Model Engine Maker
Supporting => Engine Ancillaries => Topic started by: Roger B on March 09, 2020, 01:33:17 PM
-
There have been many model aircraft and boats powered by home built IC engines but very few land vehicles. ETW designed the road roller and the 1831 locomotive. Roy Amesbury made a V8 powered Hymek. Gail made some hit and miss powered gauge one locomotives. There may be a few others.
A reason for this seems to be carburation. If the engine is driving a propeller (planes and boats) as the revs drop the load drops and visa-versa. For a land vehicle as the revs drop the load can increase such as starting to climb an incline. A simple barrel type carburettor with a suction feed has problems with this. As the revs drop the airflow through the carburettor drops and so the suction is reduced resulting in the mixture leaning until the engine stops firing. Opening the throttle makes the problem worse. As the level in the fuel tank drops the mixture also becomes leaner.
These problems were noted by ETW when designing the 1831 and he developed a carburettor with a float chamber to maintain a constant fuel level and an auxiliary air valve to maintain a fairly constant vacuum over the jet. This was/is apparently successful.
Roy Amesbury discusses similar problems with his V8 powered Hymek. He tried several designs before settling on an SU like constant vacuum type. This design also had a constant fuel level but used a pump and overflow system which he considered better for small engines than a float chamber.
As part of the experiments with my 12cc twin I am going to design and build a slightly smaller single choke version of the Amesbury carburettor including a small diaphragm pump for the fuel feed. This pump will hopefully also be useable at a higher pressure (~0.8bar) as a feed pump for my petrol injection system. The Amesbury version used a 9mm choke bore, 19mm piston and 0.8mm jet. I will try a 6mm choke bore 12mm piston and 0.5mm jet. Attached are some initial sketches. The carburettor uses a barrel throttle as that is easier to make than a butterfly type in small sizes. I will have to make some experiments to find a suitable mass for the piston as air leakage will play a significant part. The pump cam will pull the diaphragm back for the suction stroke and a spring will be used to deliver the fuel.
-
Roger, I look forward to seeing what you come up with!
-
Hi Roger,
I will be following your experiments with great interest. I think they may be very relevant to running my Seal engine in the Bren Gun Carrier. That will be exactly the type of load this carburettor is intended for.
As I understand the SU. The piston goes up and down to maintain a constant vacuum in the venturi between the base of the piston and the jet. When airflow demand is high, the piston is high and the needle releases the right amount of fuel. Conversely, when airflow demand is lower, the piston rides lower and again the needle releases the right amount of fuel for that airflow. The taper on the needle is all important as it sets the mixture for each operating condition. If the needle's taper is wrong, the mixture will be wrong at one end of the operating range. SU used to offer thousands of different needle tapers for the enthusiast to experiment with. SU also offered a range of springs to control the piston response.
Does the simplicity of Art's chicken hopper offer any advantage over the pump and overflow alternative?
This is going to be very interesting
Mike
-
Hi Mike,
I spent many a happy hour looking through the needle charts to find the best for my modified engines. I expect similar fun making needles for these experiments ::) At least I will be able to measure the piston and hence needle position whilst moving the jet up and down to find an optimum flow rather than guesstimating the position whilst pulling the choke out at illegal speeds on a deserted M27 at night.
I can't really comment between pump and overflow and bird feeder having not yet tried either. Westbury did not think highly of the bird feeder and prefered float chambers. Amesbury seemed to prefer pump and overflow to float chamber :headscratch:
-
Hello Roger,
I expect most of use have cut out teeth trying to tune SU carbs. Going flat out and pulling on the choke may not have been the most scientific approach but we all did it and lived to tell the tale.
I live beside the M27 motorway. So it was probably you who woke me in the middle of the night as you roared past fiddling with the SU carbs. :thinking: :thinking:
Mike
-
I will follow this post with great interest.---Brian
-
Thank you for your support :ThumbsUp: As ever for me this is a long term project, although my travel is currently stopped I will have to catch up with it all when the restrictions are removed.
Mike, if it was a red Triumph Herald Convertible in the 80s I admit responsibility ( the statute of limitations is long passed).
I have started on the diaphragm pump, beginning with a stub of 25mm diameter aluminium bar. I have some 0.5mm thick Viton sheet for the diaphragm on order. The minimum quantity will certainly see me out :old:
-
There are two separate issues with carburetion. Fuel delivery (to the carb) at consistent pressure is one and relatively easily solved. Air/fuel ratio over the load/speed range is the other, but is difficult with simple carbs seen in models. As noted propellers have reasonable relationship between RPM and engine air consumption which generally works well with simple throttles. Traction or other loads with simple carbs leads to unstable air fuel ratio over the speed/load range. The biggest issue with home built model engines, in my opinion, is oversize carbs. If the vacuum signal is so low at WOT throttle that jets(adjustable needles) have to be large, you have little hope at low throttle settings or low speeds. I've run model aircraft four stroke engines at 3000 RPM(vs the typical 10-11kRPM) with oversize propellers, but it wasn't on the original carb. Area was reduced to account for the anticipated air flow, and there was no trouble at all. As a side note, (glow) methanol fuel allows a wider range of combustible air/fuel ratio, and adding nitromethane widens it even more. Gasoline is much more picky about air/fuel than glow fuels, so the carb must be sized correctly.
Specifically for your engine, 6mm for a 12cc four stroke twin is too large if it only runs 6kRPM. I would try something closer to 1/4 the area. I have made 250W+ electrical with a 7.5 CC single cylinder four stroke at ~8kRPM, maybe 4mm venturi, but no throttle. If your engine only breathes enough to make 150W, even smaller venturi may be required than I suggested. We know that going too small will limit power, but a size should be found where it makes good power and has proper fueling. I think reproducing a functional SU design at model scale would be a miracle. Though, when I considered it I was thinking flying models, not stationary engines.
For the injected engine look at the Kugelfischer mechanical injection pump.
-
The trick with the SU carbs is to always keep an oil can to top off the things. My 140 series Volvo's all had SU's
Art
PS It wasn't me I've never been on the M27. :lolb:
-
Greg,
I agree that 6mm is on the large size for this engine however that is the smallest I think I will be able to make. As it has a variable venturi the problem is not so great, the piston will just not rise all the way. I will also try the same carb on my 25cc horizontal engine.
My fuel injection system is based on a diesel style pump like the Kugelfischer system and has the same problem of the petrol vaporising when the plunger makes it's return stroke. I am currently using a small electrically driven gear pump and relief valve to supply the petrol at around 0.8 bar. I hope to use the controlled pressure diaphragm pump as a replacement for this.
-
Interesting that so many had so many problems getting the settings right on SU carbs - where they not standard on those whehicles ?
I only had two motorcycles with the much improved Mikuni CV carbs and they where standard items on those models. The first was a bog standard GSX-1100EZ and I never had any reason to fiddle with the carbs (there where four in-line), other than ensurering that they all where clean, corretly jetted (original sized), needle in correct groove and correct float height.
The second one was a different matter - a GSX-R1100H (Gixxer) and it had an obnosxiously LOUD aftermarked exhaust, open airfilters and a Jet-Kit installed when I bought it - but it was extremely fast for the day. The 1/4 mile in less than 10 seconds and more than 200 Km/h. / 125mph exit speed. The engine itself where bog standard.
I knew when I bought it that I could not live with the racket, so I bought a never version of the exhaust system (can't remember brand) that was a lot less noisy, but almost exactly same power. I also changed the four individual K&N filters out with one that fitted inside the original airbox, as this also reduced noise and DO NOT have any problems running in the rain ;D
I therefore knew that I had to redo the jetting and thank God I actually wrote the settings I bought it with down above the orignal settings in the original Suzuki Workshop Manual.
I started with adjusting the float level back to the original height, put it all back together and did a test run => put bigger jets in and after a few tries all was apparently good ....
It turned out that it lost over 100Km/h (~60mph) in topspeed and used more fuel :Mad:
The secret turned out to be the float height that I had written down. The lower fuel height made the carburators mix the air and fuel much better, as it opened more of the air holes in the Needle Jet - premixing more air into the fuel (making the droplets smaller) before it was lifted into the airstream in the venturi ...!... ah the small details ;)
-
Hello Roger,
If you make the needle in the dashpot adjustable, you could use the needle to set the correct mixture at tick-over then re-adjust the mixture at WOT and all points in between. This could give you a good indication as to what the final needle taper should look like. Just a thought.
Cheers
Mike
PS for Per, young lads (would be engine tuners) just love to fiddle with the SU carb, in the hope of higher performance. They could always go back to the factory needles and settings if (when) they cocked it up
-
Did anyone here make the 1/4 scale Stromberg Carb from Strictly IC and if so how did it perform? Pretty sure that incorporated an accelerator pump, fast idle cam and a choke. It was made using EDM but the author stated that it could be milled. I think it was by Lee Root somewhere around issue 29 onwards....
-
dieselpilot - you say a 6mm carbie for a 12cc engine is too large, I'll be pleased to take your word for that no worries even if I don't fully understand why. I have a .33cu.in sleeve valve 4 stroke aero engine (about 5.4cc) that won't go. Part of the problem, which I'm working on, is lack of compression, the other is the carbie. The original carbie design that came with the plans I am unconvinced on, I put a Chris Turner designed carbie on, from memory a 7mm, again am not convinced but until I sort the compression out I won't know for certain. What size carbie would you recommend for that small an engine, and what are the guideline to determine suitable size carbies against engine size? Presumable there is some 'rule of thumb' that applies?
Roger - I hope you don't think I'm trying to deflect or detract from your thread here, rather I am trying to get a feel for why a carbie is deemed too large for an engine by relating it to a known (to me) engine and thus understand what is going on.
! will follow this thread with considerable interest as carbies in general are a bit of a black hole for me; hopefully, here I will learn something! I look forward to reading your progress........
Chris
-
Gentlemen,
There are many different styles of gasoline delivery carburetors, throttle plate, basic slide, CV (constant velocity), air bleed etc. The thing they all have in common is to deliver the fuel at the best possible mixture with the air. 14.7:1 ratio.
We'll stick with the SU (CV type) as this is what the conversation is primarily about. The CV carb has 4 basic circuits, enrichment (for starting) idle or low speed controlled by jetting and needle valve, intermediate throttle controlled by a tapered needle which works in conjunction with it's needle jet, and at wide open throttle a predetermined jet.
