Introduction: Build a Better Stirling Engine

Following my experiments with the first LTD Stirling, I decided to try and make a better one. The main problems with the first had been the glued joint leaking air, and excessive requirements for machining. This one would have soldered joints to prevent the leaks and would have far less machining and could be made with just a drill press..

The main difference between this one and the first one I made was the diffuser cylinder. This one would use a longer, thinner cylinder with an aluminium piston. I also decided to use metal where possible on this one and have roller bearings. The diffuser cylinder would be bolted on allowing strip down of the machine and a rubber gasket to hold the seal.

Step 1: Materials

For this engine I used the following materials:

1 length of 22mm (1") copper pipe
1 length of 15mm (1/2") copper pipe
1 22mm copper end cap (not shown)
3 aluminium heatsinks from an old TV (scrapyard)
1 piece of resin stock
Paxolin copper clad PCB board
HDD actuating head arm with bearings and shaft
2.5mm brass tube
1.5mm brass rod (sliding fit)
One marker pen with an aluminium body (not shown)
Part of a plastic 3" pipe or similar (not shown)
The brass centre boss from a defunct CD drive
A short length of 10mm aluminium bar from a scrap printer
A short length of aluminium channel (not shown)
Various nuts and bolts rescued from scrap items (not shown)
A small piece of high density rubber (a piece of bicycle inner-tube)
4 15mm brass disks (not shown)

The reason for the not shown items is ..... at this point I haven't acquired them yet!

Step 2: The Pistons and Cylinders

The power cylinder on this engine is made from a piece of 15mm copper pipe, 1/2" will do if that's what you have. I cut a 40mm section using a cheap pipe cutter, then cleaned the ends up with a Dremel to remove any burr and lip at the ends. The bore was polished using a small wire brush in the Dremel, followed by wire wool and very fine grit wet and dry rolled into a tube.
I then cast the tube using the resin process in my other instructable.
Just a note here. I tried both the resin and some epoxy putty cast internally as per other 'ibles on the site. I found that in both cases the shrinkage (around 5%) made the piston too loose. Maybe it's the brand I was using.
The resin blank was then sanded down in the drill press and polished for a sliding fit. This gave a piston 13mm in diameter. I cut it down to 13mm in length then cross drilled 2mm and drilled it out 6mm at one end for the con-rod connection.

I decided to simplify the con-rods for this engine. I went down to the model shop and bought some more 2.5mm tube and some 1.5mm rod. These two can be slid into each other. The piston connecting rod is 2mm steel from an old printer. To make the connectors I simply crushed 4mm of the tube flat in the vice, then cut it off at 10mm. Drilled the flat 2mm and slid it on to the piston connecting rod.

The diffuser cylinder is a length of 22mm copper pipe (1" should do). The pipe needs to be cut to length as depending on the length of the piston and the stroke of the crankshaft.:

The piston must have a gap all around it in the cylinder, the idea is that as the piston moves, air flows around it from one end of the cylinder to the other heating and cooling as it goes. There should be a gap at the sides of at least a couple of millimetres. I wanted to use a metal piston so that I could use a tea light candle to run the engine.

I found an aluminium bodied marker pen that was around 16mm across. This left a gap of about 2mm either side  when slid into the tube. Ideal. I cut the pen in half and removed the innerds which were discarded. The plastic end was broken off and cleaned with the Dremel to leave a nice round hole. Then I worked out the length of the required piston. A guess but I went for 65mm. To make the piston I cut the rear portion down to give the overall dimension I wanted, then I slid a short section of 15mm plastic plumbing tube into the end a glued it using super-glue gel. The other piece of the pen was then slid onto the other end of the plumb fitting and secured with super-glue gel. This gave a lightweight air-tight piston. The open end was fitted with an aluminium plug pre-drilled for the con-rod. That completed the diffuser piston.

