Hi Bert,

I just receive your kit today, you did a great job. I'm going to start my project and i wondering if i can use a capacitor 100uf to get rpm from the rcexl on oxs rp2040 (3.3v do i need a divider?), and is it interresting to get temperature to optimise settings ? and if it is the case, what is the max temperature? My questions are probably already asked before but i didn't find it and i'm a noob in gaz engine.

Thanks again for your job and time !

Happy to hear you received the goods!
I am not sure on your question about the RPM, because I don't know what oxs rp 2040 means...

Temperature always is interesting, but the question "what is max temperature" cannot be answered easily, because it STRONGLY depends on HOW that temperature is measured. More specifically, WHERE on the engine, what kind of sensor, and how is that sensor attached.
It also depends on whether it is a 2 or 4 stroke engine, and what kind of metallurgy is used for the running gear (piston and liner).

If you can provide info on that, preferrably with pictures of how your sensor is fitted to the engine, I can maybe give some general guidelines. Without that, if I mention a maximum temperature but your sensor is fitted to a place that never reaches that temperature, you could damage your engine without ever seeing the temperature I gave.

For now, WRT temperature, all I can say is that, AS SOON AS you notice an RPM drop when running full throttle, reduce power immediately, and first confirm spark timing (visual check if the sensor has not come loose), then confirm whether the engine maybe is running lean, by opening the main needle (count the clicks!!!) and opening the throttle again. .

If tuned to peak, and ignition timing is OK, but after a certain time the engine starts to drop RPM, it is overheating. This can be due to not yet properly broken in, although on the bench it is not uncommon for engines to have a tendency to overheat. Cooling on the bench is far from optimal. Cooling is much better in flight.
If possible, fly the plane and see if the problem persists.

Here is a video of my printed solenoid holder bubbling away on the .15 LA. You will see a crank case pressure line but it is plugged and I’m just using muffler pressure. In the video the solenoid is running at 89% duty cycle, 30hz.



https://cimg9.ibsrv.net/gimg/www.rcu...984b70486.jpeg


im not sure how this is happening but i tend to blame dimensional inaccuracy of the printed part. I am going to work a bit more on this until my CNC shows up then I’m machining it if my skills allow me such a feat…

I want to make sure I understand this valve as I modify my sleeve design and such. The two lower sections are obvious. The third section though… there are little indentions near the top. What are they for? Could this be a source of air getting into my system? This is right next to the spot where my supports hold the part to the print bed so dimensional accuracy is worst here. I can orient the part differently and try again but I want to ask if those holes are anything more than snap points for assembling the solenoid.

i experimented with a variety of capacitors from 10uf up to 100uf and they all worked with openxsensor on my arduino nano. it's been a while so give it a shot and see.

It looks to me, that what you see is normal cavitation caused by the valve. leakage usually tends to produce larger bubbles directly at the spot of ingress. That has to do with surface tension: Bubbles need to reach a certain size before buoyancy overcomes the surface tension and the bubble rises upwards.
Cavitation bubbles form IN the liquid and can be (and usually are) much smaller, like we see here. The reason it shows up so clearly, and how it collects close to the inlet of the carb, is most likely a combination of unfavourable conditions:
-Your engine has a very low fuel consumption.
-The piece of fuel tubing between solenoid and carb is fairly long. those two factors combined, my estimate, given engine size and RPM, is that fuel spends at least 5, maybe 10 seconds in that piece of tubing.
- that piece of tubing also is vertical.

The low fuel consumption combined with the rather long piece of tubing, and its vertical position allows the minute bubbles formed in the fuel to collate (cling together), gain in buoyancy, gain in vertical velocity, promoting further collating. and due to the low flow of the liquid, collect in the top and remain stagnant there.

I think you will allready see a significant improvement if you shorten the tubing between valve and carb to "as short as reasonably possible" and route it as horizontal as possible.
If you can find it, use the smallest inner diameter in order to minimize the time fuel spends in the line, and to improve the ability of fuel to carry the bubbles along.

I noticed you are running glow? Keep in mind, you will have much less control over ignition timing, and the little you have, comes from mixture strength, meaning the exhaust residue will be rather dirty, which is an issue if you do not use an effective craptrap in the muffler pressure line.

Forgot to adress this... The upper "chamber" above the O-ring has no real function other than provide the propere depth for the needs of a Stihl M-tronic carb, and can be ignored.
The holes most likely have to do with assembly or manufacturing, I would not know to be honest.
It is NOT a source for air: as said, IF there is a leak there, fuel should flow out with the checked crankcase pressure you applied previously.

Very interesting. Then why not mount the solenoid directly to the carb?

I won’t be able to print anything new to try until Monday as I took my resin printer to work (wet lab with fume hoods. Resin printing stinks and uses solvents and stinky resins - I didn’t want it in my house). I can link the current solenoid housing directly to the fuel jet with a 10mm or so section of tubing. It will sit vertically and be more or less connected to the carb.

also I can try this housing and solenoid on my MVVS .77 which gulps fuel compared to the .15.

hey if this is just normal cavitation then a printed housing is feasible.

for the .15, it will always only be on glow as the ignition module and battery and such will weigh more than the engine itself. to help with the timing issue, I opted for unleaded, unoxygenated 100 octane fuel. This should allow for a somewhat leaner (cleaner) mixture compared to pump gas. It costs as much as traditional glow fuel but I don’t really fly much so I don’t care.

Chris (Cat1) did just that by designing and machining a throttle body without a needle, instead a calibrated orifice downstream of the solenoid placed directly into the throttlebody where normally the Needle Valve would have been.
He shared his drawings with a clubmate of mine who made one for his OS FX46. It runs like a charm, but pretty decent machining skills are needed for that method.

Here's a vid of that:

It might be a touch heavy, but so is a glow battery.
A 40~50" wingspan, high wing plane, built with weight in mind, will be able to carry the weight of the ignition, and fly with the power of that .15 LA. It has been done before.

I haven’t tried it yet but I am hoping the OS gt5 gas plug will keep it lit without the glow battery. If I have to use a battery then yes, either consider spark or start working on a small, light FET-driven glow driver. I’m sure the existing SAMD could handle a few more lines of code to switch on the glow driver (PWM the FET).

Sorry i didn't bring enough details, I'm going to put the sensor on the cylinder head, my engine is an os 46AX i'm waiting for magnet to start... for oxs it's here :github.com/mstrens/oXs_on_RP2040

i will try, i don't know if the rcexl tacho output is in 5v, if it is, i will make a voltage divider, i'm going to make your mixture controller, too it's great !

For 2-atrokes this is usually the best way to fit the temperature sensor...

Attachment 2274623
Attachment 2274624
Attachment 2274625

First bench test information please?
 

I have now wired up my controller. I have it connected to the Stihl/Zama diaphragm carburettor on the bench.
I have it powered by a 2 cell LiFe battery through a diode as described over on RCopter's GitHub instructions.
The second wire to the carb solenoid is connected to ground on the XIAO.
What tests (if any) can I do on the setup before mounting it on an engine?
Should I be able to hear the solenoid when powered up?
If so, mine is not working yet.
Edit:
If I apply 3.3 volts intermittently to the solenoid, can I expect to hear it clicking?

Jim.

Yes... a soft rattling sound should be audible. It's not loud, so listen vewwy carefully (I will say this only whence :D :D :D )
I never tested the minimum voltage. Hans' drivers I believe use 5V.

I know that with a 4 cell NiMH you can hear it click. I do not know if the click is still audible at 3,3V.

Measure the resistance, the coil should be in the neighbourhood of 30 Ohm.


Mind you: Both Hans' controller, and Dave's, they only rattle when there is a servo signal applied, and that servo signal must be "somewhere in the middle of the range. At the outermost stickpositions (1000 us or 2000 us, give or take) the solenoid is either continuously open or continuously closed, and does not make any audible noise.

Mine clicked just audibly with 6 volts applied. (from a flat LiFe 2 cell battery and resistor).
Yes I did that; mine tests at around 35 ohm.
I have checked all my circuit connections multiple times. The output of channel 5, as per Dave's .otx file, in my Taranis output looks ok. It approximates the curve shown in one post on the other thread. (I have the controller plugged into channel 5).
So far, I cannot find anything wrong.
Tomorrow morning I will get out my regulated lab power supply, my little oscilloscope and test the MosFet and circuit operation properly. Maybe even swap it out for another similar one from my box of bits.
Jim.

Mind you, the sound is really soft. Most of my planes I have to listen fairly careful if I want to check if it is going.

Easiest test is to simply connect it to a linear servo channel, and using a piece of clean fuel tubing, blow into the valve while moving the corresponding stick. You should notice a resistance proportional to the stick movement.
Most conclusive test it to set up a syphoning fuel line, and observe the outflow, like the video of the "bucket test" I posted earlier.
The noise you should hear, is by the way clearly audible in that video, but nowhere eeven close to being that loud. The metal bucket amplifies the sound and the camera's mic is apparently very sensitive to clicking noises...

New XIAO needed?
 

I did all those tests plus some Bert. No success.
I used my little DSO 150 oscilloscope to see if there was any modulation signal on the MosFet gate. Dead as can be; no modulation signal.
I have tried all the methods of fault finding that I am aware of without success.

I will order some new Seeeduino XIAO boards.

More when they arrive.

Jim.

Damn... Sorry to hear that...

Dear all, today was a very good day for my gaser! 😁
I flew all the day with my trainer, swap from the 25LA to the 40FP, and the 40FP run so well with the solenoid (compared to 25LA where micro bubbles come from the cavitation of the solenoid and become big bubbles and kill the engine after one minute).

So, one more time thanks a lot to all guys which develop this solution!! (Berts, Raleighcopter, and all other)

One question, this morning my solenoid curve seems to be too rich, and this afternoon maybe a little too lean, do you observe this little change on air/fuel mix ratio during the day on your plane? Normally the BPM280 must correct this little variation, isn't it?
​​​​​​
​​​​​See ya!!

First try with brutis
 

Hello folks

First try with bert IS a success
 

Hello folks first startup and 3 tanks of 400cc 10% oil ok.
Just had to invert efi Channel
But promissing
https://cimg3.ibsrv.net/gimg/www.rcu...14fcf8f224.jpg
Thunder tiger f91s. Gas efi conversion

To be honest, that is not normal, BUT it can be your curve is not yet 100% perfect. Just keep flying, and if you detect "spots" on the curve that you think can be improved, make minor corrections. That way a curve "matures" slowly.

Do NOT make a correction every 5 minutes. Every little change of the curve has an effect on the quality and cleanliness of the oil in the engine. This effect takes time to stabilize, and the oil quality again effects how the engine runs. Your engine beng bushed, this effect is most likely more pronounced, since oil viscosity directly influences the resistance in the bushed bearing.
If you make a correction to the curve, run the engine like that for a few flights at least, before you make another correction. Try to make only ONE correction at a time.
If necessary (this is a personal choice), take notes of what you did.

EDIT: Let me explain this a bit better...

The purpose of the solenoid system is to provide the engine with an as perfect as possible mixture, in order to have the as perfect as possible combustion, THROUGHOUT THE ENTIRE THROTTLE RANGE.
Chances are, the first bench tests, curves are not yet "perfect", meaning at some speeds the mixture is "better" and at other speeds the mixure ls "less good".
Now if the mixture is "perfect" at say, 50% trhottle, and you run the engine constant at this speed, after some time, the oil that remains in the engine gets cleaner. It easily takes 30 minutes for the oil quality to stabilize.
If the mixture is a bit rich at say, 75% this means, if you run the engine constant at this speed, the oil gets polluted with soot and unburnt fuel. This also takes quite a bit of time before the oil quality stabilizes.

The state of pollution pof the oil, has a minor but noticable effect on how the engine runs.

But... we do NOT run the engine at constant speeds. During flight, we move the throttle. Which means, the oil quality is constantly changing, therefore the way the engine runs, also changes a bit.

By running the engine at constant speeds (after take off, climb to a safe height, select a throttle setting,for example 50%) and keep this setting for 1 or 2 minutes, and carefully judge how the engine runs. Repeat this for as many as possible different settings, and try to identify spots that need correction.
You can do this, for example, the first flight of the day. Take a quiet moment (not too many planes in the air at the same time).

Make TINY corrections, and leave it like that for the rest of the afternoon. Let the oil find its new "balance".

Make no mistake: The influence of the oil quality in the engine (the state of pollution due to imperfect combustion) on how it runs, is extremely small, and cannot be "seen" or "heard". But it IS THERE...
One way to tell, is to observe the oil coming from the exhaust. It will NOT tell you where the rich and lean spots are in the curve, but it will indicate that they still are there.
The lighter the oil residue, the better your fuel mixture curve is. It will never be 100% clean oil of course, but just keep an eye on it.
You can NOT find a "perfect" curve on the ground. Only in flight.
This sounds like a lot of hassle, but really, it is not: Your plane is currently flying! If you're satisfied, don't change anything. If you notice a rich spot, carefully correct it and continue flying. Finding the perfect curve is a natural proces, and you will only know you have found it, when you suddenly realize, you have not made changes for a long time... :D

Every next day you are at the field, you will notice that the engine runs better and the corrections are smaller, until, after a few weeks, you realize that you do not need to make corrections anymore.
Once the curve is settled, all that is possibly needed is corrections for large seasonal changes, and those you make on the needle, NOT on the curve.
Seasonal changes are depending on your local conditions. West and north Europe, in winter the engine needs ONE click richer, in summer ONE click leaner. Mediterranean climates might differ. Tropical climates usually are fairly stable.

The thing is, try to minimize the corrections, and develop a "test routine" that helps you identify the rich and lean spots in the curve.
There IS "learning" involved, but once you get a feeling for it, you will find it relatively easy. I had to learn it as well, and I first had to figure out HOW to learn it... :D :D :D

Thank you very much for your answer Bert! Yes, during the flight I detected some points not perfectly adjusted, often too rich.So I must now try to make the adjustments in the air with your method.This is the goal for next weekend. I noticed during the flight that if I stay for a few moments at half throttle and then I do a full throttle, the engine makes a too rich noise and after a few seconds, it comes back with a correct sound like a good air/fuel ratio.I think it is due to the engine temperature, when you stay at half throttle for a while, the engine temperature is lower and when you put full throttle back, it takes a few seconds to increase the temperature for a good combustion of the fuel. Is my analysis correct?If so, is it better to adjust the curve for a lower temperature (like at half throttle) or a higher temperature (like at full throttle)? If I set the curve to a lower temperature, I fear that my engine will stop at a higher temperature due to a too lean setting (and then I have to use my poor glider skills to find the landing strip, with a high probability of glue party 🤣)And probably when the curve is perfectly set, you don't notice this combustion change between lower and higher temperatures?