The needle valve (intermediate throttle) is mounted to the slide which also controls the venturi area. The slide has a diaphragm attached to the top of it. The diaphragm is balanced by atmospheric pressure on one side and air velocity through the carb on the other side. There is also another element that controls the slide movement and that is the throttle plate (butterfly valve) like the type found in most carburetors. When the engine is started at low speed (idle) the slide is at rest (bottom of it's travel. As the throttle plate is opened it allows air to flow through the port in the carb that is connected to the top of the diaphragm. This air flow or pressure drop is lower than atmospheric pressure on the other side of the diaphragm so therefore the slide raises in it's bore and as it rises it lifts the tapered needle out of the needle jet which maintains the proper air/fuel ratio. At wide open throttle the needle/needle jet is open enough that it will allow as much fuel to flow as the main jet will allow. The main jet is mounted at the bottom of the needle jet.
There is a similarly designed carb for RC aircraft. The venturi opening is controlled by a slide much like the CV carb but rather than being lifted by a diaphragm it has a direct link attached to it. The similarity is that it has a slide and a tapered needle valve. It doesn't have a butterfly valve so there is no atmospheric pressures involved.
The original question, as I understand it, is having a carb that will react to engine loading to maintain a constant rpm of the engine with the proper air/fuel mixture.
Here's the scenario! You're driving along with the gas pedal at a fixed point. If you accelerate you push on the gas which in turn opens the butterfly valve which then causes the slide to open which increased the speed of the car. The somewhat misunderstanding as I read it is that when you're driving along at a constant speed and you start to go up a hill and therefore load the engine with the gas pedal still at the same setting the car will start to slow down but the slide valve will raise giving more air/fuel to the engine thereby maintaining a constant speed. This is incorrect. The slide will not move unless the butterfly valve is opened allowing more air to flow through the carb and therefore opening the slide valve more.
What you are kind of describing is a governor. The governor senses a drop in rpm (engine load) and opens the throttle plate more and maintains a constant speed. If a governor control were attached to a CV carb it would do the same thing.
When I got into riding and repairing dirt bikes most of them had slide/needle type carbs but they weren't CV types. These engines were of smaller displacement and by quickly opening the throttle the engine would respond (to a point). As the engines got larger if they used the same non CV carb when the throttle was opened quickly the engine would bog because it would take a big gulp of air but the air velocity through the carb hadn't yet affected the fuel flow so there was a lean point until everything balanced out. By going with the CV carb you could crack the throttle wide open (throttle plate) but the slide/needle valve would react only as the air flow would dictate so the engine wouldn't bog. (Here again to a point).
Most dirt bikes and off road machines nowadays are 4 cycle engines and have been fitted with CV type carbs. The carbs are quite sophisticated and have enrichment control for starting, accelerator pumps for quick throttle response and some even have heating elements to warm the carb quicker.
Although I have never run any of my engines under load I have put friction on the flywheel of my Holt and it didn't die but just slowed down. It has a simple air bleed type carb and the fuel metering is controlled by a needle valve adjusted to a particular air flow. In slowing down with a fixed throttle position the air velocity will drop and therefore it should draw less fuel but the same amount of air which in theory would create a lean condition. I can't prove it because I have never done any testing to prove otherwise. Most model engines aren't working engines they are just examples of the builder's talents.
-
Thank you for thoughts George :) I am indeed looking for mixture compensation rather than governing.
If you load an engine the speed will drop and so will the air velocity through the carburettor. This will reduce the depression and hence the fuel feed resulting in a lean mixture. As you say with a CV/CD type design under load condtions the piston or slide will drop maintaining the air velocity and depression and hence the fuel flow. The (correctly profiled) needle will then adjust the fuel flow to match the reduced air flow.
I started down this route after carrying out some load tests on my vertical twin. If I bought the engine up to speed I could add the load step by step and by adjusting the throttle and needle keep things running. If I put too much load on in one go it would die. If I put the full load on the generator at low revs and then tried to accelerate at some point it would also die.
Chris, all my threads are open for discussion. I have seen various rules of thumb for carburettor size, they are typically between 1/4 and 1/3 of the engine bore. My 12cc twin is 20mm bore so 6mm is at the top end of the range. My 3cc single is 16mm bore and is quite happy with 3-4 mm bore carburettors.
My Triumph Herald was ~74mm bore and used 38mm carburettors.
-
Ah Roger, the lesson to me is to read what people write more carefully! I had missed the vital word 'twin' in your OP! The bore for my single cyl engine is 0.7875", or approx 20mm in new money, so the 6mm bore carbie I have should not be too far out! Many thanks for the info.
Chris
-
Chris, I checked a few dimensions. The homemade carb on my 3cc engine has a 3.2mm venture. It develops more power with the modified commercial carb with a 4.2 mm bore. I am using the 4.2mm bore carb on the 12cc twin at the moment. The 25cc engine runs well with the 4.2 bore carb but develops more power with a commercial slide carb with 5.6mm bore. I get similar power and better throttle response with fuel injection. maybe that's the way to go :headscratch:
I made a centring plug and a 2.2mm punch for the diaphragm pump and punched out a trial diaphragm.
-
Just to follow up on the conversation that wandered over to Stromberg carb in Strictly IC. I looked it up in the index and here is where it is.
Carburetor, A Stromberg in Miniature, by Lee Root/Washburn: 31-3; 32-3; 33-3
I don't have the time right now to look up the articles butit is spread over 3 issues.
Art
-
Art, I looked. :ThumbsDown:
Unfortunately, it is a straight downdraft fixed jet carb, NOT a constant depression/velocity carb with the flexible diaphragm.
Mike
-
A start on the piston and cover. Getting the various diameters concentric is going to be interesting :thinking: I think that I will be able to use a 2mm endmill to bore the 3mm holes ::)
-
Hi Roger,
You are off to a good start with the CD carburettor. :ThumbsUp: :ThumbsUp:
Are you going to make several of each part to allow for future experiments?
What do you think about the idea of making the central needle screw adjustable? You can then adjust the mixture to be spot on at tickover and again at WOT and all points in between. That should give you a clearer indication of what the 'true' needle taper should be. I think the adjustable needle will be give more finesse than an adjustable main jet.
Mike
-
Thank you Mike :ThumbsUp:
I intend to just start with one offs as there may need to be significant changes to the design.
The emission controlled Stromberg CD carburettors had the needle adjustable using a special tool but using the tool disturbed the piston due to it's weight. I always found the older versions where the jet housing was screwed up and down to be easier to adjust. I intend to have some form of adjusting screw (not designed yet) to adjust the height of the jet.
-
The body is from a piece of DIY store (Afler) 15.6mm square aluminium. This was faced, drilled 5.3mm and reamed 5.5mm. I have decided to reduce the bore to 5.5mm but the piston will remain at 6mm. The manifold end was turned down to 8mm to fit the existing adaptors. The body was then milled to 15mm square and the seating for the cover was cut.
I have attached a PDF of the current state of the design.
-
I agree that 6mm is on the large size for this engine however that is the smallest I think I will be able to make. As it has a variable venturi the problem is not so great, the piston will just not rise all the way.
I meant for a standard design carb, you could simply install a reduction sleeve in the carb you're using now. I agree that in theory SU would use only as much area as required, but this assumes it's well calibrated. Even so you are giving up a lot of "resolution", by using only a small portion of the needle if it only opens slightly. Looking at needle profiles, variable tapers in microns over ~50mm, seemingly brutal to duplicate at model scale. Fit of the piston to the suction chamber also looks important. This will be an interesting project.
-
Thank you Greg :ThumbsUp:
If the carb works at all I will make a 'guess' constant taper needle based around the dimensions of the one used in the barrel carb which also has a 0.5mm jet and then try and determine a better profile by moving the jet up and down. I think that getting a dimension every mm and smoothing in between should be sufficient.
Next one the trickier bits. The holes to connect the vacuum by the jet to the space above the piston end up at 0.8mm diameter to give a reasonable thickness of metal around them. These are 9mm deep so plenty of stops for swarf clearance, the cutting fluid tends to make the swarf stick to the drill. This was done at 4800 rpm in my Proxxon drilling machine. Still in the RT I moved to the Proxxon mill to cut the two lead in slots and the anti rotation keyway with a 1mm end mill at 20 000rpm.
-
As the Hobbymat was in milling mode I made the rocker arm support for the diaphragm pump as well as a holder for my 3mm shank Proxxon end mills that I am planning to use as micro boring bars. I should really have moved to the bigger mill to cut the 6mm spring seat but it worked ok.
-
Have you done any math on pressures and spring rates, which make this carb function?
My thoughts on carbs.
The piston in the SU carb is a flow metering device. It measures pressure drop across the venturi(created by the slide itself) and sets venturi area for that air flow rate. In addition the needle affixed to the slide is used to meter fuel accordingly. Really an ingenious device, though somewhat complex and not trivial in manufacture. The main benefit of this design is that needle profiles directly influence air/fuel ratio in very specific regions of operation directly related to volume flow(torque). The throttle plate is there to limit engine torque, just like any other carb. In simplest form, the throttle plate has no direct influence on slide position. The slide in a SU/CV carb is positioned only by air flow demand. Transition shouldn't be confused with basic operation, various methods are used in all carbs to assist when the throttle is opened quickly.
The slide carb used in RC models doesn't really compare to the SU/CV other than the similarity in layout. There is nothing to control the flow to slide position. I understand characteristics are slightly different than a rotary barrel, but not enough to matter. You can open the venturi all the way just the same as a rotary barrel, long before the the engine is actually drawing enough air to need such a large opening. RC helicopter are setup to run a fixed RPM, but obviously need a wide range or throttle setting depending on load. The one thing RC cars and heli have in common is a thirst for nitromethane, which allows more deviation from optimal air fuel ratio.
The trouble with simple venturi is that you are stuck with the pressure drop created by the venturi and volume flow. While the characteristic pressure drop(fuel draw) of a venturi is in our favor (almost linear going slightly richer at higher volume), there is a critical point below which the pressure drop is very non proportional and goes lean very quickly as flow drops. If a carb is so large that it's operating in this nonlinear-region the needles are very difficult to tune, and you may not get enough fuel draw to run at all. Having played with a variety of venturi sizes and throttles on production engines, I can tell you this effect is clear. Small carbs may limit power, but the needle settings are very obvious and fuel draw excellent. Oversize carbs have fuzzy needle setting and are sensitive to fuel level. Muffler pressure is also used to allow larger carbs with production engines. Once you start enlarging the carb to make the most power possible, they get much more difficult to tune and require pumps and regulators. Rotary barrel carbs are different than throttle plates downstream of a simple venturi in that they need air bleeds or extra fuel metering to prevent very rich mixtures at very low throttle. This is apparently due to the very high velocity of the flow in the bore of the barrel even with low volume.
