• Questions about NanoEFI for the GY6 (Electronic Fuel Injection)

  • Let's talk about new parts and solutions to common problems you want to see with a BDX logo. -T@BD
November 2019 Forum Retirement
Due to lack of extra personal time (and too much spam), I'm retiring the forum and setting it to a read-only state. No new registrations or posts will be accepted as of 11/09/19. However, all information will stay up and searchable. If you spend a few minutes searching, you'll find the answers you're looking for. We may re-open the forum next spring.
Let's talk about new parts and solutions to common problems you want to see with a BDX logo. -T@BD
 #11995  by myvt365
 Mon Apr 09, 2018 11:41 am
Travis @ Buggy Depot wrote:
Sat Apr 07, 2018 11:57 am
myvt365 wrote:
Fri Apr 06, 2018 9:42 am
So here is my next question like I said I just love talking about this stuff and its highly interesting. How is the injector controlled when to spray? Every fuel injection system ive messed with used an on off trigger on the injector itself or has been rocker driven this is def a new design to me and im super excited to see it in action.

Here's my article covering the basics of how my injection timing works on NEFI and my initial tests way back in 2015. I'm sure it's not an original way to handle timing, it's probably the most obvious way to do it on an 8-bit AVR micro. It's simple, fast, and accurate.

NanoEFI - Proof of Concept: Tests successful!

myvt365 wrote:
Fri Apr 06, 2018 9:42 am
As far as faults is this a separate CAN network with a dash for what the fault is almost like OBD2?
I don't intend to incorporate CAN until support for newer more technologically complex vehicles starts making it a requirement. Interfacing with ABS modules, OEM gauages, etc. As far as diagnostics, with Wi-Fi wireless capability directly on the ECU, anyt wired troubleshooting tool becomes obsolete. Even burning firmware updates will be wireless by internet, "over the air" with no physical connection.

A physical dash that uses the wireless connection is something I've considered. But the fact is, if I take that on myself it will be years before I can finish it (look at how long it's taken to get through with A-Arm's recently, and I've been making those for years).

I believe cool add-ons like this would be best developed as community projects. NanoEFI draws a lot of guys from the maker crowd, since development for the first couple of years was on the Arduino platform. Though I want to protect my core code, I plan to encourage modifying the system through an open API for added fun and customization.


AHHH I gotcha maybe I am crazy but in a simpler fashion how will the injector be fired IE you mean pulse width modulated pump so I gather that would mean the injector would only cycle as the pump does? my question is how does the injector close and re open as to not loose prime in a sense. I get that there would be no return fuel line so aspects of a dry start or air to injector is my concerns I would think that the injector would have to be electronic in this aspect.

All of this begins with an idea when I was speaking of a CAN network im thinking way advanced and maybe a future add on for abilities of troubleshooting like a simulated diag aspect of compression test and such I know this is far advanced but I can dream right :lol: :lol: I know I use CAN networks on a far extreme levels of monitoring engine operations and it is helpful but as you stated this a huge development and would take years of engineering code C++ to dial in. single cylinder engines well they are single no need to cyl cut outs and such but the ability to expand this aspect is cool to think about.
 #12002  by Travis @ Buggy Depot
 Mon Apr 09, 2018 1:10 pm
myvt365 wrote:
Mon Apr 09, 2018 11:41 am
AHHH I gotcha maybe I am crazy but in a simpler fashion how will the injector be fired IE you mean pulse width modulated pump so I gather that would mean the injector would only cycle as the pump does? my question is how does the injector close and re open as to not loose prime in a sense. I get that there would be no return fuel line so aspects of a dry start or air to injector is my concerns I would think that the injector would have to be electronic in this aspect.
The fuel pump and injector are each on their own output circuits, independent of each other.

PWM at the pump is an continuous signal at a fixed frequency, with no reference to engine event timing. It switches on and off 20,000 times per second with a continuous square wave. The frequency of the signal itself doesn't care if the engine is at 100rpm or 10,000rpm, it operates independently of other signals.

The pulse width of the injector is different. The frequency is tracked precisely to the RPM of the engine and the calculated position of the crankshaft. It's a single pulse that is active only when the processor says the time is right. The pulse width in this context literally refers to the amount of time that the single pulse is active. It's not modulated, and doesn't use a PWM module on the microprocessor.

