Old Russian motorcycle fuel injection

From Frankenstein to Proteus and everything in between
dexter_lab
Posts: 39
Joined: Sun Feb 02, 2020 8:01 pm

Re: Old Russian motorcycle fuel injection

Post by dexter_lab »

Great work, impressed. Are they suppose to produce that sort of noise?
mck1117
running engine in first post
running engine in first post
Posts: 1493
Joined: Mon Jan 30, 2017 2:05 am
Location: Seattle-ish

Re: Old Russian motorcycle fuel injection

Post by mck1117 »

dexter_lab wrote:
Sun Sep 13, 2020 3:23 pm
Great work, impressed. Are they suppose to produce that sort of noise?
That medium frequency vibration sounds like the PID is underdamped.

Kavabanga, did you use the PID autotune button to tune them?
Kavabanga
Posts: 66
Joined: Sat Apr 13, 2019 9:07 pm

Re: Old Russian motorcycle fuel injection

Post by Kavabanga »

mck1117 wrote:
Sun Sep 13, 2020 7:49 pm
Kavabanga, did you use the PID autotune button to tune them?
Yep. No else but autotune. It's just test of system. I'm gonna to connect second APPS's and ETB's channels and make more clear tune little bit later
dexter_lab
Posts: 39
Joined: Sun Feb 02, 2020 8:01 pm

Re: Old Russian motorcycle fuel injection

Post by dexter_lab »

good. about the tubes to connect the TB with cylinders, put more metal to reinforce them. you will get a big moment if you not support the TB and use air filters mounted to the TB. especially where you want to put screws
Kavabanga
Posts: 66
Joined: Sat Apr 13, 2019 9:07 pm

Re: Old Russian motorcycle fuel injection

Post by Kavabanga »

Hello! Yesterday I added second channels to both ETBs and here is my actual problem in the video....

Also, here is my little progress with ignition system
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Old Russian motorcycle fuel injection

Post by AndreyB »

you say

please upload your tune to rusEFI online

you say
My problem is:
To start my boxer two cylinder engine, I use a kickstarter (because is the only one method of starting which I have)
And I also have only one place to set a trigger wheel - a camshaft. Nothing more. The camshaft ONLY
When I kick my engine, the camshaft can ony make one full revolution. It is not enough for the Frankenso to synchronize.
You can see that video:

Also you can check #support_member_vehicles to see all my attempting to start the engine (with carburetors. In the close future I'm gonna install injectors.)
But! Usual ignition system, which was installed on camshaft and had two teeth and a hall sensor, was able to start my engine from one kick even at -10 degrees Celsius.
You can see here (Russian language, 2017):
But my goal is to have a sequential ignition and injection.
I see one suggestion here and it requires the software changes.
It's possible to install two trigger wheels: one with two teeth (TDC when eqch tooth goes from a hall sensor) and the hall sensor, the second wheel with X-Y formula to help the ECU to synchronize on higher RPM's.
And we can do this:
When tooth is inside a hall sensor, it sends +5 volts via the signal wire.
When tooth isn't inside the hall sensor, it sends 0 volts.
When tooth gets into the hall sensor, pistons is far away from TDC (it is on the both cylinders at the one time)
When tooth gets out the hall sensor, it is exactly TDC
On the 0 RPMs when the tooth gets out the hall sensor, we have to fire exactly at this moment (and inject a little bit earlier. Maybe after ECU initialization? Or when the tooth gets inside the hall sensor?)
It will alow the engine to start.
When It's started:
If we know the tooth's parameters (dimensions, angle), we can use the moment when the tooth gets into the hall sensor to predict the ignition and injection moment!
It will alow us to get the idle and, after that, get the synchronisation and even use ignition and fuel tables
Future plans:
When we get the idle, it's the engine's speed is enough to get the information from X-Y wheel (not 2 teeth wheel which we use for start the engine) to synchronise and switch to sequential ignition ang injection.
Also I recommend to check this video for more understanding:

I'm not so strong in the programming and to get into such a hard project as RusEFI to write a code, based on my idea, is a very hard, so I ask you to do it if you can. I would be really happy.
Thank you
Hope, I haven't pissed off nobody with my engine yet

so, if it's on camshaft please use "4 stroke with cam" in trigger dialog

please use "enable faster engine spin-up"
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
Kavabanga
Posts: 66
Joined: Sat Apr 13, 2019 9:07 pm

Re: Old Russian motorcycle fuel injection

Post by Kavabanga »

Hello!
I've used settings you suggested. It didn't help.
Matthew suggested put "double 9-1" trigger wheel on the camshaft and set "4 stroke without cam" and put 9-1 wheel in settings.
I did it.
I've just posted my current tune, you can check.
And today's log is also here.
But yesterday I met a fail..
I broke my kickstarter. It was impossible to repair it easy due to some problems and I was forced to deinstall my gearbox.