So with some minimum required some pressure drop to generate fuel suction (and sufficient atomization) in the carb, what is the determining factor? Engine air volume flow. It boils down to obtaining a minimum velocity through the venturi. For gasoline in model size engines 120-140m/s (taking into account spraybar area in the venturi) seems to be very reliable with no pump and small suction height(basically at spray bar level with a small fuel tank or float reservoir) assuming flow happens in 180° of the 720° cycle for a single cylinder four stroke cycle engine. I took some time to check these values for a full size application. I picked a Holley carb (rated by CFM) found it’s venturi bore size and did the math. It turns out a 600CFM (.283 m^3/s) carb has a 85m/s venturi velocity. This is low, however, these carbs have double venturi design which essentially double velocity through the venturi in which fuel is introduced. We can also see that if you split flow among two cylinders which don’t draw at the same time, a carb doesn't have to be as large as a single of the same displacement. Specific to model engineering and probable less developed engine designs, meaning volumetric efficiency is low, further venturi size reduction may be necessary. The shaft run/no load condition for display models adds another complication, but further reduction in venturi size(to prevent overspeed) should allow for even better fuel draw. 6mm on a 12cc four stroke twin is only 30m/s so it's no surprise to me that it has trouble.
-
Thank you again Greg,
I haven't done any detailed work as there are too many unknowns, especially leakage rates and the pressure on the bottom of the piston which can be atmosphere one side and manifold pressure the other. I intend to assemble the carb without the jet and needle and then see how the piston responds while turning the engine with the motor/generator.
The cover was held in a 15mm collet and faced. I removed the bulk of the material with drills and end mills and then started boring using a 6mm HSS boring bar. Using the DTI as a guide I slow opened up until the 12mm plug gauge would go in. The 3mm hole was then spotted and drilled 2.4 mm. I then mounted my endmill holder and cautiously increased the depth 0.1mm at a time until it cut. I made a further pass with 0.1mm depth of cut and finally reamed the hole 3mm.
The fits seem good :) The plug gauge takes about 5 seconds to fall out under it's own weight if I block the 3mm hole. I have no means of measuring the concentricity so that I will have to wait until I finish the piston to check that.
-
Next step check that my 3mm silver steel is straight then part off 29 mm for the piston guide. I then moved to the ER11 collet check as the piece was a bit short for the ER25 system. The ends were faced and then one was drilled 1.3mm and reamed 1.5mm for the needle. Next step Loctite the piston onto the guide and let it cure before finishing both parts.
-
Hi Roger,
More watchmaking skills in action. :ThumbsUp: :ThumbsUp:
I recognise the piston but it's a lot smaller than the ones on my MGB
Mike
-
Thank you Mike :) I would put this one 1/5 to 1/6 full size.
I turned the piston to a measured 12.02mm to match the plug gauge but couldn't check it in the bore due to the length of the plug. The plug was turned to 6.00mm as I can bore to suit. When I removed it from the chuck and tried it I got a good fit. It moves smoothly and if I block the two 0.8 mm holes there is definite resistance :whoohoo:
Next up was to mount the body in the 4 jaw independent to make the hole for the throttle and then turn it round ready to bore the holes for the piston and jet.
-
I removed the bulk of the material with a 5mm end mill and then used a 3mm end mill as a boring bar to finish off. Next I faced off the surface where the cover will fit to ensure it was a right angles to the bore. The bore for the jet was more of a problem as I could not see it properly. I started with a2.5 mm drill and then set up a 2mm endmill as a boring bar. I centred this by eye and then fed it in moving the crossslide out in 0.05 mm steps until I got a spiral of swarf to show it was cutting all round. The hole was then finished with a 3mm reamer.
-
Next up was to finish the cover fixing flange. This was drilled and milled with the Proxxon mill with plenty of aluminium snow, promptly cleaned up by Henry. To give an idea of the scale I have included an SU number 5 needle as used in a non emission controlled MGB or with a different spring in my much modified 1300 cc Triumph engine.
-
After drilling and taping a few holes I assembled the piston end. It moved freely most of the way but kept jamming in the closed position :( On close inspection there seemed to be a small ridge at the bottom of the bore ::) I set the body up again in the 4 jaw and took another skim out. The Plug is 6.00mm diameter and the bore is around 6.08mm. This time there was no jamming so I plugged the jet hole with a piece of 3mm rod, blocked the throttle bore with my finger and gave a suck. The piston rose as I sucked but tended to remain there as there in no spring yet :cartwheel: At least it has a chance of working :)
https://www.youtube.com/watch?v=spE33rp4KrQ
The next stage was to make the throttle barrel from 10mm diameter silver steel. The 3mm spigot for the lever was turned and then parted off to 15mm. The cross hole was started in the Proxxon but for the 5.3mm drill and reaming I moved over to the big drill.
-
Roger,
Thats looking good, a replica of an SU. carb in miniature. :ThumbsUp: :ThumbsUp:
What are your thoughts on the fuel supply? A float chamber, or a chicken hopper or a simple clunk tank with the ability to raise and lower the fuel level.
Mike
-
Thank you Mike :ThumbsUp: I am intending to follow the Amsbury route and use a pump and overflow system, but as ever I will see how it works and maybe try other options. If the depression over the jet is more stable maybe the fuel level is less important :headscratch: :headscratch: That's the fun of experimenting :) :wine1:
Having finished the throttle barrel I tried a few more suction trials but noted that the piston tended to rotate so the 0.8mm ports were on the wrong side. Time to make the anti rotation pin. The Amsbury drawings didn't seem to have any form of rotation control :thinking: This was just a stub of 1.5 mm silver steel turned down to 0.9mm at the end. After the initial test I removed it with a magnet (to avoid dismantling it all) and then replaced it with some medium strength thread locker so I could remove it if I had to.
Next up the throttle lever.
-
Now you are really going down in size here and like Mike I certainly follow your experiments (though I haven't got many comments so far). It will be interesting to see how far down you can go and still have it working as intended :thinking: :cheers: :popcorn:
The Amsbury drawings didn't seem to have any form of rotation control
I'm sure you remember how the full size works here (no I have never opened an SU - only the Japanese versions on Motorcycles) .... but I do not see a diaphragm in this model :mischief: so I understand why you need another device to solve this.
Per
-
We have some SU carbs at work I can look at them Monday, if I am at work. If I'm lucky they will already be disassembled. Very interesting project Roger I hope it works.
Art
-
Per, there were two types of CD carbs in common use for cars. The Stromberg version used a thin rubber diaphragm like the bikes and the SU used a closely machined piston in a cylinder. I don't think I could make a diaphragm in this size but a piston is possible.
-
Hi Roger,
I have seen you there working with little Proxxon lathe.
Are you still happy with it ?
-
Hi Achim, yes it does what I want it to do now I have sorted out the collet nut. The supplied 3 jaw chuck was no good however I am thinking of buying some Sherline chucks for it. I also need to sort out a DTI fixing for the crosslide and a toolpost grinder for making needles for this carburettor.
-
HI Roger,
When I got interested in motorcycles in my teens my first bike was a Triumph. The Triumphs had Amal carbs, first Monoblocks then Concentrics. These were slide type carbs operated by a cable. I then got into Japanese machines with Mikunis, cable operated. The thing all these carbs had in common was that the slide had a cutaway on the inlet side of the carb. When modifying a Triumph engine for more power you could actually get slides with different amounts of cutaway to tune the throttle response. On the Japanese carbs all you could do was change the jetting.
Now onto the CV type carbs. They have similar features to their cable operated brethren, cutaway slide, tapered needle with jet, low speed circuit etc. but the distinguishing feature is the vacuum operated slide.
My questions to you are, how are you going to regulate the vacuum on the slide, how are you going to calibrate the needle and needle jet size, how are you going to determine the amount of cutaway on the slide.
I only ask these questions because I like you have always wanted to design a more efficient miniature carb that would function properly. The machining to me is the easy part, the stumbling block was trying to answer the questions I posed without tons of experimentation.
gbritnell
-
Thank you for your interest George. :) I know we have discussed carbs in the past on more than one occasion.
I have attached a diagram of the type of SU I am basing my model on. The pistons (slides) do not have cutaways. The tuning options are the spring for the piston and the needle. There were 2 or 3 jet sizes and needle sets to cope with the range of carbs between 1 1/8” and 2”.
The top of the piston is connected to the manifold depression via the L shaped port(s) the underside is connected to atmosphere by a port or ports in the air cleaner flange (to compensate for pressure drop caused by the filter). The jet can be moved up and down with an adjusting nut to fine tune the mixture. It can also be lowered with a lever to enrich the mixture for starting.
If the manifold depression is sufficient to overcome the weight of the piston and the force from the spring the piston will rise. This opens the venturi and causes the depression to reduce until a balance point is found. Changing the air flow (open or close the throttle) will change the balance point. The needle moves together with the piston and opens and closes the jet orifice to maintain a suitable mixture. The full size version also has a hydraulic damper to aid sudden acceleration. If the throttle is open rapidly the depression will fall rapidly and some of the fuel condenses on the manifold leaning out the mixture and causing a ‘flat spot’. The damper delays the rise of the piston temporarily increasing the depression over the jet and hence the fuel flow.
The set up sequence was to select a suitable spring to suit the airflow range so that most of the piston movement was used. This could be done on a Dyno or be based on experience of similar engines. Next a needle was guessed based on copying a similar engine. The needle was then optimised on a dyno if you were posh or by road testing if not.
That’s the real world so how I am planning to set this up.
First I need to find an suitable weight + spring for the piston. I plan to do this by fitting the carb to the engine with a blank in place of the jet and turning the engine at a range of speeds with the starter/generator and putting small weights on top of the piston rod. A combination of different engine speeds and throttle positions should allow me to find an initial setting.
Second I need to find an initial needle. I can determine from the current carb the jet and needle diameters that it runs with. I will make a needle with that diameter in the middle and a constant taper (10°???). Next I will try and get the engine to start, moving the jet up and down to hopefully find a running position. If it won’t run I can hopefully determine too lean/rich from the plugs. If it will run I can try and determine the optimum jet position for various piston positions by measuring the piston rod height above the cover and the jet height. Knowing the diameter and taper of the needle I should be able to calculate a better profile. Repeat until satisfied.
There are a few more unknowns I may have to deal with. The full size version has a raised bridge across the jet area. I don’t know if I will need one. The weight of the piston will be constant, the force of the spring will rise as the piston rises. I will have to determine if the balance between the two is important.
This is all theory :headscratch: It may work ::) If it doesn’t I will go back to fuel injection as I know I can make that work with petrol (gasoline).
-
Hi Roger,
Thanks for the response. Many years ago when the NAMES show started there was always a booth setup from Strictly I.C. Engine magazine. I had won an award through there program and thus became friends with Robert Washburn the editor. He had a friend, Lee Root, who made some remarkable miniature engines. He also contributed an article to the magazine which provided plans for a somewhat replica of a Stromberg 97 carburetor. This carb had a float chamber, air bleeds etc. I built one of them but it was out of scale for my engine so I never used it other than to try it. The biggest issue was getting enough float leverage to open and close the needle properly.
