Got to run the Pressure fed setup today - works great on a quick bench test setup curve - Curve is very linear.

I did lower the feed pressure from .8 to .4 PSI as the curve was very low - 0.4 seems to work well

I did have an issue with idle as I couldn't lean it fully - Found it was my "green" solenoid that was leaking by a small amount - Changed out for a white one and all's well - Was able to program a "Fuel Cutoff" that shuts the engine off at any speed.

assembling another "pump" and plan to try it on some other conversions to see if the System can "scale".

https://cimg1.ibsrv.net/gimg/www.rcu...3647b6e2dd.jpg

Chris, do you think that raising the fuel pressure and narrowing the carb fuel nozzle would help to pre-atomize the fuel better? There are some smoke jets made by Bad Dog that look to perform quite well.

I was thinking this might be a good test Lonnie. Hard to see in the video but the fuel is leaving the spray bar in a fairly straight stream and is likely hitting the other side of the barrel.. less than 1psi is not much pressure but I'm sure some improvements could be made here...

This is what you get for $17 CAD delivered. :)


https://cimg2.ibsrv.net/gimg/www.rcu...e2097c9877.jpg

Used one to make another pump setup to try on other engines - The inlet and outlet is different on this one to make it more compact. The pressure is "fixed" too and set with the spring length - This one is set for 0.35 PSI.


https://cimg1.ibsrv.net/gimg/www.rcu...2d76acf221.jpg

Squeezed down on the jet size - Now near 100% duty cycle at full speed - 0.4 PSI bias. Runs well.


Can still see the fuel stream coming out in a laminar stream and hitting the opposite side of the throttle bore - took a video and slowed it down - between the Camara phasing and the solenoid opening it makes for a neat video - you can see when the port is open (intake draw) and the fuel pulse starts the fuel breaks up better. I think when the port is closed and the spray comes - it hits the other side.

Ideas to help atomization????

Yeah, that's a super cool vid! The engine runs quite well.

I know that very small orifices can be be difficult to drill accurately. Kind of why I wanted you to see the Bad Dog video. I think he grinds the orifice on a 45 degree angle, like a Saito spray bar. This creates a slit that has a football-like shape. Saito creates that profile, and when combined with an overlapping low speed needle tube it works quite well. It allows for a variable orifice size as the throttle opens and closes. Different with your carb, not having a low speed needle of course, and you could cut the slit 90 degrees from what's shown in the attached Pic.

Using a closed ended nozzle with a slit pointing down the carb you wouldn't have fuel spraying against the opposite side of the throat. It may be easier to control orifice size using that machining technique, offering the the ability to run higher bias pressure and increase pre-evaporation of the fuel? Whether or not that would enhance running behavior, I have no idea, just throwing ideas out there.

https://cimg8.ibsrv.net/gimg/www.rcu...02469e4e8a.jpg

At some point, close to 100% dutycycle, the atomising effect of the solnoid is lost.
I get the feeling, that the nozzle jet is too "clean". How about a "mushroom" of sorts, a tiny disc on a stick, opposite of the jet, approx in the middle of the air intake? It would spread the fuel more evenly over the cross-section of the intake without hindering the airflow too much?
Or perhaps some sort of emulsion tube? Downside would be that there is an object in the air intake, which in case it would come loose, could cause consideable damage.

Another idea would be to do what is done by some motorcycle companies: placing the injector concentric above the air intake, the jet directed straight into the engine.

Another option could be to relocate the nozzle to somewhere on the crankcase, injecting fuel into the whirring and stirring engine parts, but that would only work for 2-strokes.

Although it looks like a garden hose stream in the video the jet in the runs is only .4 MM (16 thou) I have tried a few ways to make a Saito/OS/Supertiger style slot and could make a "glow size" one but getting it accurately made smaller is troublesome and I would loose my "orfice size data" reference.. I think I have an idea to do a Splash plate style one... The engine is running very well and I don't know if this is really required - it just looks bad. The solenoid is only cycling at 30 hz - or about 1800 RPM. on the two stroke cycle - at half power (and half duty cycle) only about every 1.5 turns has felt flow and 1.5 turns does not - even at full power you see interruptions. This all seems to work out in the wash as the engine runs great. A solenoid shut off signal kills the engine immediately at anything above half and even at idle it shuts off quite quickly so it's not loading up much fuel.

I like the emulsion tube idea - Properly sized it could be included in the "injector body" But if pressure feed allows for a "big bore" inlet then there would not be much motive draw (vacuum signal) for the air.

Now that winter has arrived here I find myself with more time to "play". My next project will be to try and add the "pump" to my boxer twin setup. While the air pump pressure setup on there now works - the plumbing and the programming is much too complicated to be reliable - I had an issue this summer when It suddenly refused to run (in the air). I do have a "spring" available to get the pump pulse from the 4 stoke intake but can I not just use the Case pressure (pressure vacuum pulse) or is that too dramatic on a 4 stoke?? never ran a diaphragm carb on a model four stroke.

Crankcase pulse should be possible, with a "T" connection on the crankcase vent.

Thanks Bert. I will try that method first.

