definitely,,..apparently piston replacement was mentioned also..parts are no longer for those big twins , i would have assumed the guy was running the right oil ratio, i mean he is in his late 60,s or so , well on the other hand probably 30 to 1 ratio as i have read that mm had recommended that..

Ah, yes, well... The MMM conversions are worth crap. The guy has a thousand vids of customer converted engines, but YT vids of engines fitted with his conversion kits in actual flight are rare, far and few inbetween. He indeed recommends IMHO irresponsibly low oil rates.
I have re-done two of such conversions, both times the owner told me I could keep all the crap I pulled off the engine (which I didn't, what am I supposed to do with it?)
Believe it or not, one of those engines was initially modded by me (simple carb groove mod) and ran fine when I delivered it. Owner never flew it because "it didn't look nice", bought a MMM mod for it and sent me back the carb that I modded.
Back then, like today, I did those mods for free, only materials used.
Then he found out that his nice looking but expensively modded engine now ran like crap, so he shelved it for a year or so before giving it away (unknown to me at the time) to another guy that eventually contacted me if I could have a look at it.
After some back and forth talk, I figured out the history of the engine, realized that I still had the carb originally belonging to the engine and told him it would be a 10 minute job to fix it. So he brought it over, and left an hour later (test runs take time) with a happy face.

Thats a bit of a coincidence for you , and brilliant you still had that carb to put back on..
i really do believe that a walbro hanging of the underside of a boxer engine is a terrible look...
im amazed that the man didn't like what you had done to his engine, couldn't imagine your set-up looking terrible, even if it was a free mod ,
I would sure like a larger boxer engine, or even a pegusus to convert,

I'd be okay with the look of a walbro if it worked well. There is no denying the modded meth carb is more compact. I think the solenoid and controller has better performance given it has the atmospheric compensation even with the added bulk of the crap trap, solenoid, and controller.

Hopefully, the new controller I'm working on (Chris developed the air pump part) with air pump for tank pressure will make it even more convenient. I'm planning on knocking out the code and schematic this weekend so with a bit of luck we'll have something to test and talk about pretty soon.

yes true, would not be too concerned about looks over functionality, thats great to hear more progress!
a new controller, sounds greal, .. well eventually for me i would love to givetthat system a go ..

I plan to have some working code (based on Chris's work) and a working controller by this weekend if possible.

I've been working through power requirements for the new controller. The air pump i have (https://www.ebay.com/itm/263145594968?) runs on 5v but also runs (slower) on 3.3v. I'm currently running the solenoid on 3.3v, sourced from the xiao's onboard voltage regulator. I'd like to keep the parts count as low as possible and it's currently using about 150ma (screen, microcontroller, bmp280, solenoid). I think the pump will draw about 200ma which, with the other current loads, would be close to marginal for the onboard regulator, especially when powered with a 2s lifepo. I may run it from receiver voltage.

Above all, the controller reliability is paramount. I'd rather not add an external regulator if I can avoid it. I may power the solenoid from the receiver battery directly if required.

Chris, I have several questions for you...
1. Are you using 5v air pumps or did you order something that runs on higher or lower voltage?
2. Are you powering your pump through a regulator or are you using your receiver battery?
3. What tank pressure are you generating at full throttle? (Or if you have any numbers, bert, feel free to share)

I haven't measured the pressure my pump generates. I do have a calibrated digital manometer that can measure down to a few hundredths of an inch water column so I can match your numbers. My pump generates a bunch of pressure on 5v but I'm not sure how well it'll throttle down and I'm not sure what it will generate at 3v so I'll be dusting off my manometer soon.

​​​

Dave, FWIW, I'm not worried at all about adding extra components if reliability is more robust. The Xiao is already nano compenentry in size and weight.

On gasser carbs: They look fine to me and the few engines I have tested with CH ipIgnitions full gas conversions run really well, both on the ground and in the air. Not sure what MMM is doing across the pond but it's not rocket science to get a gas conversion to run good.

I'm also trying to keep costs down. 3 bucks here, 4 bucks there, it adds up.

Considering the cost of the solenoid, solenoid housing and crap trap or air pump the controller is dirt cheap imo. The reliability here is no no place to skimp, 3 or 4 bucks is insignificant to me in trade for that. I spend close to $60 just for a reliable device to turn on and off my ignition without sending interference back to my RX.

Hopefully, no crap trap. I do believe Chris built a barf tank that recycles the gas back to the tank though. The details are a bit hazy for me at the moment.

My engine mods look like a tie-rap tying a sensor to the front, and a magnet Epoxied into the propdriver, instead of a niche shiny green anodized CNC milled sensor bracket and magnet carrier ring...

I should share my 3d printed hall sensor bracket file. It looks good.

All I can tell, is that I on average see something like 15~20" of fuel column.... ish...

...of course you have numbers. What was I thinking.

If it's the one you posted a pic of a few months ago, yes it's very nice. Unfortunately not compatible with my Saitos.

The issue is simple: Since the originally smallest commercial utility engine used to be 23 cc there never was much development in carbs for smaller engines.
And a 26 cc fourstroke twin requires a totally different carb than a 2-stroke single of the same displacement.

In all fairness, MMM's hardware is pretty good: Very well fitting brackets and rings, and very well made carb adapters.
But either his carb selection totally sucks, or his supply is inconsistent and he slaps on whatever he can get, I don't know. Fact is, if you watch his YT vids on a large screen and half speed, you will notice a very inconsistent throttle response that (deliberately or not) is hidden by manipulating the throttle such that the engine sounds as if it responds excellent. Only if you pay close attention, you will see that engine response not always matches throttle movement.

It is one of the things I lately at least TRY to get very unambiguous into my videos.

:d :d :d

couldnt complain with that setup..especially if your told whats what before the alteration was done..

bert just sent this link to me: https://jivaro-models.org/glow_essen...html#gsc.tab=0 .

congrats on the story about you, bert.

i threw together a driver for the air pump with what i had on hand: a 2n3904 transistor with 1.2k (anything 1k-3k will work) base resistor and a 10 ohm series resistor for the motor and a 1n4148 snubber diode directly across the motor. the final version will be more robust, i'm sure. this is something like $0.10 worth of parts.


.

What kind of signal does it use? An analog proportional from the controller? Seems to me it is not designed to handle a direct RX output to be driven from a transmitter generated curve?

So Chris was taking a channel, doing some math with the channel value, and adding it to the result of the engine rpm after he applies an exponential curve in the controller to the rpm.

First step is figuring out how to best drive the motor and that was a simple transistor switch that I put a pwm signal to. I'm going to add a capacitor across the motor to see it it performs better today.

I'd like to start reading s.bus since we're using 2 channels for the controller. Sbus is inverted serial so I need a transistor to inVert the s.bus so hopefully I can read it with the serial port on pin 7 of the Arduino (currently not used by the Xiao controller.)

​​​​​​i think I also want to buffer the rpm signal to protect the rpm pin.

Those plus the 2 drivers for the solenoid and air pump makes 4 transistors.

Rather than r separate transistors with their 4 separate base resistors plus series resistors....

I just finished this clock...

with digit driver's using uln2003a chips which are 7 Darlington transistor pairs with base resistors all wired up in a 16 pin dip package. It'll work great with 3 extra Darlington pairs unused (for now). This should make wiring very easy.

I'm off to breadboard the circuit and write some code....

...I wonder if it really needs a channel to be able to modify the air pump or if it just needs to react to engine rpm?

... or even, what's the tuning procedure when using the air pump? Is the hope to tune the engine on the ground using the solenoid with the air pump reacting only to rpm changes and then in the air tune only the air pump? This would have the air pump handling all of the mixture corrections due to engine unloading while the solenoid handles changes due to air pressure and temperature changes only.

Is that workable? will we be able to achieve a set it and forget it type of operation? Will it be easier to tune? My gut tells me we should set the pump up to mimic the pressure curve we would get from muffler pressure and that attempting to modify that pressure profile will prove difficult but I haven't tried it yet.

If I would have to guess, I'd say in real life the muffler pressure is a function of both throttle position and RPM (in helicopters, despite constant RPM muffler pressure rises and falls with throttle position, I know that much, and in airplanes, with constant throttle, muffler pressure also rises and falls with RPM, or the leaning out due unloading engine could not be counteracted with muffler pressure).

I would think a few testruns with measuring equipment should result in fairly generic ratios that allow for some sort of simple programming, analog to how the gas-law in the original controller compensates for temperature and barometric?