(Sorry for music... I was so concentrated and didn't noticed it when I was shooting the video...)
By the way I installed my record: 31 minutes to deinstall the gearbox, ~1 hour to install it

After I fix my kickstarter, I came back to try to start the engine.
The "double 9-1" wheel made ste situation a little bit better. The engine even tried to start a couple of times. And it almost didn't tend to spin in the wrong direction. But it still doesn't start...

As you can see, it steel can't synchronize properly...

One additional thing:
Sometimes, after kicking, the engine stops before the TDC and compression in the cylinders forces the engine to spin back a little. It makes my Frankenso think that the RPM increased to ~1000 rpm.
By the way, I also have a "trigger error" message....

But with my "point ignition" system, which uses a MCPU to calculate dwell, the engine still works great!
I still think, we need the software changes as I showed above. But I don't know how to do that....
Attachments
9-1double.csv
(59.58 KiB) Downloaded 859 times
Kavabanga
Posts: 66
Joined: Sat Apr 13, 2019 9:07 pm

Re: Old Russian motorcycle fuel injection

Post by Kavabanga »

Hello guys!
Today I did:
A stroboscope based on LED
Managed to set the trigger angle advance allowed me to control the engine. The system works very bad but the engine can work at least
Managed to did the same with the VR sensor
Remember the rocks on the RPM sniffer? I was right, I had noises on the hall sensor's line. Schematic was: 1kOhm between +5v and the signal wire and 1kOhm between the signal wire and PE11. I connected the sensor via analog input on the Frankneso and noises gone. But I didn't satisfied because I don't wanna spend one analog input channel on the hall sensor. I'm gonna develop a schematic which will allow me to connect the signal wire to PE11 or another digital pin without any noises.
But now, at least, my RPM sniffer doesn't have noises and I can control the engine better.
But about the VR sensor: I still have noises... Even with the microphone cable...
I used 2-0 trigger wheel in each video I made today.
But at the last moment I installed the 4-1 wheel and started the engine again. After I started, I connected the Frankenso to the ignition coil and the Frankenso didn't managed to control the engine.... But I have a high speed and usual log, so you can check.
Also, check my videos please:





Kavabanga
Posts: 66
Joined: Sat Apr 13, 2019 9:07 pm

Re: Old Russian motorcycle fuel injection

Post by Kavabanga »

Also here are my logs and tunes:
2-0 with hall sensor before I connected the signal wire to the analog input pin
Attachments

[The extension msq has been deactivated and can no longer be displayed.]

First log.csv
(97.46 KiB) Downloaded 819 times
Kavabanga
Posts: 66
Joined: Sat Apr 13, 2019 9:07 pm

Re: Old Russian motorcycle fuel injection

Post by Kavabanga »

2-0 with the hall sensor after I connected the signal wire to the analog input and managed to control the engine:
Attachments

[The extension msq has been deactivated and can no longer be displayed.]

Kavabanga
Posts: 66
Joined: Sat Apr 13, 2019 9:07 pm

Re: Old Russian motorcycle fuel injection

Post by Kavabanga »

My attempts with the VR sensor and the 2-0 wheel:
Attachments

[The extension msq has been deactivated and can no longer be displayed.]

Log with VR.csv
(228.18 KiB) Downloaded 845 times
Kavabanga
Posts: 66
Joined: Sat Apr 13, 2019 9:07 pm

Re: Old Russian motorcycle fuel injection

Post by Kavabanga »

My attempts with the VR sensor and the 4-1 wheel:
Attachments
Log with VR 4-1.csv
(235.03 KiB) Downloaded 818 times

[The extension msq has been deactivated and can no longer be displayed.]

Kavabanga
Posts: 66
Joined: Sat Apr 13, 2019 9:07 pm

Re: Old Russian motorcycle fuel injection

Post by Kavabanga »

Hi guys!
I finally can start my engine by the Frankenso! :)
It starts after the first or second kick
If it doesn't, my chances decrease after each kick, so I just restart the ignition and start kicking again.

Check the video
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Old Russian motorcycle fuel injection

Post by AndreyB »

Wow that's interesting progress!

Please post this winning tune on rusEFI Online

What trigger wheel is that?
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
Kavabanga
Posts: 66
Joined: Sat Apr 13, 2019 9:07 pm

Re: Old Russian motorcycle fuel injection

Post by Kavabanga »

Thank you!)

I posted the tune https://rusefi.com/online/view.php?msq=492

The wheel is set up on a camshaft and has two teeth. Each tooth is 90 degrees. With a hall sensor it starts and works good.
Kavabanga
Posts: 66
Joined: Sat Apr 13, 2019 9:07 pm

Re: Old Russian motorcycle fuel injection

Post by Kavabanga »

Today I did some "endurance" tests.
First, I used the software version released on March, 10, 2021.
I had big troubles with engine start (if my engine didn't start after the second kick, the red LED on my Frankenso turned on and the Frankenso didn't do sparks at all 'till I restart the CPU)
But when my engine started, everything was great.
I passed 20km and decided to come back to my garage and install the latest version of the software. So, I had a laptop in my garage and could check why I my red LED turned on
Check the video... I don't know how to explain this...

Then I decided to install on of previous versions of the software released on December, 12, 2020 and I've never had the red LED. I passed 30km after installed this version of the software and my engine always started after the second or third kick.
I think, we have troubles somewhere in the software.
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Old Russian motorcycle fuel injection

Post by AndreyB »

red error means fatal problem. please connect console and read fatal error text

i suspect what it could be but please read it from console
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
Kavabanga
Posts: 66
Joined: Sat Apr 13, 2019 9:07 pm

Re: Old Russian motorcycle fuel injection

Post by Kavabanga »

The TS already gives me this (check the video)

Isn't it enough?
Attachments
Screenshot 2021-05-01 034826.jpg
Screenshot 2021-05-01 034826.jpg (28.18 KiB) Viewed 29986 times
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Old Russian motorcycle fuel injection

Post by AndreyB »

https://github.com/rusefi/rusefi/issues/2552

This validation is temporary disabled in latest nightly so latest nightly is not affected
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
Kavabanga
Posts: 66
Joined: Sat Apr 13, 2019 9:07 pm

Re: Old Russian motorcycle fuel injection

Post by Kavabanga »

Thank you!
I tried the latest nightly version of the software today
Works great!
Kavabanga
Posts: 66
Joined: Sat Apr 13, 2019 9:07 pm

Re: Old Russian motorcycle fuel injection

Post by Kavabanga »

My butterfly didn't handle dependability tests...
It collapsed on a highway in 35 kilometers far from home... The butterfly's mileage at the moment of the collapse was 140 km... The engine stopped at the one moment and I found out that I was on the highway at 11 PM alone...
Thanks a lot to one guy who helped me and tow me to my garage (hope, he'll read it because I showed him this topic and the whole forum at all).

Looks like a guy who made the butterfly for me used a "bad" type of metal instead of a type I asked him to use. He set me up!

Find yourself at a highway at night far away from home with no chances to start the engine is not the best thing to feel
Attachments
f_st99ZCBb8.jpg
f_st99ZCBb8.jpg (492.64 KiB) Viewed 29866 times
Mh1lixq6icQ.jpg
Mh1lixq6icQ.jpg (726.8 KiB) Viewed 29866 times
Last edited by Kavabanga on Sun May 02, 2021 12:56 am, edited 1 time in total.
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Old Russian motorcycle fuel injection

Post by AndreyB »

Ohhh that's brutal.
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
Kavabanga
Posts: 66
Joined: Sat Apr 13, 2019 9:07 pm

Re: Old Russian motorcycle fuel injection

Post by Kavabanga »

Good news!
Today I managed to start my engine!!!
TWO ETBs. TWO WBOs and it works!!!!!
My problem was in a bad battery, which was falling down to 5.5 volts when I started the engine even though it had 12 volts when ignition was turned off
I connected another battery and engine started
WITH EFI!!
WITH A KICK START!!!!!!!
Only one cylinder works (the second) because the mixture is extremely rich (need to tune) and a spark plug on the first cylinder gets wet every time I try to start the engine and can't do sparks. But I have to admit, I use old spark plugs with different gaps and so on. I'm gonna buy new spark plugs.
I've got a little closer :)

Kavabanga
Posts: 66
Joined: Sat Apr 13, 2019 9:07 pm

Re: Old Russian motorcycle fuel injection

Post by Kavabanga »

Hi everyone!
Today I had my first EFI riding!!!!

2 ETBs
2 WBOs
A hall sensor as a CAM position sensor
But not everything was so good and after a half of an hour riding I had to push my motorcycle back to the garage (thankfully, it wasn't far away)
I was riding with an autotune turned on and ony of my cylinders (exactly that cylinder which's WBO the autotune used) dead because of dirty spark plug...
And after that I could use the idle only, because on other regimes the engine tended to stop...
So, I turned on the first gear and get back to my garage on the first gear and the idle
But when I almost go to the garage, the riding on the first gear got boring and I tried to shift the gear. I was on a hillock, so, the engine stopped...

Is it possible to use two VE tables for different cylinders (or banks)?

And is it possible to tune each cylinder (or bank) via autotune using two WBOs for two cylinders (banks)?

If you're interested in, here is my tune
https://rusefi.com/online/view.php?msq=549
and a log
https://rusefi.com/online/view.php?log=254

Also check how I use an APPS from Nissan as an acceleration handle position sensor


and you can see motorcycle's pre-assembly


And also how it looks now
Attachments
photo_2021-06-06_04-18-09.jpg
photo_2021-06-06_04-18-09.jpg (176.03 KiB) Viewed 28978 times
photo_2021-06-06_04-18-13.jpg
photo_2021-06-06_04-18-13.jpg (124.77 KiB) Viewed 28978 times
photo_2021-06-06_04-18-27.jpg
photo_2021-06-06_04-18-27.jpg (156.88 KiB) Viewed 28978 times
Kavabanga
Posts: 66
Joined: Sat Apr 13, 2019 9:07 pm

Re: Old Russian motorcycle fuel injection

Post by Kavabanga »

Hello everyone!

My Ural has already passed about 350 kilometers with RusEFI! I'm playing with settings, coils, sensors and other stuff, but today I installed very rare exhaust pipe which I've been trying to found for about six years!

I'm waiting for your advices about cranking settings and other things
Hope you enjoy :)



User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Old Russian motorcycle fuel injection

Post by AndreyB »

Love is #rusEFI mentioned right in video love :)
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
Kavabanga
Posts: 66
Joined: Sat Apr 13, 2019 9:07 pm

Re: Old Russian motorcycle fuel injection

Post by Kavabanga »

Hello everyone!

I've benn riding my motorcycle for one year with RusEFI. A lot of things which are now able with EFI and ETB's made my riding more enjoyable. But the most important thing was a really weak point: starting/cranking. Due to the RusEFI scheduler's architecture and kicktart cranking features both, in was really hard to start the engine. It usually tended to start in the wrong direction. My kicktarter was broken several rimes this year due to this. I couldn't rely on my motorcycle in critical situations because I never new if it starts or not... I don't have abilities to install an electric starter to my engine and I needed a special mode for start.. I called it "the kicktart mode".
Matthew Kennedy helped me to figure out what strategy of code changing is better to use and explained about the scheduler, ingition logic and so on. And he even improved my code (which is posted here: https://github.com/VasiliyKh/rusefi/commit/8a05bc3ffa7e047f4aa12069b1ed075818104f0b and here https://github.com/VasiliyKh/rusefi/commit/5823ca85cb8ceb4639b792a6cd407f443a23daed) Matthew's changes are here. I use it. https://github.com/rusefi/rusefi/compare/master...mck1117:kickstart-testing?expand=1

The logic is simple: ignition is turned off 'till the engine gets 800 RPM and more. Each time the ECU sees a trigger tooth goes out from the sensor, it charges the ignition coil and asks the scheduler to spark in X ms (however you want).

And those simple changes made my engine start really fast!!!!!


Later I tuned the priming pulse parameters and now my engine starts really fast (after 1st to 4th kick. And it's with the broken kickstarter which's stroke is less the a half compared to the normal kicktarter!!!) . I'd make the software get to a special injection mode during the kickstart cranking, but I'm gonna think about it little bit later.

Now you can compare. How it used to be:


And now:



I even got rid of a rare optical sensor which I've used for some time and which didn't seem reliable for me as in the datasheet said that it can work in conditions when the temperature is less than 80 degrees Celsius (It's much warmer under my air cooled engine's case). So, I've got back to widely spread LADA hall sensor!

Now I can tune my engine in a proper way!
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Old Russian motorcycle fuel injection

Post by AndreyB »

Search hint: First successful rusEFI kickstart test above
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
Post Reply