The pulsing of the pump is like watching a "video", with many frames streaming continuously uninterrupted. The pulse of the injector is like looking at a still picture, it is only one frame.
 #12004  by myvt365
 Mon Apr 09, 2018 1:15 pm
yep that's exactly what I was saying im used to "pulse width" at the injector being the duration of the nozzle being open not the signal of open close I think we are talking about the same thing just lost in the terminology of cultural differences of GA to VA ahhahahaahahah jking that's all I was wondering how the injector was going to relative to timing and duration or if it was even electronic haha.
 #12006  by Travis @ Buggy Depot
 Mon Apr 09, 2018 1:37 pm
myvt365 wrote:
Mon Apr 09, 2018 1:15 pm
I was wondering how the injector was going to relative to timing and duration or if it was even electronic haha.
Yup, 100% electronic. In a nutshell here's the source and flow of events:

1. Flywheel magnet passes the trigger sensor at 25°BTDC.

2. NanoEFI ECU hijacks that signal but doesn't "pass" it to the CDI or injector.

3. ECU calculates RPM based on how many microseconds (millionths - 1,000,000ths of a second) have passed since the last trigger signal was received. This is extremely responsive tach measuring versus what you find on Trailtech Vapors and other tachometers that use the "fixed window" approach. They're tachs only update values once per second or so. The NEFI internal tach value updates every revolution of the crank, with error rejection.

4. ECU then calculates how much fuel and how many microseconds need to pass before the crank will be at the desired angle for opening, and a separate variable for when it needs to close. The opening event comes first, and the processor sets a timer for that number of microseconds.

5. When the timer expires (overflows), the injector circuit is immediately driven active. The timer is set now for how many microseconds need to pass before the injector needs to close. This is what amounts to your pulse width.

6. Timer expires again, and the injector is driven to close. Rinse, repeat.

A similar strategy is used for driving the CDI and manipulating when it fires spark.
 #12007  by myvt365
 Mon Apr 09, 2018 1:55 pm
Travis @ Buggy Depot wrote:
Mon Apr 09, 2018 1:37 pm
myvt365 wrote:
Mon Apr 09, 2018 1:15 pm
I was wondering how the injector was going to relative to timing and duration or if it was even electronic haha.
Yup, 100% electronic. In a nutshell here's the source and flow of events:

1. Flywheel magnet passes the trigger sensor at 25°BTDC.

2. NanoEFI ECU hijacks that signal but doesn't "pass" it to the CDI or injector.

3. ECU calculates RPM based on how many microseconds (millionths - 1,000,000ths of a second) have passed since the last trigger signal was received. This is extremely responsive tach measuring versus what you find on Trailtech Vapors and other tachometers that use the "fixed window" approach. They're tachs only update values once per second or so. The NEFI internal tach value updates every revolution of the crank, with error rejection.

4. ECU then calculates how much fuel and how many microseconds need to pass before the crank will be at the desired angle for opening, and a separate variable for when it needs to close. The opening event comes first, and the processor sets a timer for that number of microseconds.

5. When the timer expires (overflows), the injector circuit is immediately driven active. The timer is set now for how many microseconds need to pass before the injector needs to close. This is what amounts to your pulse width.

6. Timer expires again, and the injector is driven to close. Rinse, repeat.

A similar strategy is used for driving the CDI and manipulating when it fires spark.
:o :o Does this mean the chance of an ecm controlled tach that is accurate ?????
 #12010  by Travis @ Buggy Depot
 Mon Apr 09, 2018 2:42 pm
myvt365 wrote:
Mon Apr 09, 2018 2:33 pm
I almost bought a trailtech the other day
Save that money. That functionality comes with NanoEFI, but performs better with NEFI and is ECU fed. In-app speedometer will be a part of an update after release, using the GPS you've already paid for on your phone. Much more precise than a magnet pickup.
 #12011  by myvt365
 Mon Apr 09, 2018 2:48 pm
Travis @ Buggy Depot wrote:
Mon Apr 09, 2018 2:42 pm
myvt365 wrote:
Mon Apr 09, 2018 2:33 pm
I almost bought a trailtech the other day
Save that money. That functionality comes with NanoEFI at no extra cost. In-app speedometer will be a part of an update after release, using the GPS you've already paid for on your phone. Much more precise than a magnet pickup.
Very cool that's one thing the cop said im lacking so I almost just said screw it and pulled the trigger hahah