Now I wanted to be able to strip and re-assemble this engine so as few glued joints as possible. You could use a plain solder end cap on one end of your cylinder, but because of the soldered bolts I didn't want to disturb with more heating I used a compression fitting. This will become the hot end. Next I sourced 4 brass motherboard mounting bolts from a scrap yard PC. These come in a number of thread sizes, most common sizes are M3 fine thread or M3 course thread but there are some imperial ones around. the trick is to find 4 the same (which should be the case if they came from the same PC). Find 4 bolts that fit the threads. These need to be at least 15mm long. If they are metric you should have no trouble finding bolts. If they are imperial suck it and see.
Cut off the threaded part of the bolt, leaving just the brass 'nut' which should be about 6mm long. We are going to solder these onto the cold end of our cylinder and then use them to bolt the cylinder to the chassis. If you manage to solder them at exactly 90 degrees each then you don't need to mark the cylinder. I didn't trust my accuracy so I scribed a top mark so that each time the cylinder is bolted up it will go back in the same place allowing for misalignment of the soldered stand-offs.

I made a simple wooden jig to hold everything in place. i tinned the bolts and one end of the pipe and then used a micro pen torch to solder the bolts on. I tested bolting it all up before going back to finish the chassis.

With the exception of the con-rods this completes the cylinders and pistons.

Step 3: Main Chassis

Assembling the main chassis.

This engine differs from the last in that the cool plate will house a separate diffuser and power cylinder one either side. Both will be sealed in their own right and linked by some tubing.

The chassis is made up of a sandwich of four plates. Two heat-sinks (they had power transistors on them) from a TV I found in the scrap yard. They have pre-drilled holes in them from the transistor mounts and are identical. This means that I can reverse and screw them together to form a single thick plate with the cooling fins on.

Each cylinder is formed from the copper tubing, the power cylinder is soldered to a Paxolin board that is bolted to the two heat-sinks. The diffuser cylinder will be bolted to a second Paxolin board and bolted to the chassis via a gasket. This should enable me to seal the two cylinders and then simply link them with some tubing. (I hope).

I cleaned up the two heat-sinks with the Dremel to ensure good mechanical contact between them. I cut and marked out the two Paxolin boards, then drilled the power cylinder board to bolt to the heat-sink.

The diffuser cylinder board was marked up and drilled for the bearing and the feeder tube, then tacked on with super-glue. The four mounting bolt holes were drilled through and then the two feeder tubes were soldered into place. The power cylinder was soldered on and the nylon bearing for the diffuser piston glued in.

Finally the feeder tube was connected up. That completes the main chassis assembly

Step 4: The Flywheel and Boss

I wanted to make the engine look nice, so I decided to try and make an elegant flywheel.

I had found an old brass centre from a defunct CD or DVD player, then I downloaded a protractor template from the net and marked lines at 0, 120 and 240 degrees to give three equal spokes. I still had the top ring of a food container I had cut down for my first Stirling, so I used that as the flywheel outer ring. You could cut a ring from 3" plastic drainpipe instead.

I placed the brass boss onto the protractor with double sided tape and then measured and cut 3 lengths of 1.5mm brass rod from my stock. These were then supported on some levelling materials and soldered to the centre boss. The outer ring was laid onto the protractor and marked then drilled 1.5mm to accept the spokes. Because the ring is slightly flexible it was easy to deform it enough to force the spokes through. Once it was all back in shape and trued up on the protractor, the spokes were super-glued into the ring.

A piece of 10mm aluminium was cut to size and drilled ready to accept the crankshaft. It was then super-glued onto the brass boss.

That completes the flywheel.

Step 5: The Support Pillar and Bearings

Recognition for the idea for the support pillar and bearings must go to eVolti which was an inspired bit of 'ible.
I decided to use a very similar system, just a bit simpler in execution.

Using the head arm and bearings from the HDD that I got the platters from for my LTD engine, I drilled the shaft 2.5mm to accept the crankshaft. The head had originally had four read arms, three of these were ground off with the Dremel and all the ancillary coils, wire and the actual heads were removed.