I spoke to Lee on several occasions and found that he was, to me, rather guarded in providing information about his builds.
Lee had built a 4 cylinder engine that had a throttle body fuel injection system on it. https://www.youtube.com/watch?v=wxsSx7OJ29k The engine ran well and it was intriguing to look into the throat of the carb and watch the minute spray of gas coming out of the nozzle. I questioned Lee several times about how it worked but like I said he was kind of closed mouth about it.
It's a shame that he expertise wasn't passed along but that's the way some fellows are.
I adopted my signature 'Talent unshared is talent wasted' partly because of my experiences with some builders over the years. When you go to meet your maker it doesn't matter how good you were but how you interacted with others.
It would have certainly been a great help for model engine builders to have had some of Lee's knowledge.
gbritnell
-
The next piece was the throttle control. This was mostly a drill, hacksaw and file job based on the carb I made for the 3cc vertical engine. When it was assembled I had another suction trial.
https://www.youtube.com/watch?v=ZIJ_X8WyEco
-
Hi Roger,
The petit SU is coming together very quickly.
How closely do you think your (lung) suction matches the aspiration of an engine? Could you create a more constant suction with Henry the Vacuum cleaner or the inlet of an air compressor?
Mike
-
I don't think I could make a diaphragm in this size but a piston is possible.
That was partly a tongue in cheek and partly a serious comment about the diaphragm - and like you I can't see it made in a much smaller size as the thickness and maintaining strength and flexibility get really complicated. The original diaphragms are rubber casted around a woven piece of "cloth" .....
George - I'm sure that the other slides etc. for the Japanese carburetors were made - but I wouldn't have a clue about how to get them outside Japan .... oh and I do know this from the original workshop manuals - where I can see that the US models has different parts (not only Jets) to those here in DK, OZ, etc.
George made some valid questions and I have thought a bit about it ....
It is possible to get accurate flowmeters for the air in the required amount (the fuel is almost more in Chris old line of work with pico liters) - but is it also in a reasonable price range ?
Fuel in an experiment could be measured with a small diameter fuel hose filled all the way to the Jet and an open end (not in fuel). Have a constant airflow empty the line and measure how long it takes ....
Repeat experiment with many different flowrates and plot into a graph.
This would be simpler if one has access to some automatic measuring equipment (I presume that automakers (or fuel injection companies) has something along these lines ....
Per
-
Thank you all for your interest :) :ThumbsUp:
George,
I did try and make a float chamber for my first engine but as you say getting enough force on the needle to close it was a problem. ETW noted the problem and recommended using a lever system to increase the force on the needle.
I found a couple of pictures of Lee Root’s engine. There seem to be two electrical connections to the injection unit, one to what looks like an electromagnet and one to what is probably a throttle position sensor.
Looking at the video the fuel pressure pump seems to be separate from the engine, the fuel connection is on one of the engine mounts.
If I was doing this I think I would make some sort of electromagnetic shuttle valve that would deliver a defined quantity of fuel per pulse and then control the fuel quantity by varying the pulse rate.
I agree about sharing knowledge and always try to do so. This usually results in me sharing how much I don’t know ::)
Mike,
I think that my suction is probably much higher volume but lower pressure than my engines, Henry will be much the same. To avoid having to quantify this I will use the engine being driven by an electric motor as the test vacuum source.
Per,
My initial set up attempts will have to be much more practical. At these small sizes the measuring devices will have a significant disturbing influence on the system, suitable ones are also not cheap. A basic differential pressure instrument like this would probably have too much internal volume.
http://www.drucksensorik.org/products/en/pressure-gauge-HMG/HMG1/HMG1-1000-mbar.html
Small volume U tube manometers will probably be influenced too much by capillary forces.
-
Hello Roger,
I know this topic is about CD carbs but you touched on PI in your last post
If I was doing this I think I would make some sort of electromagnetic shuttle valve that would deliver a defined quantity of fuel per pulse and then control the fuel quantity by varying the pulse rate.
You may already know that the diesel garage heaters, discussed of the forum awhile ago, have such an delivery pump. The dosing pump on my heater delivers 0.02ml per pulse. At the lowest heat setting the pulse rate is 1.6 HZ which delivers 0,1152 litres per hour to the spray jet in the heaters combustion chamber. This is still much higher flow than your engine would need but these dosing pumps (made in China) are/were cheap. Cheap enough to take apart, copy or modify, reduce the stroke, perhaps. Sorry to distract you.
Mike
-
My initial set up attempts will have to be much more practical.
I thought that my idea using a clear fuel tube with an internal diameter off perhaps 1mm. a meter long or so would be cheap, easy, depentable + easy to calibrate - was just such a device, for measuring the fuel use at a give flow through the carburator. With such a small diameter, the fuel will not back out after opening the end with the sucktion from the carb and measuring the time it takes to empty => you will get an accurate measure.
I will admit that measuring the air flow is a lot more complicated - unless you can fill the air sucked out off the carb into a "ballon" for an exact time too.
There are flow meters that has no moveable parts, but likely not cheap - though modern flowmeters for measuring hot water for heating appartments, to calculate the size of the bill are cheap. They can work a number of different ways - one being Ultra Sound.
Sorry I'm just thinking out loud ....
-
Hello Roger:
One of the tools we use in two stroke boat racing is a powered flow meter to "guestimate" the needle valve opening based on the load (prop pitch and diameter). Initial runs are performed and tuned by ear to get proper needle settings, and when one likes the power curve and after the boat returns to the bank, the carburetor/needle valve is connected to this meter and air is supplied by twin pumps, through a blood pressure gauge, and powered by a regulated voltage supply. This establishes a base line for subsequent runs. Minor needle tuning is performed based on the atmospheric conditions for the time of day etc. The difference with this meter and say a mass flow meter is that the associated number one obtains from their setup is based on the back pressure of the system. Accurate(?) and reproducible flow rates can be established based on orifice size vs. mmHg.
One question I have had reading through this and other posts is, does everyone run the same fuel in their engines and has any consideration been given to the various stochiometric variables involved with each type fuel?
Please excuse any ignorance on my behalf.
Best Regards...…….Rick
-
Lots to think about :thinking:
Mike,
I have done some rough calculations based on the Root engine:
The Root engine is around 6ml per cylinder. The density of air is around 1.2kg/m3 so the air charge will be 7.2mg. For a mixture of 14-1 this will require a maximum 0.5mg of fuel. The fuel density is around 0.8kg/dm3 = 0.8mg/mm3 so ~0.6mm3 of fuel will be required per stroke.
This is a 4 cylinder 4 stroke so a maximum 1.2mm3 of fuel per rev will be required. A combustion heater pump delivers 0.02ml = 20mm3 per stroke so one pump stoke will supply the fuel for 16.7 revolutions of the engine. At 6000 rpm this would require 360 strokes per minute or a rate of 6Hz for full output. At idle this would be significantly less, maybe 20-30 strokes per minute. With a small accumulator between the pump and the injector nozzle to smooth out the flow and a reasonably large inlet manifold to even out the mixture that might work.
The control system would need to take into account rpm (not neccessary with my mechanical system), throttle position and ideally manifold pressure. I think I will leave that project to someone else ::)
Per,
I'm not sure I actually need to know what the actual flows and pressures are :headscratch: The load on the piston just has to match the engine the carb is fitted to which as long as you can turn the engine at operating speed with an external motor should be ok to do. The fuel flow also just has to match the engine and using the moveable jet it again should be feasible to find a working needle profile. It's the difference detween making a piston to fit a cylinder by trying it, boring/laping it a bit and trying again and accurately measuring the piston and cylinder to achieve the desired tolerance.
Rick,
An interesting technique :ThumbsUp: :ThumbsUp:
There are several fuel options all with diferent requirements. I am running pump petrol (gasoline) or one of the alkylate variants. Some people add a little oil on four stroke engines. Two stroke engines have a significant proportion of oil so more fuel is required. Glow plug engines use alchohol as the fuel basis.This again requires more fuel in relation to petrol (gas). I have run my engines on cooking alchohol instead of petrol which requires a significant opening of the needle valve/increase in injection pump stroke. Model compression ignition engines also have an additional proportion of ether.
I would guess that for nominally equivalent size engines a glowplug two stroke would require twice the fuel flow of a four stroke spark ignition engine.
Keep the thoughts coming :) :wine1:
-
Mike,
I have done some rough calculations based on the Root engine:
The control system would need to take into account rpm (not neccessary with my mechanical system), throttle position and ideally manifold pressure. I think I will leave that project to someone else ::)
Keep the thoughts coming :) :wine1:
I thought of the diesel burner pump as the metered fuel supply to your mechanical high pressure injector pump. But as you say, the development of a multi parameter, computer controlled ECU is best left to someone else.
I think you are heading in the right direction with your self regulating, self metering CD carb experiments.
Got to it.
Mike
-
Good Morning Roger:
"I would guess that for nominally equivalent size engines a glowplug two stroke would require twice the fuel flow of a four stroke spark ignition engine."
You are correct. The amount of oxygen in the alcohol and nitromethane fuels replaces the amount of air required to about a 5.7:1 A/F as compared to a 14.7:1 for gasoline. Also, if any other petrol mix is used, such as E85, will require recalculation of the A/F ratio. Synthetic oils are used in the 8 to 17 % range and contain no usable burnable compounds. Of course, the less oil, the more oxygen = bigger bang. And that leads sometimes to more parts. :facepalm2:
I think you are doing a great job in this research. Following along.
Regards...…Rick
-
I'm not sure I actually need to know what the actual flows and pressures are
True to a point or seen another way - it's a question of from which angle you start to solve a problem with many unknowns ....
I can certainly see the idea of using an external motor to turn the "real engine" to the desired RPM. That could be used to make the Vacuum Piston open the whole way and check that it goes up and down as desired with the RPM's as the first thing on the list.
To me it will then be interesting to know / measure how much Air and Fuel goes through at those RPM's in order to see how much I'm away from the mark that would make it run - but I can also see the point in using an external Fuel metering device as found in many of the latest model airplane engines and just adjust it to make it run as it should @ max RPM first. That would allow me you to do it the next steps as you would have done in our youth :old:
-
To me it will then be interesting to know / measure how much Air and Fuel goes through at those RPM's in order to see how much I'm away from the mark that would make it run - but I can also see the point in using an external Fuel metering device as found in many of the latest model airplane engines and just adjust it to make it run as it should @ max RPM first. That would allow me you to do it the next steps as you would have done in our youth :old:
Per, can you provide a link to these modern external fuel metering devices you refer to? It is easy to loose track of the latest airplane engine developments
Mike
-
Sorry Mike but I haven't seen them for model engines.
