would prefer an air valve that shuts during the explosion.

          I am half tempted to switch to a 4-cycle configuration, so I can use the piston to pull in a whole cylinder of fresh gas and oxygen, but this would really complicate the design and surely break my “Only parts from the hardware store” goal. I want to end up with a design that any somewhat-handy dad can build with his kid, without a machine shop.

  25. Lou: Could it be that your Mixture is not lean enought? I am not that good in NG-engines (and never seen a NG-twostroke) but your ignition problem in the second video could be from not enough oxygen.

    Secondly you could try to couple the ignition to the flywheel. Get some Hall effect sensors a Magnet and maybe a Mosfet to do your ignition. If your Flywheel is stainless steel you can move the ignition point around very easely.

    Manually turning the flywheel a few times would then start the cycle automaticly.

    1. Yes. I am pretty sure my mixture is not lean enough, but I am getting closer. Notice the oxygen valve opens much more than the propane valve, when the cam comes around.

      The first prototype had the ignition fire right after the inlet valves closed. I will do the same with this one, which will put it at top dead center, when the piston is furthest from the flywheel. I just want to get it firing again, at all! I think I may switch to compressed air, if the little oxygen tank can’t give enough volume. We’ll see.

  26. Can’t tell for sure from pictures, but for Otto cycle engine you may need to increase compression ratio (by making compressed volumen at top dead center) smaller, and increase the piston size. The ratio of piston size (i.e., force exerted) to seal friction is important, so you may want to try that first. Too much power is in friction loss.

  27. 1. you want to be careful there you may have a bomb if too much fuel builds up and the neighbors and even the law may not like it.

    2. there may be a timing issue with the ignition switch.

    3. smoke coming out the pipe means you may be burning your seal

  28. What’s important is the stoichiometric ratio of propane to oxygen. Finding that does not appear to be easy, I could only find lists of stoichiometric ratios for AIR and various fuels.

    An absolutely ideal combustion of propane and oxygen leaves little behind other than some water and carbon dioxide, but you won’t be getting that with this engine.

    What you could try is use a barbecue grill regulator on the propane and just fully open the oxygen cylinder valve. That way you can regulate the fuel down to the right ratio for the available oxygen flow.

    Another way to admit more oxygen VS propane would be to use two cams, one for each, with a longer duration for the oxygen.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.