The arm was carefully ground to fit a piece of aluminium channel, then the channel was drilled and the arm bolted to it. A little epoxy glue was added for a really good strong joint.
The channel was cut 60mm long, but you can set the size to whatever you require.
The bottom of the channel was bolted to the third heat-sink I found in the old TV which had a good 90 degree bend in it. The bend will be bolted to the main chassis plate.

Step 6: The Crankshaft

The crankshaft consists of a short length of 2.5mm tube. This is passed through the bearing head and carries a flywheel on one side and the cranks on the other.

The cranks are formed from 15mm disks of brass. You could cut these off a piece of brass bar, as I intended to do, but whilst I was in the plumbing section of the hardware store buying the compression end cap for the diffuser cylinder I came across a pack of four 15mm 'blanking caps' for 79p. These will do the job very well.

I drilled the centres of three of the caps 2.5mm to accept the shaft, then I drilled 1.5mm offset by 5mm from the centre for the crank rods. This will give a stroke of 10mm.

The first disk was slid onto one end of the shaft and soldered in place. The distance between the cylinders (26mm) was marked onto the shaft and the pair of disks for the second crank were slid on and soldered in place.
The crank rods were slid into the disks, the con-rod connectors slid onto the rods and then the rods were soldered.

Finally the extra metal on the rods and the main shaft of the crank were ground out using the Dremel. The shaft was secured to the bearing using some super-glue gel and the flywheel fitted.

Step 7: Final Assembly

I made up a small connector for the diffuser piston, then the con-rods were measured and cut to length. These were then glued into position using super-glue gel.

The engine was tested and adjusted to ensure that everything turns freely, and for balance. The flywheel will need some work to get better balance but the engine should work.

I need to make a stand so that it can be positioned above a tea light candle but that can wait.

I may get around to posting a video at some stage.


Brandonoutdoors made it!(author)2016-10-12

Would love to see some video of this running :)

al46320 made it!(author)2016-02-16

Hi. Nice design, but, from a thermodynamic point of view, I'm not sure that copper cylinders (or any metal or good thermal conducting material) are a good idea to improve engine efficiancy : you will lose energy directly in the surrounding air. I think that glass or resines or other kind of low thermal conductivity should fit better. But, of course, it has to sustain the temperature of your heating source !!!

An+Gel+ET made it!(author)2016-05-28

better temperature conducting, for better energy transfer and better performance, ;)

jimandre22 made it!(author)2016-02-17

What is the problem with this? There's a leak at the other cylinder. Can you give us some advice on hoe to fix this? It's not working too. :(

rodgeyrog made it!(author)2016-02-28

Use Sugru

Jim+Andr%C3%A9T made it!(author)2016-03-07

what sugru?

Jim+Andr%C3%A9T made it!(author)2016-03-07

what sugru?

jimandre22 made it!(author)2016-02-17

What is the problem with this? There's a leak at the other cylinder. Can you give us some advice on hoe to fix this? It's not working too. :(

HenryE6 made it!(author)2015-09-18

Hi Marshon, I had an idea for a sterling design but i dont know how plausable it is.

Would it work if you had two burner cylenders connected to each other with a very thin skin pipe that has a it more air volume than the displacer takes up in the burner cylinder. The pipe is submerged in a cooling system to quickly cool down the air.Thus the air would move from the burner to the pipe and back, not from burner to burner.mount the two at 180 degrees and it should run. What do you think. Contact me at : Henry Eckard

copen84 made it!(author)2015-09-16

I am in the process of building one of these awesome engines, and came across yours on here. Your ideas have helped me immensely.

T0BY made it!(author)2015-06-23

My goodness! You have really done a beautiful job here!

