All sorts of interesting custom bikes and engine projects are coming out of Sweden all of a sudden, or at least, I’m just coming across them, and John Ellwood thought we might like to see one more, it’s his Ellwood Hybrid, a single cylinder, 4 stroke with reed valves, from a 2 stroke, mounted on the crankcase directing the crankcase pressure to charge the intake twice per power stoke. The exhaust stroke pumps the intake once and the reed valve holds the charge, the intake stroke pumps it the second time when it goes into the cylinder, so you have a 2 stroke like pumping action from under the piston and a 4 stroke firing cycle.
John Ellwood, who built this engine, says hybrid engines working on this principle began back in 1915. John got the idea to build one back in 1994 and he’s been working on the constantly evolving project ever since.
The first Ellwood Hybrid was based on a Godden engine, commonly found in speedway motorcycles, he’s also built one with a Jawa engine. The first one was converted in a pretty straightforward manner, as John describes it:
A hacksaw, drill and file were enough to slaughter the speedway engine. A bit of rubber pipe, two reed valves from an RD 350, a seal on the crankcase and a carb from a Rotax converted a relic into a Hybrid.
He has both a 500cc version which he says puts out an estimated 50 hp and a 1300cc version doing somewhere around 150 hp. The 1300 is all handcrafted and it’s fuel injected and water cooled with 4 crankcase mounted reed valves, a belt driven rotary cross head valve, an intercooler and 15:1 compression.
John has been racing the 500cc version and plans to race the 1300 in Supermono racing in the U.K. so it must run pretty well. I just think it’s an interesting combination of 2 and 4 stroke technology.
John has done a lot of work without the resources of any expensive engine building facilities. I wonder how this would run if really refined and, with the right equipment, it would be much easier to determine how much extra power this setup actually provides. Definitely different.
John adds: “…here in Sweden we’re having a flying kilometer race at the end of March, Bonneville Salt Flats style – only we do it on ice. I’ve entered the black 500 Hybrid, and will put on the laughing gas tube, to get a bit extra boost.” I checked out the ice racing website and got a chuckle, there’s a note, “Don’t bring any salt.”
Link: Ellwood Hybrid
Link: Landracing.se
Tin Man 2 says
Very Interesting concept, Kind of using the rotating crank as a small Supercharger.
Bob Jones says
This doesn’t supercharge at all. Other than the piston pushing air in during the intake stroke I can’t see any advantage at all. Certainly not enough to overcome the weight penalty of the crank shaft that has to be designed to fill the case.
1. Intake stroke, the piston forces air out of the crankcase.
2. Compression stroke, the piston draws air into the crankcase.
3. Power stroke, the piston temporarily compresses the air in the crankcase.
4. Exhaust stroke, the crankcase is already full of now-heated air, no more air is going to come through the reed valve.
Actually, one other thing happens that might be advantageous. The compression of the crankcase charge robs power from the power stroke, but that work is returned to the piston in the exhaust stroke, so this will smooth out the power pulse. Of course, with a case-filling crankshaft it should have all the flywheel effect it’ll ever need anyway.
James Bowman says
Very interesting I think that what would make it work much better would be a boost storage chamber maybe in the frame to act as a filter/buffer to give a consistant pressure, then he could just follow more explored paths with boost such as longer exaust cam duration etc… It surely seems a waste to use the non firing stroke beyond pushing out the exaust gasses. I think then we might see the types of power that go well beyond a normally aspirated 4 stroke.
Hawk says
I see this as similar to the Ducati V-One that was discussed a few weeks ago. What sparked my interest with the V-1 and presumably here too, is that if the plenum chamber is increased in size, it results in a prolonged (10 second) power boost at sudden full throttle.
This is the kind of thinking needed because we aren’t going to get away from internal combustion engines any time soon. Making them more efficient by supercharging, heat recovery, better flame pattern or whatever will help to reduce our fuel consumption.
Dorzok says
if he’s getting these hp numbers with no other high tech gadgetry then there’s gotta be something to it. Not enough details here for me to follow why it works but if those numbers are real then good on him.
John says
What about oil mist from the crankcase polluting the intake charge?Or if that is not a problem are the mains and rod bearings getting sufficiant oiling?