I don't think crankcase pressure from a 4 stroke will run a fuel pump.

Why Lonnie? Because of no additional air being drawn in. There is definitely a pulse there at leas when slow but not sure what happens at speed.

I do know one of my clone carbs was for a four stroke trimmer motor and had the spring and a port to the carb bore back side of the throttle.

Yes, not enough air movement to operate a remote pump. Every 4 stroke I've seen uses the low pressure signal from the intake and a return spring on the carby diaphragm to operate the pump.

You need to place the "T" as close as possible to the crankcase nipple, or use a separate connection, but the crankpulse is actualy stronger than the intake pulse.

The thing that is reducing pulse strength is in the pressure drop caused by the movement of the effluent. Make a separate connection that only transfers pressure, not flow, and you will have an excellent pulse.


Using the intake pulse requires a different spring at the pump diaphragm

LED blinking or steady lit?
 

After uploading the sketch to my Seeeduino XIAO, if successful, should I see the orange LED blinking or lit steady?
I tried uploading Blink which resulted in it blinking.
However, since uploading Dave's sketch, it still blinks the same (approx 1 second intervals).
I did reset the board before doing so.

Are you sure it flashed my software? My code doesn't blink that led.

I am now sure it did not flash as required.
It appears that I have a charge only cable. I have ordered this one (charge + Data) from ebay: https://www.ebay.com.au/itm/19578432...r=495639279378

I expect that will work and get my board uploaded properly.
Thanks for the reply.

Jim.

Winter has arrived in our location with "Teeth". We have already had about 12" (30cm) of Snow and backed by wind we have some impressive drifts.. Temps hit -34C last night and -40 with the breeze. Nasty.... This has prevented any real experimenting.

I did get a pump completed and run on the Boxer and It shows the same promise. Make the setup dead easy compared with my air pump setup. a lot less plumbing and fuss.

Just a run to prove it worked - Pump at 0.4 psi Enough fuel for a solid top end - No tuning done just a trial to see if it would supply fuel. No tuning tonite as it's around -30 C - Couldn't keep heat in the motor.

The pump is set up on the pressure pulse - I tapped an additional nipple into the back of the cam cover in line with the crank - Put a small restrictor in the drain line and it seems to have a solid supply pressure with fuel to the carb in a few seconds with an electric starter. I didn't want to mess around tapping into the intake anywhere.



https://cimg3.ibsrv.net/gimg/www.rcu...4b4bec1d90.jpg
https://cimg4.ibsrv.net/gimg/www.rcu...add2fac9ca.jpg
https://cimg5.ibsrv.net/gimg/www.rcu...83f4df1e2e.jpg

Both OS and ASP foresaw an alternative connection there in the backplate, not drilled out but a cast ridge for just in case. You used that for the additional tap?

I really like the idea of a pressurized fuel supply, just one remark: my set-ups basically function with the exhaust pressure rising with RPM, compensating for unloading in the air.

I would expect a setup with a fixed pressure, to need a more rich groundrun in order to lean out in the air for peak performance.

You MIGHT get better results with a set-up where for example either muffler pressure is routed to the compensation hole in your pressure regulator, or maybe as an alternative, a pitot tube connected to that compensation hole. I expect muffler pressure to be best.

Your 0,4 psi preset pressure would then get additional the muffler pressure added to it, leading to more fuel delivery if the engine unloads in the air.

I have thought about this Bert and next flying season I will try this with and without the compensation. I do know some Walbro type setups show this tendency. Not sure a pitot would serve the purpose except for simple cruising type flight - downlines with the throttle pulled would be an issue I think. The one thing I've thought about is the sense line might need a little trap to catch any oil residue and we are kind of back to a "crap trap"though on a much smaller scale as there is really no "flow" of gas, just a pressure signal.

The other thing is I love how this Boxer sounds on straight stacks.. :)

Since I have a telemetry RPM signal to my radio, I have considered a straight electronic solution - We could have the system add a small bias either through the radio or through the controller with RPM input and change the PW to the solenoid. Again more complexity though.

You/ve got a point there I overlooked... Most definitely. Pitot, not the best idea...

I think a simple small buffer volume (could be as simple as for example a larger fuel filter housing without the filter cartridge) would allready do the ob, with perhaps a periodical cleaning, line once per year or so.

That would also be a solution, albeit indeed a trickier one to dial in.
But it COULD be dialled in to perfection... Theoretically at least.
Given how hard that can be, not my preference. I think the muffler pressure would be the easiest to deal with.

Happy to see crankcase pressure worked well enough to operate the fuel pump. Always heard it wouldn't. Maybe being a boxer helped?

I think, because it has its own separate pulse-tap...

The crankcase ventilation nipple is an orifice, and due to the strong flow the pressure in the ventilation line is strongly reduced, Bernouilli's law)
But the separate tap, not carrying any flow, the full pressure pulse is availlable (Pascal's Law)

How do I confirm this? Is there a simple way of knowing it has flashed successfully? Apart from powering it up in the circuit and determine if it is providing and output to the solenoid or not?
I ordered a new DATA cable and tried uploading the sketch again to the XIAO.
No success yet.