The reason I kind of know about them are twofold. One from an old proffession of mine, as I in my youth worked for Navitronic - a company that specialized in equipment for making Seamaps, amd in order to make acurate soundings you need to know the Speed Off Sound where you're measuring. We (among many other products) made a "Fish" that was lowered into the water and it measured the SoS + temperature for every fathom (aprox 2m.) you lowered it. One of my old colleague that I talked to many years later told me that he was approached by Kamstrup who makes some very advanced Joule Meters today. These meters use some of the same priciples, but much more advanced and they are placed around the heating water tube and the return colder water tube. This allows them to measure how much energy you took out of the system and wirelesly report that to the supply company as the measure temperature and flow amount down to centi liters.
Some off the other informations are from old "Das Motorrad" magazines I don't have anymore and Internet sources I have forgot, but mostly online Motorcycle Mags and their linked tech info + some YouTube videos - again I do not remember the exact source.
But as mentioned - I would use the trick with the clear fuel tube for test purposes. I got the idea from seeing airbubbles in the fuel lines.
Per
-
Next up was the jet sealing block. This could have been cut from a piece of 3mm plate or turned from some 10mm square bar. As the bores were important I decided on the latter. The seating for the O ring was bored, one again, using a 3mm end mill. As I had to remove the cover and piston to drill the fixing holes I have added a picture of the anti rotation pin. I followed Amsbury's thoughts and didn't turn a register on the body for the cover, a mistake :( , for the next body there will be one. I put it all back together including the fun of re-centring the cover and fitted it to the two cylinder engine.
-
I set the engine up with the starter battery to try the carb. It cranks at around 2800 rpm with 12V at 11A and the carb piston responded well :cartwheel:
https://www.youtube.com/watch?v=VXN49TcK34Y
To try the full rev range I will have to borrow 24v ~20A power supply from work but looks good so far :)
I measured the needle of the needle of the carb I have been using to get a base point for this needle. The jet is about 0.5mm (drilled) and the needle was around 0.42mm.
-
Hi Roger,
This is getting exciting. Can we fast forward?
Mike
-
Unfortunately you may have to wait a while with my rather slow rate of progress :( I still have some new injectors to build for the diesel ::)
I made a selection of brass weights to go on top of the piston guide rod to try and get an idea of the additional load requires. The biggest, 13mm diameter calculated weight 6g, was going in the right direction. I borrowed a chunkier power supply from work (spare focus magnet power supply for a 3 million volt electron accelerator) so I could try higher rpm. I got up to 5000rpm where the 6g weight was not enough, the motor was also getting rather hot after 10s of seconds running at 25V 17A, over 400w.
https://www.youtube.com/watch?v=GoPOQdOmHI8
https://www.youtube.com/watch?v=8c9za4AgkR0
I cannot fit a weight that size on top of the piston, I would have to remake the piston in brass. I noted that Amsbury made his piston from mild steel :thinking: The other option is a spring. I couldn't find anything standard but Lee springs calculator came up with a possible design. Maybe I will have to look at winding my own. Once again the 0.355 wire in the Lee Spring's design is close to the 0.4mm wire in the Amsbury design.
-
Great result so far on the first tests and thanks for the videos :ThumbsUp:
I know that you're mostly testing - but a spring works in any direction - so I would prefer that if I was you ....
You never know if you suddenly want to have a future vehicle "bumping down the track" so to speak.
:cheers: :popcorn:
-
Thank you Per :ThumbsUp:
I searched further through the Lee Springs list and found a better option with a rate of 0.02N per mm. The 6g additional weight delivers a force of 0.06N so a compression of 3mm will give that and there will be a rising rate to 0.18N when fully open.
I also think that the bridge under the piston on the full size carb is important for the idling/low speed performance as it defines the venture better than the bottom of the bore. The only sensible way I could see to make one was to bore right through and then make a suitable shaped plug and fix it in place. Boring through is also much easier than a blind bore as will be keeping the bore for the jet concentric.
As I was going to bore through I also decided to turn a short spigot to centre the cover. Amsbury's idea of centring by hand each time is a pain. There was just enough material to make a 0.5mm high spigot without breaking through the bores from the input end. Measuring was interesting but possible. I had a couple of problems facing back out, hence the two dig in grooves but I think I need to redesign the whole top of the body. The cover was a good snug fit so I drilled out the bulk of the material for the bridge plug with a 5.5mm drill and then bored in 0.05mm steps until the tool just touched the existing bore.
It now all assembles easily without any jamming :)
-
Hi Roger,
I too believe the bridge under the piston to be an important detail for the low end performance. The proposed insert plug looks like the most sensible way to implement it. It's attention to these fine details that will help you achieve your objective
Have you had any more thoughts about the virtues of an adjustable needle in the piston verse the adjustable jet in the bridge. I believe the adjustable needle has a lot in its favour
Stay in, Stay safe
Mike
-
Thank you Mike :ThumbsUp:
I'm not sure how to install an adjustable needle in the piston, the guide rod is 3mm diameter so I would be working with M2 or M1.6 threads and I can't see how to lock that against vibration :headscratch:
I have started on the bridge plug. Machining the two ramps at 25° required some thought but I came up with a solution of clamping my small Proxxon RT in a vice at the required angle and then turning 180° for the other side.
-
Thank you Mike :ThumbsUp:
I'm not sure how to install an adjustable needle in the piston, the guide rod is 3mm diameter so I would be working with M2 or M1.6 threads and I can't see how to lock that against vibration :headscratch:
Hi Roger, A few turns of cotton wrapped around a M1.6 thread would be sufficient to seal and provide anti vibration. If you were to have a movable needle as well as a movable jet, you could see which one gives the better mixture control. I always felt the mixture enrichment by dropping the jet on the SU was far to crude. It really only provided lots of extra petrol during a cold start, rather than an enriched mixture. Screwing a needle in and out must give better control.
Stay in, stay safe
Mike
PS that bridge looks just right.
-
Hello Roger:
Attached are the adjustable remote needle valve assemblies I use in rc boats. They come in two sizes, course and fine needles. I'm not sure if you could use anything like these but they work very well for adjusting lean or rich settings for my application.
Regards....Rick
-
Thank you Rick, those are interesting. Where do you get them from? I don't think they will help me directly in this carb but they are good small control valves for controlling the flow to a overflow type fuel feed when I get that far.
Mike, I think that if I can adjust the jet height with a fine thread (0.5mm pitch or less) it will be easier that putting a hex key or watch makers screwdriver down the piston guide. The Stromberg CD carb had an interesting mixture enrichment for starting. There was a rod with a flat on it as the manifold side of the bridge which was rotated to lift the piston and increase the fuel orifice as well as restricting the air flow.
-
Mike, The Stromberg CD carb had an interesting mixture enrichment for starting. There was a rod with a flat on it as the manifold side of the bridge which was rotated to lift the piston and increase the fuel orifice as well as restricting the air flow.
Roger, That is a clever way to do the enrichment. Does it look feasible in model scale, the lift would only be 1mm or so. May have a problem with air leaks at the side of the rod, where it goes through the main body?
This way sounds a better way than the SU way of dropping the jet down inside the bridge, which gives poor atomisation.
I think that if I can adjust the jet height with a fine thread (0.5mm pitch or less) it will be easier that putting a hex key or watch makers screwdriver down the piston guide.
In my minds eye, I saw the adjustable needle going all the way down the 3mm diameter x 29 mm central rod in the piston. The adjuster knob would be outside at the top.
Stay in, stay safe
Mike
-
Thank you Mike, I now understand your ideas about the adjustable needle :ThumbsUp: I shall give it some thought :thinking:
Here is some more information on SU and Stromberg CD carbs for those who may be interested:
http://sucarb.co.uk/technical-su-carburetters
http://zenithcarb.co.uk/zencdcdsseries/?___store=zenith
-
I decided to use some medium strength Loctite and a grub screw to retain/seal the bridge plug in case I need to change the profile.
Following the original route while I think about the adjustable needle I started on the jet block. The was made from a 6mm slice of 12mm square brass. On the first attempt the while drilling for the jet the drill dug into the hole for the petrol feed union and gave an oval hole. For the next attempt I finished both holes with a 3mm 'throwaway' milling cutter.
-
A quick check that the stop screw clears the jet seal block and then on with the pivot support and jet adjustment lever.
-
The next step was to try and make a needle. I had scaled to 0.5mm diameter from the Amsbury design but felt it might be a step too far. The tapered part needed to be around 8mm long, 5mm for the bore and then 3mm to allow some jet drop. Starting with some 1.5mm diameter silver steel I tried to turn it down to 0.5mm. Using a new aluminium insert in the Proxxon lathe at 3000 rpm the first 0.2mm cut was ok, the next at 0.1mm failed :(
I then set the slide over at 1° (I don't know if this is a realistic angle or not :headscratch: ) and tried again. I took a first cut at 0.25mm and then dropped to 0.05mm. The workpiece was visibly deflecting so I reduced the cuts to 0.025mm (It did cut and produce some hair like curly swarf). I got down to 0.5mm but the deflections meant that the needle was almost parallel :(
Maybe 0.5mm is not practical so I tried again to produce a 0.8mm needle. This was cut using similar steps to before to reach a starting diameter of 0.8mm. The tip was 0.6mm which calculates as an included angle of 1.4° rather than 2°, probably due to deflection.
That was enough difficult bits so I decided to make the fuel connection nipple for the jet block. This was silver soldered in place as I expect I will have to change the jet more than once so that will be soft soldered to allow easy removal.
-
The final result looks good - but I'm sure it was hard work to get down to so small dimensions :ThumbsUp:
I can't help wondering if it almost requires a tree sided cutter ala some thread dies to avoid deflection :noidea:
-
Roger,
Could you turn the tapered portion between centers then cut the end off, leaving the tapered portion.
Art
-
The old trick for carb needles is to repurpose sewing needles. There are various types of sewing needle and sizes to choose from you should find something suitable.
If you wish to 'turn' one then you are in the watchmaker's turning domain: Supporting the needle in a groove or hole in a jacot tool from the tailstock then using a very fine needle file and a burnishing tool to hand shape the point. I'll see if I can find the details of how to do this :thinking: ...
Jo
-
If you don't want the bother of making or buying jacot tools then a piece of wood or MDF with a Vee filed into it held in th etool post will support the work while you file and then stone the taper. If you look at Ramon's builds you will see him doing this which is where I got the method from. Though most of mine are just turned on the chinese "watchmakers" machine ;)
(https://img.photobucket.com/albums/v156/jasonballamy/Engineering/5cc/HPIM1403_zpsckkual3a.jpg)
-
Thank you all for your interest and suggestions :ThumbsUp: :ThumbsUp:
This was a first attempt to see what I could do the 'easy' way. I have similar issues grinding the injector needles for my fuel injection system and noted there that I need an additional support.