Swagger says
Bob, I don’t see a revised crankshaft mentioned anywhere in this article, do you see something we don’t? Since the carb is dumping into the crank case like a rotary valve 2-stroke, one can fairly safely assume that it’s probably running premix and lubing the bearing thusly.
Since the volume in the crank case decreases as the piston drops in the bore and the pressure is stored then further compressed in the chamber, he’s essentially getting piston pump supercharging. Looks like it works to me. Neat idea!
Bob Jones says
Swagger. if you look at a 2-stroke crankshaft, it’s designed to fill the case as much as possible so that the maximum percentage of the volume is under the piston. The natural result of that is the crank is larger than it would have to be. The volume of the crankcase does decrease during the intake stroke, but there’s no supercharging. The charge pressure can’t go over atmospheric other than by the heating of sitting in the crankcase… which doesn’t increase the actual mass of air sent so isn’t at all beneficial.
As for the output, a mass-produced CRF450 puts close to 50 rwhp stock, so I’m not hugely impressed with this amount of power out of a hand-built engine.
coho says
How does it compare weight-wise to an equivalent conventional 500 single?
If it’s enough lighter…
todd says
Bob, I think you forgot that there are reed valves on the crank-case. When the piston comes down, it doesn’t compress the air in the case, it forces it into the second intake manifold (or charge tank). That manifold benefits from two fillings before the intake valve opens. It also has an air-to-air intercooler. If this is sized just right, at some RPMs the velocity of the air flowing into the manifold may be high enough to create a vacuum in the case – drawing in the next charge. This is fairly typical in conventional engines, one way of achieving over 100% volumetric efficiency.
Fuel injection would allow a larger charge tank in the intake like the Ducati V-One. With a carbureted version there is the problem of fuel vapor falling out of suspension , i.e. puddling. The “modified crank” allows the mass of the flywheel to be moved adjacent to the connecting rod, creating a stronger crank and greatly reduced timing fluctuations (less crank flex). The case itself can be designed like a two-stroke with a unitized, albeit divided, transmission. It’s not all win-win for sure; blow-by would contaminate the intake charge, necessarily long intake runners limit high RPMs (but help increase torque), ring and ring groove life is decreased due to the higher charge pressures and the extra pumping, without baffling oil might accumulate in the intake… This is all measured against the gains.
Indeed, 50HP on a 500 single is certainly remarkable. My 4-valve, hemi-head, OHC GB500 is rated at 35.
-todd
Bob Jones says
I did in fact miss the second set of reed valves. Mea culpa.
OMMAG says
The fresh paper sort of thinking that produces this sort of innovation is a real pleasure to see.
Even if it’s not immediately practical it shows that there is plenty of innovation going on all over the place.
One thing in this case that gets me is how a little bit of recovered energy makes such a big impact on the overall output. Assuming that the power is actually near that 100hp per liter and not just a calculation from volumetrics.
gafe03 says
Clench your teeth! That 1300cc engine rigidly mounted in the frame is going to vibrate the whole plot like a milk homogenizer!
They used to take Manx Nortons out for open class racing & hill climbs, some up to 700cc. But they’d come back cracks everywhere or dribbling off bits & pieces due to the vibration.
But what a blast til then! The first Sadler sports cars only had 1 gear as transaxles were not yet available. Just tune the big V-8 for a broad torque band & slip the clutch for starts! Worked for them! Why not for the brave Ellwood!
rafe03
rafe03 says
It’s neat to see someone else try the Roland Cross Rotary Valve again, Norton had 1 running the Isle of Man TT back in the 30’s. No better than the regular double knocker so they didn’t take the time & effort to spend on development.
One of the big oil companies (Shell?) made an X-4 research engine with Cross Valves that looked great in the pics but never reappeared. In the 70’s I believe. Got some pics somewhere.
Cross & Aspin Rotary Valves had the best chance for technical & commercial success but neither were taken seriously by any of the big makers. Perhaps the advanced tribological technology available today might overcome lubrication, the biggest single problem in their day
Good Luck John Ellwood!
rafe03
PS: For further info on these & other interesting engines, see:– “Some Unusual Engines” by LJK Setright. Institute of Mechanical Engineers, 1975. ISBN 0 85298 208 9
— No! my copy is not for sale!!
rafe03
rafe03 says
Have a look in Motor Cycle News from back in the 70’s & 80’s. Several clever chaps (Peter Inchley? Arthur Lampkin?) tried similar, some with company sponsorship. (Honda?)
Take a standard 250cc 2-stroke engine bottom end & top it with a standard 4-stroke SOHC head. Chain or belt drive to the cam. Oil injection pump to the bottom bearings. Plenum & reed valves would have to be designed to provide resonant tuning or the pumping losses would eat away any gains. Intake valve cams might need some tweeking as well. I think the bikes were to be used for trials but I’m not sure after all this time.
Ah! The good old days! Glad to see John Ellwood revisiting them to see what good got cast aside with the clinkers! Good Luck!
rafe03
John Ellwood says
Hi guys, thanks for the positive feedback, however critiscism is also welcome, as it is an important factor in development.
The 50kg watercooled 1300cc engine pumps twice underneath piston, has a normal homemade crankshaft, and cork filled piston. Never had a problem with the 2stroke oil pumping into the mains bearing, or bad smoke emissions.(apart from when I forgot to close the oil tap and it filled the crankspace overnight before the race, resulting in massive smokescreen)
A big surge tank is presently being built as a onepeice seat/rearend before intercooler. Big plenum chamber with reverse shower injectors after the intercooler.
This bike is a purpose built 1300cc for British Supermono racing (not an up-borred weaker structure), and can also be de-stroked to 799cc for other Supermono events.
The decision to run only with a clutch (no gearbox) is to save weight and friction.
The rotary head valve is there for my interest in such mechanisms, plus its easier for me to construct ( a 4 valve normal head would take me 2 years to make), And the piston can`t get broken by smacking into normal valves at high revs (any problems and the cambelt just snaps).
I realise there will be mega vibes, I shall use Fakir handgrips to damp this effect.
Hybrid enthusiasts :- Dont miss on my website the book (Oil Sweat and Inspiration) on all my 15 Hybrid experimenting years, no secrets held back !
Ellwood
Gavin says
Approximately 15 years ago I built exactly the same device with the assistance of a local engineer. Our prototype utilised a Honda CG100 top end crudely mated to a Yamaha AG100 2-stroke bottom end. A carb and reed valve inlet assembly was fitted to the front of the AG crankcase, with a second set of (discharge) reeds fitted to the top of the casing with a steel tube plenum feeding the CG inlet tract.
Camshaft lubrication was a total loss system consisting of an oil supply tank and an electric chemical dosing pump. The engine ran on 2-stroke mix, and from memory the CG piston still had the standard 4-stroke rings.
Our initial impression of the engine when running was that it produced good power, as its throttle response was instantaneous and far more immediate than either of the donor engines would have produced. However, this was soon dispelled when we ran the hybrid alongside a standard CG100 on a Dynojet dynomometer. Unfortunately there was no measurable difference in power output, and unconvinced we resorted to racing the two bikes side by side. This only served to confirm the dyno results.
During the time that this project was being developed I obtained a provisional patent for it here in New Zealand, which was allowed to lapse once we reached the conclusion that there was no real mechanical advantage due to pumping losses and limited supercharging effect. A few years later I read that BMW (I think) and a British specialist engine development company (their name eludes me at present) were doing some serious development on this same principal with claims that significant performance improvements were likely.
I have read nothing more since regarding either project so can only conclude that they reached the same conclusion that we did.
That said, John’s work is very impressive and I wish him every success.
C-Petteri says
Something similar was done to this old british Douglas motorcycle
Read Doug´s answer from thread to get a sense what this engine is all about.
pat neal says
rafe 030 had it right;it was Peter Inchly in 68,My late bro in law,Pat Norris and Peter worked for Joe Erlich [EMC] and that was one of Joes funded projects [MZ money ?] They got 28 hp out of a 125 twin with less harmonic pulsing than a single and used reeds on the intake and rotary valves on the pressure side.Unfortinatly they got more power with a 1/2 turbo experiment,using a exhaust driven turbine stepped down 12:1 directly back into the crank ala wright radial and napier deltic so this took precident………………Pat
Steve says
It looks good, and true to the saying, ‘want it to work?…keep it simple’
but power to weight is suspect i feel….especially put against the work currently being done by Norton on their rotary
Paulinator says
I had a Norton Commando…Did the “thought-experiment” of feeding the carbs into the crank-case thru reeds and running crank-case compressed mixture up to the head thru the push-rod passage between the barrels. The in-take ports would then be blocked and re-routed into the rocker boxes. The swept volume was 827cc – but this volume would be sqeezed into one cylinder of 2 every 360 degrees. I know that it wouldn’t generate 14 psi boost, but it would’ve realized some gains. I gave no real thought to the librication issues, however.
I did some inventing on two-strokes with assymetric port-timing. I built a couple of those. One was a “runner”. I smiled so much that my lip split.
pat neal, turbo-compound Douglas. Who knew?
John Ellwood, That’s good looking stuff. Rock on!!!
rafe03 says
Today’s noninee for the Golden GearHead Award for “Thinking Outside The Box”:—
—– The STIHL 4-Mix ENGINE!
It’s got a 2-stroke below the belt & a 4-stroke in the head! Matbe someone was looking over John’s shoulder? Or Gavin’s shoulder? Or Peter’s Shoulder? Or Pat’s shoulder? Or ……
see -<http://www.stihl.ca/4mix/english/stihl4mix/mixprinzip.htm
for an explaination & annimation
The main thrust of Stihl’s development seems to be saving weight while still insuring that lubrication is adequately handled in any position. I don’t think they consider power as long as it is “adequate”!
This string is really bringing out the shadowy “CrackPot” ideas & undeveloped projects from under the stairs & the back of the shed. Good to see that lateral thinking is still alive & well!
rafe03
Mark says
Paulinator, I want to talk to you about your inventions and an idea of my own. I love 2-strokes with a passion, and you seem to know a bit about them.
rafe03 says
On John’s proposal to save weight by running a clutch & no gearbox on his 1300cc “vibrator”, there is a class of Karts racing out there that use a centrifugal clutch direct connected to the final drive. In order to widen the usable speed range of the Kart, a shot some kind of light lubricating oil/solvent/(?) is blasted into the clutch on demand (at low speed?) allowing some drive slippage for a short time. The oil will quickly burn/evaporate off allowing the clutch to regain it’s grip.
Bearing in mind that I’m not up-to-speed in Karts, this could allow the engine to run higher in its speed range for:- 1. More power at higher revs (?);- 2. Avoids having the engine “fall off the pipe” or bog just when HP is needed out of the corner(?);- 3. Overdrive for passing in slow corners (?);- 4. Laying down a quick smoke screen while the “squirt” burns off in the heat of friction (?)
There’ll be someone reading this who has a finger in the Karting Pie & perhaps could elaborate from a basis of knowledge (rather than speculation like I’m doing!)
Keep up the good work, Paul! & Guys, keep those cards & letters pouring in!
rafe03
John Ellwood says
Wow, the Hybrids are coming out of the woodwork now. Maybe the day isnt far away before we have Hybrid JEGS ?
Just been to Charley Karlings workshop today for some tips on a clutch. Probably be going for a Bandit Sportster style on splines with a spider and outside actuator. Not sure yet if to opt for the centrifugal weight plate as well for an extra drag-like getaway. Anyone got any tips on this?
mike gullatt says
We are working on a tech 4 chevy s-10 engine right now putting 4 big reed valves in the intake and 2 big dellorto 2-barrels making an IR setup with a long duration intake timing which usually kills the low end and midrange torque on these engines. The reeds control the overlap and reversion at low speed letting it pull like a short timed cam but able to use the extra cam timing as rpm goes up. Best of both worlds and amounts to poor mans variable valve timing like my wife’s nissan. Mileage is also very good since the intake is tuned length and the reeds stop most all the backflow that usually blows a fuel fog back out the intake on quick throttle. Seems pretty practical so far if reed life isn’t a problem. Anyone playing with this?
Big Sven says
Sorry, Mr. Ellwood, old idea warmed up.
In ’76 CCM in England ‘accidently’ let slip about their ultra-secret new and revolutionary ‘5-stroke’ engine for motocross. Husqvarna’s Reubin Helmin raised an eye at this wonder engine, and as I was going over to bury my mum and a friend had broken the ‘unbreakable’ 3-speed gearbox on his CCM, I was asked by a journalist from Sweden’s top bike mag to cable and ask to visit the factory with the broken bits and explain why it had broken, maybe getting an interview when there, and keep my eyes open for this wonder-engine. “5-stroke’s are impossible,” I’d been briefed, “But we think we know what he’s on about. Doesn’t work. Some Austrian guy is already racing something similar. The response seems ok, but the power isn’t impressive as the charge from the crankcase gets hellish hot.
So I was at the CCM factory and saw the 5-stroke. It was just like your engine, just like the Austrian’s engine. The ‘5th’ stroke wasn’t, it just described the charge being pushed up the pipe to the head. I had already bawled-out the guy who made the frames, Eatough (?) in a small factory just up the road, as the frames cracked around the tiny oil-filler hole by the steeringhead. “That’s impossible, our frames NEVER crack, this is the strongest frame ever made, it’s never happened before!” didn’t bite, I had the photos to prove it. So I was wired and ready for Clews! Clews had heard me coming, and was wary, begrudgingly eyeing the broken gearbox. I began pushing him, vying for an article. Despite me waving an introductory letter from the top Swedish motorbike magazine editor under his nose Clews wasn’t keen on the idea, but I was was in the mood to be ‘persuasive’! So I eventually found out he’d used-up all the genuine crankcases machined by BSA’s supplier and had farmed-out the machining of the casting he had left to some cowboy (my description) and thus all the shaft holes were out of true, none of the gear-shafts ran parallel, the strain on the gear teeth now far beyond the design parameters. Big bang. Measuring-up an old 4-speed case SKF machined the cases and bushed them, curing the problem. But who was going to pay for the new gearbox, expensive, I probed? Despite he realising he was risking CCM’s name being slagged in the press he didn’t offer even a percentage. “Racing bikes break. He must have done something wrong, didn’t service the bike properly.”
Mmmm. As the chap concerned had a degree in civil engineering, and had helped in the design of, and personally handmade and race-tested, the first prototype HVA 360cc motocross engine (his was actually a 380) based on the new mag-engined 250, that HVA then developed into the bike Heikki Mikkola won a World Championship on, and was a up and coming development ‘whiz-kid’ at Volvo during the day, I somehow doubted that. But Clews didn’t bite. I was pomposely refused to take any photos of the ‘5-stroke’ it being secret. But the press already know about it? He ‘had to protect his investment as the patents weren’t approved yet.’
And never would be, the Austrian had already taken-out a patent!
Nice try, Ellwood, beats looking at TV or getting rat-arsed down the pub, but ‘no-go’ I’m afraid. Don’t give up your day job.
John Ellwood says
Hi Big Sven,
I heard CCM tried this too.
The pumped air is 68 degrees C, so my new engine has a whopping intercooler and reservoir tank, and water injection, runs on methanol / ethanol/ petrol.It even has an oil separator.
Hope to have the whole bike ready for next years Bike Exhibition in Stockholm so try to come by and view the beast there.. Patents and keeping up good reputations dont affect my project, and its accepted to race in British Supermono, so next season will be very interesting.
P.S. The landracing gang will be staging another flying kilometer on ice again in March, so Hybrid fans can see the original 500 Hybrid attempting to better its world record.
Regards, Elwood
Olaf Larsen says
This is a very interesting development in engines. I have been playing with the idea ever since I worked with motorcycles back in the ’70s, but as with so many of us – it stayed in the brain and didn’t get on the road! Mike Gullatt’s idea of using reed valves as a supplement to ordinary valves on a 4-stroke, is exactly what I am designing now, and I would very much like to get in contact to learn more about his experience with the Chevy engine. ( his link doesn’t work) Anybody else doing similar stuff?
I’m at olafla att ifi dott uio dott no.
Frank says
Excellent work John – good luck on the ice!
John Ellwood says
The 2010 flying kilometer on ice went reasonably well with the 500, after a very bad start managed an average speed of 94 mph (151.14 kmh )
Garry Briggs says
Hi all,
Found this discussion after a quick search for reed valved 4 strokes. Reason being, has anyone tried putting a reed valve between carb and inlet valves to stop reversion with large cam overlap. What will happen? Will the flow losses to too great? Should I try it?
Thoughts please
John Ellwood says
Hi Garry, yes its been done with a reedvalve between carb and inlet valve. i read somewhere years ago. it should help the bottom end i would think, but would restrict the full throttle part. swings and roundabouts. but maybe if a servo was present to manually open those reeds at full throttle—who knows.
John Ellwood says
Just for interest, if I should build a series, would anyone actually consider buying one of my engines, for use in mc, snowmobile or boat ???