NeilRG made it!(author)2015-05-29

I was inspired by the way you fabbed the crankshaft. Very clever. I was aso inspired by your use of the hdd bearing. Very nice presentation. Thank you for the post. Neil R

fdiaz9 made it!(author)2013-11-15

approximately how much power can you get from your engines?

conopodium made it!(author)2013-03-25

Re: "I then cast the tube using the resin process in my other instructable."
Is this the process with the car filler?

rimar2000 made it!(author)2012-09-24

WOW, this is a very good instructable! Congratulations.

I don't know why I did not see it when published.

Greyfox95 made it!(author)2011-10-16

looks very steampunk

cgc123 made it!(author)2011-08-17

does the small piston size matter?

marshon made it!(author)2011-08-23

Only relative to the larger one. Generally it seems that there is great debate about the relative merits and drawbacks. Experimentation is the key here.

The displacer cylinders function is simply to help move the air from the 'hot' side to the 'cold' side and vice versa. There are versions of the Stirling that require no displacer cylinder at all.

To run, the displacer cylinder must be at least 1.5 times the volume of the power cylinder, but I have run engines that had 10 times the volume. You see the displacer piston is not a gas tight seal fit to the cylinder whereas the power piston is.

winn2 made it!(author)2011-08-06

Oh..and one more thing- are you experienced with this sort of technology? If so, would I have to be skilled in order to construct one of these engines as well? I'm probably driving you nuts with all of my questions, but if you could answer these questions it would be great.

marshon made it!(author)2011-08-07

Well. I'm no engineer that much I can tell you.
I have only my own experiences to draw on, no formal taining whatsoever.
I have simply aquired skills here and there, done a LOT of reading and watching the right TV programs.

winn2 made it!(author)2011-08-06

How much did it cost for all these items and where did you get them? ......You must be a genius to make all this stuff! Are you a skilled engineer?

marshon made it!(author)2011-08-07

Nothing, they were all hanging around the workshop. It pays never to throw anything away.....
I have no idea what they would cost to buy.

winn2 made it!(author)2011-08-05

How long did it take you to make this sterling engine?

marshon made it!(author)2011-08-06

I did it over a weekend.

Chowmix12 made it!(author)2011-07-05

Oooh pretty... 0_o me like.

RocketManDave made it!(author)2010-11-22

Makes you wonder why computer fans can't be powered by the heat of the processor? Or am I going made here? ;) :P

Geoff56124 made it!(author)2011-02-24

Hello Rocket man MSI did develop one but I have searvched all over and never found a n example of it.


ARJOON made it!(author)2011-03-02

is this it that you were searching?

RocketManDave made it!(author)2011-02-25

Brilliant, although it looks like the offspring of a toilet and a swamp boat. Mmm efficiency.

stormthirst made it!(author)2011-01-31

Because electric fans are smaller, and given the cost of copper these days, probably cheaper. It's all about how cheap components are these days. Ever wonder why the fans in computers are so noisy? Its cos they've got cheap bearings causing the fan to vibrate more.

RocketManDave made it!(author)2011-02-25

Fair point, I wasn't looking from a manufacturers POV just considering the possibilities for a mad inventor.

kcbford1 made it!(author)2011-02-19

nicely done, i will working on this soon!! Thanks

Egoritz made it!(author)2010-12-08

Great Build, I liked your use of the brass blanks on your crankshaft.

Bongmaster made it!(author)2010-10-01


its ok havent smoked in many years ;)

texhon3 made it!(author)2010-06-26

how can you use the power? I almost see a sterling attached to something to do real work..

LinearB made it!(author)2010-09-02

Hi. I was asked to help design a dam / hydroelectric thing, but the fall was insufficient. It flows into a pond with about a 40 degree temp differential, so i looked at stirling engines. This company claims they've got one that will both power a house and heat the water. Haven't seen one yet, but it will apparently run off propane.

marshon made it!(author)2010-06-26

The two engines I have made so far are purely experimental. Ultimately I intend to make one to run a squirrel cage blower for an aluminium forge, driven by a heat tap from the forge itself.

savaSS made it!(author)2010-08-01

Where did you buy the copper clad paxolin. I'm having a very hard time finding it

marshon made it!(author)2010-08-02

I had mine in the scrap box. Just use single or double clad PCB board instead. You can get that from RS, Farnell or Maplin.