A point on the end of the needle is no problem so possibly a small inverted centre would work. If I can taper the outside down to 0.5mm diameter it won't interfere with the lathe tool. Another option is form of fixed steady mounted on the saddle again possibly with a spring loaded inverted centre.
These needles with have a defined profile so they will need to be turned or ground (I also need to add a DTI to the cross slide).
-
:thinking: That support looks familiar.
Jo
-
You seem to be making some good progress. For ease of adjustment I recall needles need to have a finer taper than those we are used to seeing on glow engine carbs. As glow fuel is run around air /fuel ratio 8:1 (?) and of course petrol nearer to 15:1
-
:thinking: That support looks familiar.
Jo
:) :) :wine1:
-
I made some pivot pins for the jet adjuster and then made a jet. I started the hole with an extended 0.5 mm drill from my fuel injection work and then finished with a 0.8mm drill which I assume will be oversize.
Next step was to make a tailstock inverse centre and try another needle. The concept worked but I need to reduce the size of the centre so I can get the end of the needle to 0.5-0.6mm.
-
That is some fine work - and yes, pun intended. Fine in every sense of the word!
-
Thank you :ThumbsUp:
I thinned down the inverse centre and had another attempt. I let this be oversize to check the actual jet bore. It appears to be around 0.83mm so I will have to design the real needle around that. The needle seems to have a similar taper to the first one so deflection may not be the problem. The inverse centre certainly allows bigger cuts. I check my precision cheap angle gauge and zero is not zero :( Time to spend money on something better ::)
-
I remember buying a large square at the hardware store for woodworking projects. As warned by a friend ahead of time, I checked them in the store - one edge against the side of the shelf to my right with second side across the shelf - light pencil line, then flopped the square over so the first edge was over on the left, looked at the line again against the second edge.... went through the bin till I found one that actually WAS square! Amazing how far off some of them were.
-
I have a better angle gauge on order and thought I should also check my collection of squares. They were not as bad as expected being cheap imports, but one had a tapered blade. The inside 90° was almost spot on, the outside was way out, ~0.5mm in 70mm.
Getting materials is starting to be a problem :( I had identified a suitable piston spring from Lee Springs but they won't supply to private customers at present. I asked in the toolroom and work and they suggested a Swiss supplier who also had something possibly suitable but they also won't supply to private customers ::) In this case I can order through work but it will take a little time. I found a supplier of small fasteners in the French part of Switzerland who appear to be able to supply some of the small screws I need, especially some M1.6 x 2mm grub screws to hold the needle in place. I wait and see.
As I can't move on with the carb at the moment I went back to the fuel pump and started on the valves. The Regner unions I usually use have a bore of 2.5mm which is a bit large for the seating of a 3mm ball so I made a copy with the bore reamed 2mm. The seat was finished with a 3.2mm drill ground flat. A quick suck test showed a good seal without tapping the ball to get a seat. I then pilot drilled the ports in the body and milled flats for the valves to seal on.
-
As I was in milling mode I moved onto the ball and socket for the diaphragm rod of the pump. The seat was cut with a 3mm radius end mill and the form tool was roughed out of key steel with same cutter. I filed some clearance and then heated to dull read and quenched it. The ball was cut from 6mm hex and taped M2.
The next step was to finish the valve seats in the pump body. The inlet was just drilled 4.5mm and tapped M5 x 0.5 as the seat is in the union. The outlet was reamed 2mm and the seating was cut with a 4mm end mill.
-
I received my new protractor which is somewhat better than the old one. It has an additional arm that can be fitted for measuring dovetails and similar. I found that with careful adjustment I could measure the needle directly :)
I finished off the pump arm's various holes and slots and will then move onto the rest of the piston rod parts.
-
The clamps for the diaphragm were made from 6mm brass rod and tapped M2 and parted off to 2mm long. The piston rod is 2mm diameter stainless steel threaded M2 both ends. I put it together with some M2 bolts as temporary piviots to make sure nothing jammed.
https://www.youtube.com/watch?v=65RV1i12yf4
Next up is a bracket to fit it on the Proxxon lathe for trials as if driven by a camshaft I can expect 3-4000 rpm. This is just hacksaw and file work on a piece of 30mm aluminium angle.
-
Next up I milled the valve stops from the piece of brass I turned earlier. For ball valves the diameter of the opening and the ball lift are fairly important. The bore should be around 0.7 times the ball diameter and the lift should be around 0.1 times the ball diameter. My bores are 2mm for a 3mm ball so ok. The lift needs to be around 0.3-0.5mm. The stops are to be cross forms to be pressed into place. The 3.3mm diameter piece was mounted in the Proxxon RT and milled 1mm x 1mm at 90° intervals. This will be parted off in 1.5mm pieces and pressed into the unions/cages.
I turned up a trial excentric from a piece off 12mm dia. offcut by offsetting in the self centring 4 jaw using two 2mm thick packing pieces to give around 1.5mm offset for 3mm max stroke. I was a little concerned about parting off into an offset bore but there were no problems. With a quick check nothing broke at 5000 rpm :)
-
I received a few bits, some tiny grub screws and springs. The smaller spring barely moved with the largest of my test weights but seemed reasonable with a quick breath test. Lots more to try :thinking:
https://www.youtube.com/watch?v=TY6A0YuYrGY
-
I cut some 1.5mm slices from the valve stop blank. One was pressed in above the ball for the inlet valve and the other into the union to go above the delivery valve. Both balls were given a light tap with a piece of brass rod to seat them. I gave the pump a quick trial with water and all looks good. It sucked around 30mm and delivered consistently :)
https://www.youtube.com/watch?v=oD62DQ5g-9E
-
That is one itty-bitty set (grub) screw!!
-
I drilled and tapped the piston M1.6 for the needle retaining screw and assembled the carb with the first 0.8mm needle. Fit the tank with the extension plate to get the fuel level near jet height, a bit of alcohol, press the button and away :whoohoo:
https://www.youtube.com/watch?v=JJBTzttLoE0
Still a lot more to do but at least I have a base line. The spring is obviously not strong enough, the needle tends to jam in the jet if it is raised too far and it will respond differently with petrol.
-
Hi Roger,
That is a fantastic result for the first run. Congratulations on your success :ThumbsUp: :ThumbsUp: :ThumbsUp:
Now the fun of tuning the individual components can begin. But at least you have a working baseline that you can always return to.
It's interesting to see the needle/ dashpot jumping up and down, I had not expected that. But you can understand why the SU (and also the Stromburg??) has the oil filled damper.
A just reward for your faith and effort
This storm cannot last forever. Stay strong.
Mike
-
Outstanding work Roger.
-Bob
-
Thank you both :ThumbsUp: :ThumbsUp:
Mike, I think that I could open the top of the piston rod out to 1.8 or maybe 2mm and make a simple damper. The problem would be finding a low viscosity fluid, possibly diesel fuel.
I also need to find a place to fit the fuel pump which was not in the original engine design :thinking: It needs to be driven at camshaft speed and probably from the timing gear case. The other end of the camshaft is too near the exhaust. I have a rough design for an excentric to fit over the boss on the gears which I planned to use for the fuel injection pump trials on this engine which I think I can adapt :headscratch:
-
Roger,
I used '3 in 1' oil in my full size SU carbs during cold weather
Mike
-
Hi Roger,
really a fantastic progress.
Back in the late 1970 I have owned a Volvo with this type of carbs.
I cant remember well, but I think it was SU.
The only thing I can remember very well, it was a horrible nightmare to find a proper adjustment.
-
Thank you both :)
Mike, I will have to think further about a dashpot and oil :thinking:
Achim, SU's were very sensitive to wear and hence air leaks on the throttle spindle which upset the idling. Fixed jet carbs had an auxiliary jet by the throttle for idling that reduced this problem.
I had another play with a stronger piston spring. The results were better in some respects but showed up some other problems. The piston tended to stick at the lower and upper limits. I think that I will have to start again with the build and make the following changes:
-The top will have to be flat so I can machine a proper register for the cover. This will also lengthen the piston so I can increase the clearance without increasing the leakage too much. There are rubbing marks with current set up.
-The surface and bores will be bored together rather than in several stages.
-I may reduce the throttle diameter to 8mm so I can move it up and give a greater bearing length for the jet.
-The cover will have to be lengthened so the piston won't hit the end of the bore when it vibrates/bounces.
Lots of fun to have ::)
https://www.youtube.com/watch?v=9hMxHmUUqBo
-
Hi Roger,
You have good lift and response from the piston, therefore it appears to have more than enough 'seal'. Could you get away with a rub of wet n dry on the high spots to reduce the rubbing?
The motion of the piston is very regular, does it coincide with engine speed? Could the siamesed ports for cylinders 2 + 3 be having some effect?
One last thing. What effect would some finger pressure have in damping the piston motion?
Stick with it, you may be closer than you think. :ThumbsUp: :ThumbsUp: :ThumbsUp:
Mike
-
Thank you Mike :ThumbsUp:
There are too many compromises in this build as I have changed several things so I will start again and hopefully retain a better alignment/concentricity.
This is a twin with a 180° crankshaft so the pulses are not even as they would be on a normal twin carb 4 cylinder engine.
The first trials were far better than expected so I will keep going as I will with the diesel ::)
Currently work is busy/overloaded and we are also supporting our daughter who is home schooling the three grandchildren so progress may be slow.
-
While sorting out the design improvements for the carb I thought I would test the pump. It delivered the hoped for 0.7 Bar (10 psi) against a blocked end but the flow is excessive. I can adjust the speed of the excentric and the stroke (by moving the cross slide). At 1000rpm and 1mm stroke it was filling my 10cc measuring cylinder in less than 5 seconds. At 3mm stroke it would make an excellent coolant pump.
https://www.youtube.com/watch?v=vTN50baq4yI
-
I have started to build the mark 2 version. A lot of the operations are the same as the first build so I will mostly show the changes. The cover was turned as before using the plug gauge and the original piston fitted and sealed well :)
The original jet adjustment mechanism was too big and in the way so I have redesigned it with a smaller jet block and the adjusting lever will be pivoted under the air intake which required two 2mm slots for the links These were cut using the Proxxon mill at 11000 rpm and 0.3mm depth of cut.
The throttle barrel bore was drilled 9.8mm and reamed 10mm as before. The piston bore was drilled 5.5mm, bored out to around 5.8mm and them reamed 6mm. The top of the body was faced to give a 12.5mm diameter 0.5mm high spigot which was then finish turned to the same diameter as the plug gauge.
-
The cover is a good fit on the spigot with barely perceptible play :) Next up was the piston and I have changed the sequence slightly. The main aluminium part was turned to 0.5mm over finished diameter and drilled and reamed 3mm for the rod. I moved to the ER11 chuck to drill and ream the end of the rod 1.5mm for the needle. The rod is a bit short to clamp well in the ER25 collet. The piston body was then Loctited onto the rod and left to set. The plug part was then finished to 5.90mm and the main piston to 12.02mm. This worked smoothly in the cover and body and sealed such that it wouldn't drop under it's own weight :whoohoo:
-
As I had the ER chuck fitted I made the jet from a 12mm piece of 3mm brass rod drilled 1.5mm for 10mm and then a 0.8mm orifice.
The piston rod was clamped in the Proxxon RT for drilling and milling. In the first version these operation were done before the body was Loctited to the rod for final turning. I think that the interrupted cut due to the anti rotation slot together with the thin, 3mm, rod resulted in the plug part not being round.
Unfortunately I broke the 0.8mm drill in the first suction hole. This is 12mm deep so a bit of a problem ::) As this was the first hole I could just move everything round 60°. I found that I needed to withdraw the drill every 1 - 1.5 mm stop it and remove the swarf from the flutes with my fingers. The M1.6 thread for the needle clamping screw as well as the anti rotation slot and the two suction slots went without problems.
-
Happy to see you back at it Roger. This is a great project.
-Bob
-
Thank you Bob :ThumbsUp:
The cover fixing holes and flange were done as before except my CNC (count number of cranks) lost count at one point :( I then worked through the various other fixing holes. The top fixing for the throttle stop was only 0.5mm from the edge and an image of the threads appeared on the top surface ::)
-
I was talking about this build to some work colleagues and they were asking about the precision and clearances I was working to. I said I really cannot measure them with what I have but I thought the piston clearance was less than a hair. I duly plucked one of my remaining hairs and it jammed the piston completely.
-
If nothing else - it is certainly a reference all people can refer to :Lol: :ThumbsUp:
You're still making good progress as far as I can tell and I'm still following and enjoying :cheers:
-
Thank you Per :)
Next is the jet assembly. This is much as before with slightly more clearance for the O ring and a much smaller jet block.
-
The 2mm slot for the control lever was milled in the end of the block and the jet and nipple were soldered in place. This time I decided to solder both rather than silver soldering the nipple and then soldering the jet. This was possibly the wrong choice as it was hard to keep both aligned. In the end I kept the jet true and let the nipple droop a bit.
Next the blanks for the control lever links were sweated together then drilled and reamed. I made a couple of filing buttons to round the end from 4mm silver steel but must have over hardened them as they fell apart when removed ::) I made another pair which survived. These are definitely a swarf gnome favourite, I think I spent more time on the floor looking for them than I did in making and using them :toilet_claw:
-
The two links were separated and the operating lever was made from 2mm thick aluminium. A few pivot pins were turned from 2mm silver steel and the anti rotation pin was turned from 1.5mm silver steel and Loctited in place.
Next up I need to make a few more trial needles ::)
-
The Mk2 version is looking so much neater than the prototype. :ThumbsUp: :ThumbsUp: :ThumbsUp:
Is it still being made to the basic dimensions and drawing you posted at the start of this topic?
Stay safe
Mike
-
Thank you Mike :)
Yes, only the details have changed. I have attached the current drawings as PDF and ACD R12 (I think you can read that version?)
-
Before I could start on needles I had to try and measure the jet. Using a previous oversize needle and measuring how far it would enter I came up with almost spot on 0.8mm. Not bad for a drilled hole. I could then confirm that the needle used on the last trials was slightly too big for this jet and would jam.
Next I used my new angle gauge to set a hopefully more accurate 1° and turned a new needle (without tailstock support) to a large diameter of 0.79mm. The small end was measured at 0.55mm giving an angle of 1.7°. Given all the potential setting and measuring errors not bad :)
Now I need to assemble it all and try it on the engine.
-
I started to make another needle with 0.75° (1.5° included angle). This one did not turn nicely and I couldn't part it off, it just kept wrecking my narrow parting tool :( :headscratch: I tried cutting it off with a junior hacksaw and that just slipped :toilet_claw: it appears that my new batch of 1.5mm silver steel (purchased some years ago) is off spec. It is hard, a file just slips, and is oversize, ~1.505 mm, definitely not h7.
I annealed a piece (heat to dull red and air cool) and it became cutable but the accuracy is not there. I have ordered some more and will have to wait. 1.5mm h7 or 8 is not easy to find in small lengths. 20m would be here tomorrow :ThumbsDown:
-
Did you just write 20 meters ? If so, I will agree on your thumb down :o
It's never nice when you discover that your source material isn't what you expect :rant:
-
I looked back in my records and I purchased 3 lengths of 1.5mm silver steel in 2017. Two were fine but I think that the third length was Piano (Music) Wire. I received some more, correct, material from another supplier and remade the needles. To the accuracy of my measurements I am seeing very little deflection even without a tailstock support. The 1° included angle needle was 1.0° as far as I could measure. Time for some trials.
-
I made a quick trial with the new carb to see if I got similar results to before. Seems ok :)
https://www.youtube.com/watch?v=Lm0jonOZEqY
To start the actual calibration I need to have a constant fuel level so on with a smaller diaphragm pump and the level control tank. This pump has a diameter of 9 mm instead of 12 mm so nearly half the working volume. I am making the body square rather than round which will allow more flexible installation.
-
Hello Roger,
I've moved this post to this thread
I believe Perry make a miniature diaphragm pump for model aircraft engines. It's a bit spendy to buy one just to see if they are any good.
BUT, I do know the diaphragm is not crank driven. Instead, they have a small weight which is vibrated by the engine and this appears to be enough to pump the pump and give small enough delivery rates..
Is this idea worth an experiment on your engine test rig. :noidea:
Stay safe
Mike
-
Thank you Mike :ThumbsUp: I think that would work to feed the constant level tank. As I understand Walbro (??) also use the manifold or crankcase pressure changes to generate a pumped feed. I am developing the diaphragm pump to reach around 0.8 bar as a feed for fuel injection which I am not sure the other techniques would meet :thinking: May be crankcase pressure on a 2 stroke single could come close :headscratch: :headscratch: too much to learn ::)
-
Finally back in the workshop for a bit more work on the fuel pump. The clamping ring was made as before and the pivot was turned from 4mm brass hex using a ball turning attachment. The pivot block and the bell crank were roughed out with a hacksaw and finished to size on the mill.
-
A bit more on the operating mechanism. Most of this was done on the small Proxxon machines.
-
A little bit more on this one. I had originally planned to use 1.5mm stainless steel for the pump pull rod with M1.4 threads on the ends. This proved hard to find. I could have turned 2mm down to size but found an alternative. 1.6mm 316L is a standard TIG welding filler rod so I scrounged one from work and increased the thread size to M1.6. This machined fairly well for something optimised for a completely different purpose. I made a new pivot and locknut from 4mm hex with M1.6 threads and so far it all looks good. Next up are the clamping discs for the diaphragm and the diaphragm itself.
-
Next steps with this one. The clamping nuts were made from 4mm hex and tapped M1.6. A guide bush was made for the central hole in the diaphragm followed by a 1.6mm punch. The diaphragm was then cut from 0.5mm thick Viton sheet as before.
-
-
-
Looks like the punch makes some very sharp and well defined holes in the diaphragm :ThumbsUp:
-
Thank you Per :ThumbsUp: I have found that these very simple punches work well for gasket material as well as Viton sheet.
-
The valves were made as before. The inlet valve was turned from 7mm brass hex, the outlet valve is a modified Regner union. I still had enough of the milled cross to make the stops. The body and the valve parts went into the ultrasonic bath together with the parts of the new fuel injection pump. The seats were formed by giving a 3mm ball a light tap with a brass rod and the valves were assembled and checked. I turned a 6mm dia. spring seating to go on top of the diaphragm nut and assembled the pump. A quick check with water showed good suction and the level rose around 5mm with each stroke.
Next mounting the pump on the engine and making the overflow chamber.
-
Some more, rather slow, progress on the fuel pump.
The mounting plate was cut from 3mm aluminium sheet and screwed to the back of the timing gear housing. The eccentric strap will be cut from a length of 25mm wide 2mm thick brass and the eccentric is turned from a dog end of 25mm diameter mild steel. With the light loads I don't expect any wear problems. I have used a similar combination for the fuel injection pump on the horizontal engine.
-
Hello Roger,
I'm still following along. :ThumbsUp: :ThumbsUp:
It's been a long detour making the fuel pump, before you can start testing the SU carburetors again.
Stay safe
Mike
-
Thank you Mike :)
Time is limited at the moment but I am plodding on. The excentric was turned to 18mm diameter in the independent chuck and then the strap mas set up to bore out to 18mm. Back to the 3 jaw chuck to part off the excentric and then some hand work to finish off the strap. The 18mm bore was slightly off centre but looking at the picture of the setup I can see that the centre was not quite on the mark. There is enough material to recover the problem.
-
Has anyone attempted a sliding plate carburetor?...instead of the usual cylindrical throttle?
Can anyone point me in the direction of one with a set of plans?...
Dave
-
I'm not aware of one. The problem is to find a suitable depression zone to atomize the fuel at low speeds. The Amal/Bing type use a cylindrical slide with an needle and are generally found on motor bikes. Slide throttles are common with fuel injection for competition engines. I guess you are thinking about your 917 project.
I finished off the excentric strap and excentric and it all seems to fit together and move as it should :)
https://www.youtube.com/watch?v=7Vq3JPmhhKE
This is driving the engine via the generator with a 6V battery. I had to hold the exhaust valves open to remove the compression.
Next up is the fuel reservoir, starting with some 15mm diameter brass tubing.
-
That pump seems to be a nice bit of kit. Thanks for your wisdom on the slide carburetor. You're quite right.
Dave
-
Great to see that you're still making progress Roger :ThumbsUp:
Top line motorcycles did get flat-slide carbs 5-10 years before all manufacturers were forced to switch to injection (EU Norms regarding pollution). My Suzuki DR350SE from 1992-4 (under restoration, with parts from different bikes) has a Flat-Slide CV carb and so did my GSX-R1100H from 1987 - though to be honest, they are semi-flats, as there is a circular section around the needle.
Truly advanced Flat-Slide carbs, are the American Lectron carb from 1974, and it's brand new successor - SmartCarb - from the same constructor. If you want a slightly more detailed description about the differences - have a look here (second post down) :
https://ktmtalk.com/showthread.php?541931-38mm-stepped-down-to-a-36mm/page3 (https://ktmtalk.com/showthread.php?541931-38mm-stepped-down-to-a-36mm/page3)
Truly astonishing that it get a modern EPA approval and in some respects are better than EFI :o
There are a few pics of it installed near the bottom.
-
Thank you both :) Those carbs look interesting :thinking: There are several other designs I have looked at like the Reese Fish weir designs:
https://en.wikipedia.org/wiki/Reece_Fish_Carburettor
I'm not sure I could make a miniature one of these :headscratch: ::)
A little bit more on the fuel reservoir. The fixing bracket was made with hacksaw and files rather than complicated milling set ups. There seems to be good penetration of the silver soldering of the base. :) The overflow tube and pipe connections can then be soft soldered without disturbing the structure.
-
Great to see that you're still making progress Roger :ThumbsUp:
Top line motorcycles did get flat-slide carbs 5-10 years before all manufacturers were forced to switch to injection (EU Norms regarding pollution). My Suzuki DR350SE from 1992-4 (under restoration, with parts from different bikes) has a Flat-Slide CV carb and so did my GSX-R1100H from 1987 - though to be honest, they are semi-flats, as there is a circular section around the needle.
Truly advanced Flat-Slide carbs, are the American Lectron carb from 1974, and it's brand new successor - SmartCarb - from the same constructor. If you want a slightly more detailed description about the differences - have a look here (second post down) :
https://ktmtalk.com/showthread.php?541931-38mm-stepped-down-to-a-36mm/page3 (https://ktmtalk.com/showthread.php?541931-38mm-stepped-down-to-a-36mm/page3)
Truly astonishing that it get a modern EPA approval and in some respects are better than EFI :o
There are a few pics of it installed near the bottom.
Thanks for that Admiral! Great information Very interesting flat pin metering valve on that ...
Dave
-
You're welcome Dave - I didn't respond in you own question thread as I have no plans for model versions ....
Best wishes
Per
-
I'm following this thread out of fascination with the subject. In the late 1960's I designed a continuous flow FI for my Mk1 Sprite. Never built (I installed a 45DCOE instead) but learned a lot.
These days I go to patents to get more insight into the workings of items, in this case the rotary disc carb. The full description of the Lectron is here: https://patents.google.com/patent/US8616179B2 with a treasure trove for referenced patents. And also patents that cite this one. This can be quite a rabbit warren to get into, but printing a few out for reading in the reflection room is always rewarding.
The SmartCarb patent is I think: https://patents.google.com/patent/US10371100B2
Gerrit
-
Thank you all for the interest and additional information :ThumbsUp: :ThumbsUp:
I stripped the carb down again to drill and tap the fixing holes in the body. 3mm holes were then drilled in the reservoir for the fuel inlet and outlet nipples. To keep the overflow pipe centred and at the right height I turned an aluminium spacer with a 3mm hole in the centre for a piece of 3mm rod to keep the pipe in place. This was all soft soldered together and I made a start on plumbing it all together.
-
Hello Roger,
Still following along with interest :ThumbsUp: :ThumbsUp:
Just a thought :thinking: When you get the engine to run with this pumped fuel system, you may like to try an experiment to simplify the pump. Try replacing the eccentric drive with a small nut and bolt through the diaphragm; and see if engine vibration alone will be enough to move the diaphragm and pump the pump. If that works, it would sure simplify future pumps.
I think Perry Carbs have a pump something like that.
Just a thought :thinking:
Mike
-
Thank you Mike :ThumbsUp: When (if) I get it to work I will try that. The biggest problem is probably the initial priming :headscratch:
The first trial did not go well :( The fuel in the reservoir had a much bigger meniscus than I imagined (around 3mm) so the carb flooded, the oil was washed off the cylinder walls and the engine seized :toilet_claw: I removed the plugs, squirted some oil in and managed free it up but am not sure what damage has been done. It is fairly easy to shorten the overflow pipe and reduce the level but I'm not sure it will be ready for the show :(
There is still compression and it fires but I think I should strip it down and check.
-
That's a bummer Roger :-\ .... well live and learn (thought this at times are a drag).
I hope that the damage is only minor.
Mike - I feel sure that the Perry Pump is a membrane pump that is 'driven' off the (single cylinder) crankcase pressure fluctuations ....
A bigger version of the same principle is the Velo Solex (French moped), where the pump has enough pressure to drive the fuel injection system directly - kind of funny when you consider that the Solex company made carburetors back then.
-
Hello Per,
You are correct, most Perry pumps use crankcase pressure fluctuations to vibrate the diaphragm. However I believe (cannot find the reference at the moment) they also offered an engine vibration version, driven by a weighted diaphragm. I may be wrong, memory can play tricks these days.
Roger will need to give the engine stand a good shake to prime the vibration pump. if he tries the experiment. :ROFL:
Never thought that petrol would have a meniscus. We learn so much from these experiments. It's not easy.
Mike
-
As you say lots to learn ::) I turned an additional 5mm thick spacer and moved the overflow pipe 5mm down. The penetration on the original soldering was excellent :) I hope it is the same after the move. The engine still seemed very tight and after some investigation it seems that the actual seizure was the fuel pump eccentric :toilet_claw: It was a battle to remove and free up but without it the engine felt normal. Once again I didn't allow enough clearance :(
-
Roger,
That must be a relief for you to have identified the source of the seizure problem, which thankfully was not a piston/ cylinder issue.
I guess its the sealing which prevents a simple variable depth overflow. Could you put a thread in the existing overflow pipe with smaller a screw in (and therefore height adjustable) overflow pipe. :thinking:
Mike
-
I have used the perry vp-20 oscillating pump for years on 4 stroke model engines. They require the higher levels of vibration of the 4 stroke engine as well as the freedom of movement as the engine runs in the model aircraft. http://www.perrypumps.com/prod02.htm
-
I had a chance to make some longer trials on the balcony (at 9°C no one else is going to be sitting out on theirs).
The engine seems to have developed a water leak around the water pump inlet and the cooling fan only works intermittently when the engine is running. I'm guessing electrical interference or possibly vibration. You can see it starting in the video when I switched off the ignition.
I tried the three different trial needles I had made and the best results were with the smallest taper, around 0.6-0.7°. The engine seemed to pick up the load fairly well but I need to get the cooling sorted out so I can run it for longer before it boils.
https://www.youtube.com/watch?v=aFVkmtzUvWg
-
Sounds like you got it to run really well Roger :ThumbsUp:
The cooling system problem is annoying - but at least easily cured :cheers:
-
A quick trial after improving the cooling system:
https://www.youtube.com/watch?v=wXndv7xm82k
I need to remake the needle as it is a little rough at the idling end and sometimes sticks. Interestingly the profile is almost identical the needle used in the Amesbury V8 :thinking: The engine is smaller, the carb bore is smaller, the spring is much weaker and the needle is the same.
-
I took a few more measurements. At first I thought that the needle was sticking but it was at the idle level. if I pushed it down it came back up to the same point.
The jet is about 0.2mm below the bridge.
The bore is 5.5mm
The bridge is 0.5mm high.
The spring starts to have an effect at ~0.9mm piston lift.
It idles at around 1.7 mm piston lift.
Full open is around 4.4mm piston lift
The jet is 0.8mm diameter
At idle position the needle is ~0.77mm diameter
At full open the needle is ~0.74mm diameter
As long as I have concentricity better than 0.01mm the needle won't jam at the idle position (maybe, don't know :headscratch:)
I think the piston is reaching full open too soon. If I use a stronger spring the needle/jet clearance may be too small.
It seems to work quite well but more trials are needed ::) :thinking: :wine1: :wine1: :wine1:
-
Progress Roger, :ThumbsUp: :ThumbsUp: :ThumbsUp:
Could you simulate a stronger spring by putting a small weight (or your finger) on top of the piston to stop it rising too far too fast?
Mike
-
That's a good thought :ThumbsUp: :wine1: I have some small brass weights that go on top of the piston rod from the first trials.
-
I tried the largest of my weights, 4.5g, which didn't seem to have much effect :headscratch: This was a short trial as it ran out of fuel ::)
https://www.youtube.com/watch?v=MCmn2VqGorU
-
Hello Roger,
The piston and weight appear to rise and fall in proportion to the throttle opening. Thats what I would expect to happen.
If you want a smaller piston rise per throttle angle then fit a slightly heavier weight/spring. The needle taper is another matter, you will need to decide if you are rich or lean or OK at each piston height.
Maybe with a longer run we could see whats happening a little better.
It's progress
Mike
-
Thank you Mike :ThumbsUp:
I did a few sums and as the spring rate is 0.07 N/mm the 4.5g weight will cause around 0.5mm deflection. I was able to confirm this by carefully balancing the weight on a spare spring ( I am very good at getting the decimal point wrong when going from g to N ::) ). I do have some of the same type of spring with a free length of 22mm which I could cut down for further trials. The Amesbury version had a mild steel piston which I was worried would be too heavy and made mine out of aluminium like a full size SU (it's also easier to machine and doesn't rust) maybe he had already learnt some stuff :thinking:
Now what I need is some time for some longer trials ::)
-
I made a cover for the fuel chamber as I was getting some splashing and then packed the two cylinder engine and a few useful bits in a box to try it in the welding area at work. Initially it jumped around a bit on the steel table but a spill absorber pad underneath kept it still. I tried adding and removing the load at various speeds/throttle openings which seemed to go quite well. It generally sounded better under load. I didn't take any electrical measurements this time, but with full load the resistance is 1 ohm so I will be getting 60-80 w. The generator and resistor bank heated up quite well and with an extended full load run the engine cooling system was on the limit. A bit more detail work on the needle and spring and I think I can call it a success :) :wine1:
https://www.youtube.com/watch?v=PzqN-UTOmjU
-
Well done Roger :ThumbsUp: :ThumbsUp: :ThumbsUp:
That was a good result and a nice reward for all your persistence and dedication to the task. Am I right in thinking the choice of spring rate was the main driver? Are you going to up-issue your drawings? I would like to try one for myself
Hope you get the same success with your fuel injection.
Stay safe
Mike
-
Thank you Mike :) :wine1:
I am still using the original spring but I think a little stronger might be good :headscratch: I have attached the latest drawings of the carb and fuel pumps as R12 DWGs (I think you can read them). They are certainly not up to Brian Rumpnow's standards but are hopefully understandable.
I am slowly learning about the diesel injection systems but I think there is at least a year to go with the current rate of progress ::)
-
Thanks for sending the updated files.
Unfortunately I cannot open them. I can still open "CD carb2 R12" dated 13/5/2020 but not the 08/11/2020 version. My AutoCAD says 'created by incompatible version'. Dont you just love computers :killcomputer: :killcomputer:
Mike
-
I have cleaned the carb drawing and resaved as R12. The start of the file has the correct code so I hope this works ::)
-
Thanks Roger,
Got it this time. :ThumbsUp: :ThumbsUp:
Mike
-
Goodo :) I have done the same with the pump drawing.
-
:ThumbsUp: :ThumbsUp: :ThumbsUp: