Speeduino 2T EFI Project

[TZ350]

http://www.supermototecnica.com/2015...tta-due-tempi/

I like the way he squirts above and below the piston. Not so sure about squirting down the side of the piston. But it is 100% time available.

https://patents.google.com/patent/US...njection+rotax

This arrangement looks very time limited.

The disadvantage is that the air stream transports the spray droplets instead of evaporating them

Yes, I very much agree with you, Flettners idea of straight down the B ports is the best for evaporating the fuel droplets and getting a homogeneous fuel mixture.

[TZ350]

Got my Speeduino fuel injection CPU kit. Put it together and started to find my way around it and understand its methodology.

Squeeze bottle is for the MAP sensor. Leds are Injectors and the pot is for TPS.

Interesting system. It has staged injection but instead of it being the small or large injector. Just one or the other as in the Ecotrons system. The Speeduino fires both together but varies the on time of them individually to suit.



I have run into a bit of a problem. I have the Speeduino system in two stroke mode and have been sending a 100Hz ignition signal to it. This should equate to 6,000 rpm but I am getting a healthy 12,000 four stroke reading.

Not sure what is happening. No doubt I am doing something wrong, but what? .....

[TZ350]

http://www.supermototecnica.com/2015...tta-due-tempi/.

Interesting link, I notice the guy says he has good drivability. Drivability is my problem.

People have had a lot of success with the Alpha-N methodology on bikes like the CR250 and have got good drivability.

But the 250 and bigger singles have a RPM ceiling less than 10,000 rpm and a lower than 10 bar BMEP compared to tuned 125's and 50's that typically rev to 13-15,000 rpm and develop more than 10 bar BMEP.

Over 10 bar BMEP a two stroke motor becomes very reliant on the wave action in the pipe for its power and this wave action is very finicky at smaller throttle settings and prone to collapsing with a subsequent loss of air flow through the motor. Alpha-N (TPS vis RPM) cant cope with this effectively because it cant see changes in airflow, it can only guess at it.

It seems that less than 10,000 RPM and 10 bar BMEP you only need one logical injector and an Alpha-N map will do. Over the magic 10 threshold you need staged injection and a way of seeing changes in air flow for the same RPM and Throttle position.

To get good drivability with a tuned 125 or 50 something like a VE volumetric map and MAP or MAF sensor value is required. Something that can adjust the fueling to match the changes in the finicky airflow at small throttle settings. VE is RPM vis MAP or RPM vis MAF so it can see changes in air flow and adjust the fueling accordingly.

Four stroke methods for measuring MAP or MAF do not work very well in two strokes. Another, more two stroke method has to be used. I am not sure what that is yet.

[Niels Abildgaard]

Four stroke methods for measuring MAP or MAF do not work very well in two strokes. Another, more two stroke method has to be used. I am not sure what that is yet.

Many years ago I made hot wire anemometry for some district heating caloric meters.
The signal was from .5 to 250 hz and had to pass through a layer of epoxy and stainless steel.
Without that I think a sensor in one of the transfer channels could detect amplitude using the first half of the open period .
Would that leave enough time for injecting?

[TZ350]

Many years ago I made hot wire anemometry for some district heating caloric meters.
The signal was from .5 to 250 hz and had to pass through a layer of epoxy and stainless steel.

Without that I think a sensor in one of the transfer channels could detect amplitude using the first half of the open period .
Would that leave enough time for injecting?

It is an interesting idea. I would like to know more about how they are made. The frequency range is good.

Injection is timed to finish at BDC. But the hot wire reading could be used on the next cycle.

The bulk of transfer happens well after TPO transfer port opening and just before BDC where the pipe should be sucking its hardest.

A sensor in the transfer channel would have to cope with regular doses of fire, you can see the combustion staining all the way down the transfer ducts.

[ceci]

This arrangement looks very time limited.




I do not understand why you say that the injection time is less.
It is injected into tranfers B in all cases

[TZ350]

I am not sure that washing the oil off the piston is the same as the direct down the transfer duct like Flettner and KTM's arrangement.



That is a great picture you posted of the injectors firing straight down the B transfer tracts.

Looks like Flettners and KTMs injector arrangement and is what I have tried to copy with my own engine. Having experimented with lots of different injector placements it looks like we are all starting to following Flettners lead. Straight down the B transfer duct is proving to be the most sucessful.

[TZ350]

Mass flow in the transfer duct, its easy to see why finishing the injection squirt around BDC is a good idea.

[TZ350]

.
Really Happy. I am starting to understand how it all works and have got all the gauges working. Everything is dummied up and my calibrations are crap but for the moment it is just about understanding how to wire things and set up Tuning Studio.

With my previous EFI experimentation the system completely switched between the small injector to the big one so it was all small or all big and that seemed to work Ok. I will be interested to see how the Speeduino approach of running the small injector all the time and adding the big injector as required works out in a two stroke.

I have been able to simulate a multi tooth wheel and get sensible RPM readings. And a lot of things look familiar, like the pulse width and duty cycle is the same at various RPM that I recognize from my previous work with that other EFI system. So things look very promising.

The only thing I have not simulated yet is the fuel pump. With that other EFI system at "Key On" the fuel pump would run for a short moment to prime the system then turn off and wait until the motor is cranking before turning on again. I will wire in another LED to check to see how Speeduino handles it.

Very little wiring required out to the injectors Etc so putting it all on the bike should be easy.

The engine needs a re build, may take a couple of weeks to get sorted and back into the bike, then I am looking forward to having a go with the Speeduino.

I am not sure how I am going to handle the VE mapping but I am ... ... Very excited to be making progress.
.

[TZ350]

.
I needed to figure out how to get the fuel pump switching on and was told that there is not a dedicated output for a fuel pump but I could re assign a spare MOSFET to do the job or I could use the development area of the board to add some components for switching a pump relay.

Assigning a spare MOSFET sounded easier so on the Speeduino user projects thread https://speeduino.com/forum/viewforum.php?f=19 I asked what I needed to do to re assign a MOSFET and whether I needed to go into the Arduino code to do it or should I be looking somewhere else?

This is the reply.

Just assign in TunerStudio to the desired output pin in the Accessories > Fuel Pump section. The post here is one that explains, also with two following posts for two methods (schematics and utils.ino) for determining the current pin assignments of specific PCB versions.

On the TS pin reassignment for Fuel Pump, note there is a slider on the side of the selection list, giving you many more options than 3 through 9 if you scroll the list down. v0.4 FP output is default on Mega pin 45 (goes to the proto area). Boost output is default on Mega pin 7.

• Set FP to pin 7.
• Burn.
• Set Boost to pin 45.
• Burn and Save tune.
• New FP output is now through Q8 to J5 pin 1 and IDC pin 35 (your choice for wiring).

The output will ground when the FP should be active. It should be active when Speeduino is powered-up with 12V, for as long as the Fuel Pump Prime Duration is set (default 6 seconds), or when Speeduino detects RPM.

David

[TZ350]

.
Well that was easy, I followed the instructions and it all came together real easy. The third LED represents the fuel pump. The fuel pump itself draws to much current for the Speeduino board so a relay would be needed and we would wire the relay coil where the LED is.

1-Primary Injector, 3-Secondary Injector, 35-Fuel Pump.

I have labeled everything important, wiring is real simple.

[Number-One]

Hi, do you know RusEfi?

https://www.youtube.com/watch?v=X7IXrSoFngc

[pete376403]

TZ, what platform are you running TunerStudio on? I have been trying to get it running on Lubuntu, (i've downloaded the correct versions for Linux) about to give up and go back to Windows.

Edit - never mind, I can see from screenshots that you're on W10

[TZ350]

.
I have been having a bit more success. I was using a signal generator to simulate the cranks toothed wheel but without a missing tooth Speeduino seemed to be having a bit of trouble interpreting what was going on.

I followed some good advice and got Ardu-Stim, compiled it and installed it on a Arduino UNO. Using Device Druid to configure it I was able to simulate a 24-1 tooth wheel and Speeduino seemed much happier and Tuning Studio displayed realistic results so it looks like I am getting close to the pot of gold at the end of the rainbow.

It is not that obvious what to click to down load the Ardu-Stim zip file.

Arduino toothed wheel simulation code.... https://gitlab.com/libreems-suite/ardu-stim

Serial communication Library required by Device Druid to talk with the Arduino DUO.... https://www.arduinolibraries.info/libraries/serial-ui

Device Druid, used for configuring the Arduino simulation application.... https://devicedruid.com/

YouTube video clip of the TuningStudeo gauges moving about.... https://youtu.be/YeBDTIAjPys



The UNO is used to generate a simulated toothed wheel signal for the Speeduino EFI board. The Tuning Studio gauges all seem to be doing sensible things.

[TZ350]

Now as soon as I can get this lot back together I will be able to try my Speeduino EFI CPU out in the real world.