6raleyh made it!(author)2010-07-26

Hi Would you be able to post a video of this model working? Thanks

marshon made it!(author)2010-08-02

There's a link at he top of the page

Nerdz made it!(author)2010-06-21
For those who want some mathematics involved (ie formula) here is a great reference on a forum:

One of the comments mentions a book:

I might look into getting this book. Ive been working on my own version of yours for a while, Ive spent 14 bucks so far (since my last comment anyway).

Hagar1 made it!(author)2010-06-25


Hagar1 made it!(author)2010-06-18

I have know the formula to calculate the power for an engine like this. I don't know how it went exacly but it must have been something like this. Anybody that does know the formula? I know the power ~ ΔTemperature³ x volume² x 10^-8 x Cte x ... (or 1/ volume(²)) The power depends on the temp difference, a little on the volume (altough it could be it was 1/ volume² and then some constants depending on the density of the gas and it's heatcapacity, the stroke and bore of the engine, how closse the components resemble the perfect adiabatic components of a perfect engine... So does this ring a bell with anywone?

Blackice504 made it!(author)2010-06-17

i have been interested in this type of engine since i have seen them however most of the time i have never been able to contact its owner so i ask you what sort of torque or rpm do you get from this device. Thanks for the great info on how to build and i now understand more on how this type of engine works.

marshon made it!(author)2010-06-18

I'm no engineer, and certainly didn't work anything out using gas flow thermal dynamics. You'd need to ask someone who knows about these things. However, over the two first experimental engines I have found that (testing them to destruction) the limiting factors are mechanical design, and temperature (differential). The LTD ran between about 50 RPM and 300 RPM before the temperature started to soften and deform the walls of the diffuser cylinder (made from a plastic tub). I am working on a glass cylinder to improve this. The metal engine shown here runs between about 200 and 700 RPM but the bearing for the diffuser decided to bind and this bent the crankshaft and the con-rod, so I'm working on improvements here too. The torque and therefore the power is more dependant on the temperature and the speed at which the air can be heated and cooled effectively. I have no idea as to the Newton Lbs Ft figure, but for these small engines it would be quite low. Theoretically, these engines could be nearly as powerful as a steam engine, but that would depend on materials and design. Eventually I'm aiming to build one to drive a squirrel cage fan for a forge.

Bubbler made it!(author)2010-06-17

This is a great looking Stirling engine, and you have done a great job. A video would be much appreciated, and perhaps you can enter it onto Boyd's Tin Can Stirling web site?

atraxrobustus made it!(author)2010-06-17

Very cool! I'm thinkin about it. Perhaps have power one or two leds.. Nice Job!!

Wesley666 made it!(author)2010-06-14

Very nice! For the piston, I have one I made awhile back, I used Epoxy Putty and it woked fine. I had to make two cause I didn't grease the pipe enough to get it out on the first try, but the second one is amazing. I also had the idea of using copper pipe, but I used some high strength epoxy instead of soldering mostly cause I had alot of it anyhow, and it can withstand 2000C heat. Cool Instructible! :D

marshon made it!(author)2010-06-14

Yes I have seen some nice examples using epoxy putty. I did try using white Milliput but I broke it trying to get it out of the pipe and internally cast resin shrank too much. I may try it again sometime though.

About This Instructable




Bio: Untidy, disorganised and a bit silly. I am a photographer, artist, body artist, sculptor, prosthetic maker, model engineer, and general idiot who likes making stuff ... More »
More by marshon:Upgrade Courtenay Solaflash 1000Making Faux Feathers for Costumes and PropsHalloween Gargoyle Wings
Add instructable to: