[4wheeljive]

A bike like that with those specs would be fun to play on


Mr Dan you have mad skills

[oozuk]

how's this progressing mate?

[JrZook]

Not too much interesting progress to report so far but I have completed the PCB and gave it power! It didn't end up in a puff of magic smoke so my confidence is still high!

Tested the PCB and all seems well! Wrote a bit of test code to check the inputs/outputs of the board and its all fully functional!! Woooooo win!!

Now its on to coding and interperating things like battery voltage, MAP and temperature sensor inputs before I go on to decoding the flywheel reluctor inputs of the crank.

Here's a few pics of the hardware I'm going to use on it:



38mm CV type throttle body


Fuel rail and reg to be modified, denso injectors and atmospheric MAP sensor (9V batt for scaling)


Fuel pump module from a CBR125


Actual fuel pump from the module that will be converted to external use.

Dan

[royce]

why does the throttle have a piston and butterfly? is it some kind of strange AFM?

[JrZook]

royce wrote:

why does the throttle have a piston and butterfly? is it some kind of strange AFM?

Nope! Its a constant velocity type EFI throttle body . Don't have to run servo controlled secondary butterflies to stop acceleration lag this way and I reckon it will be easier to code the acceleration enrichment using this.

[JrZook]

Well first descent fail of the project. Blown injector drivers!!! Seems like they don't do what the data sheet tells me! FARK!! Time to modify the circuit.

In other news im reading the raw data from the TPS, battery voltage and MAP sensor. woooooo!

[fordem]

Ahhh - assuming the injector drivers are the TO220 transitors along the edge of the board - do they have internal freewheeling diodes?

[Fatzook]

I'll come for a drive this weekend and see what needs doing for the manifod. Been doing some more stainless lately.

[royce]

JrZook wrote:

royce wrote:

why does the throttle have a piston and butterfly? is it some kind of strange AFM?

Nope! Its a constant velocity type EFI throttle body . Don't have to run servo controlled secondary butterflies to stop acceleration lag this way and I reckon it will be easier to code the acceleration enrichment using this.

Nope you've lost me, cant see the point in a piston causing a restriction in the intake if its not performing any other function like metering fuel based on air flow or something

[JrZook]

fordem wrote:

Ahhh - assuming the injector drivers are the TO220 transitors along the edge of the board - do they have internal freewheeling diodes?

They do have internal freewheeling diodes built into them (IRF840's). Maybe I should have been abit more clear. The FET's didn't die but the FET driver (ucc27324) did and im still clueless to how that happen. Data sheet seems fine for them ie logic inputs 0-5V to drive non-logic mosfet at 0-12V.

[JrZook]

royce wrote:

JrZook wrote:

royce wrote:

why does the throttle have a piston and butterfly? is it some kind of strange AFM?

Nope! Its a constant velocity type EFI throttle body . Don't have to run servo controlled secondary butterflies to stop acceleration lag this way and I reckon it will be easier to code the acceleration enrichment using this.

Nope you've lost me, cant see the point in a piston causing a restriction in the intake if its not performing any other function like metering fuel based on air flow or something

Yea I know you know your CV carbs, the slide meters the fuel. In this case I inject fuel so that aint needed but since its a rather large TB the slide will dampen the change in vacuum when the throttle is cracked from idle to wide open so the engine wont fall flat on its face! Yes it is a sort of restriction but the only other way ive seen it done on bikes running rather large TB's is with a secondary, servo controlled butterfly set.

Else wise for ultra crisp acceleration with this fairly large TB Ill require a pretty complex enrichment algorithm to deal with the wall wetting/stripping effects in the injection port. Then again that still wont sort the large vacuum loss/flow velocity into the head once the TB is cracked wide open. Dunno how well it will work yet, just an idea I had. Will easily be able to hold the slide wide open and try without the dampening effect also.

Thoughts?

[JrZook]

Fatzook wrote:

I'll come for a drive this weekend and see what needs doing for the manifod. Been doing some more stainless lately.

Fark man that looks neat! Yea would love a hand even with the intake manifold at this stage (turbo later ). My copper prototype is ugly and still doesn't fit! Going to bundy this weekend for some booze supplies, how bout the next?

Cheers!

[royce]

JrZook wrote:

royce wrote:

JrZook wrote:

royce wrote:

why does the throttle have a piston and butterfly? is it some kind of strange AFM?

Nope! Its a constant velocity type EFI throttle body . Don't have to run servo controlled secondary butterflies to stop acceleration lag this way and I reckon it will be easier to code the acceleration enrichment using this.

Nope you've lost me, cant see the point in a piston causing a restriction in the intake if its not performing any other function like metering fuel based on air flow or something

Yea I know you know your CV carbs, the slide meters the fuel. In this case I inject fuel so that aint needed but since its a rather large TB the slide will dampen the change in vacuum when the throttle is cracked from idle to wide open so the engine wont fall flat on its face! Yes it is a sort of restriction but the only other way ive seen it done on bikes running rather large TB's is with a secondary, servo controlled butterfly set.

Else wise for ultra crisp acceleration with this fairly large TB Ill require a pretty complex enrichment algorithm to deal with the wall wetting/stripping effects in the injection port. Then again that still wont sort the large vacuum loss/flow velocity into the head once the TB is cracked wide open. Dunno how well it will work yet, just an idea I had. Will easily be able to hold the slide wide open and try without the dampening effect also.

Thoughts?

How do other map based systems cope with the same thing? I see what you are saying baout what it will do but it might not be a concern? most engines will drop to 0 vacuum when you open the throttle right up

[JrZook]

royce wrote:

How do other map based systems cope with the same thing? I see what you are saying baout what it will do but it might not be a concern? most engines will drop to 0 vacuum when you open the throttle right up

Imma not a EFI engineer but they do use some pretty extensive algorithms to get around this (plus the secondary butterflies). Mite not even be a concern but we shall see.

All in all its a pretty descent size TB for only 250cc. 38mm!! Then again these TB's came of a 600, so that's a 38mm TB per 150cc!

[royce]

I assume its going to be something along the lines of rate of change in the throttle opening compared to vacuum? youll get that sorted out quick enough

my bike is 1 38mm carb per 250cc

[Fatzook]

JrZook wrote:

Going to bundy this weekend for some booze supplies, how bout the next?

Cheers!

Yeah no worries.

Grab me a bottle of reserve or distillers and I'll be your best freind

[JrZook]

royce wrote:

I assume its going to be something along the lines of rate of change in the throttle opening compared to vacuum? youll get that sorted out quick enough

my bike is 1 38mm carb per 250cc

38mm CV carb?

Yea that's the inputs to the basic algorithm for acceleration enrichment (TPS opening rate of change). Then you have the wall wetting/stripping algorithm, enrichment hold time and tail off enrichment decay time and slope. Ill get something to work in the end. Guess thats the bonus of writing the software from scratch, you can change/manipulate everything at the expensive of ripping hair out

[matt.scorgie]

this guy is a super smart man good work dude and if ya get it properly then it should be in the paper or some shit cause you are doing what people have been trying to do for years

[JrZook]

Well finally have some more success to report! Made a makeshift injector FET driver and it works !! Will do for now till I get the proper driver IC's in. Just setup some test code to read the TPS input and interpret that into injector pulse cycle speed.


Messy test bench

Makeshift driver for the moment

[JrZook]

matt.scorgie wrote:

this guy is a super smart man good work dude and if ya get it properly then it should be in the paper or some shit cause you are doing what people have been trying to do for years

Cheers man! Yea this may be a first for this bike but its still based on the traditional fuel injection algorithms. I just like challenging projects

[Teracis]

Hey man looking good, that's hectic how far you're going.

The vid is super dark though and I can't see anything but can hear the clicking.

[want33s]

Teracis wrote:

Hey man looking good, that's hectic how far you're going.

The vid is super dark though and I can't see anything but can hear the clicking.

To stop industrial espionage....

[JrZook]

Teracis wrote:

Hey man looking good, that's hectic how far you're going.

The vid is super dark though and I can't see anything but can hear the clicking.

Yea dont know what happened there, photobucket seemed to darken it out for some reason. Ill try and get a better one to put up when I can. Oh and teh clicking your hear is the fuel injector pulsing.

[DarkHorse]

Wow Dan, you really are a nerd!

Cool project though - the electronics is black magic to me, so all credit to ya. Be interesting to see how the old Virago copes with 50HP! I have images in my head of the swingarm collapsing!


(This from the guy who wants to put 180kW of SR20DET in a 93 Vit...)

[oozuk]

really cool to see it in the first stages of operation

[JrZook]

Cheers Guys!

Yea its a bit of a relief to finally see things start to work! The actual injector driver MOSFETs on this board don't even get warm so no heat-sinks will be required

Here's a clearer vid for all.



Dan

[JrZook]

Update:

Managed to get my boost referenced MAP sensor! Lil sucker will be able to read up to 21psi above atmospheric (2bar).



So far I have interpreted the raw sensor data in to KPa as well as PSI. I have also interpreted battery voltage to the resoultion of 0.1V and TPS throttle opening from 0-100% in 1% resolution. Stuff is starting to fall into place!!

Now on to getting a few temperature sensors!

[JrZook]

Fuel pump relay driver is now online!! Also got a cam sensor sorted and working that will detect a small neodymium button magnet from roughly 10mm away .

Now to get on to the crank triggering setup and start writting some real functional code

[JrZook]

So next stage of testing passed with flying colours!!

Connected the board up to the running bike for the first time to test some real-time signals and see how well my filtering circuit worked on the crank position inductive sensor.

Trying to filter the signal of this type of sensor was a little bit tricky as it outputs an ac signal with anything from 5volts peak-peak cranking to something like 30volts p-p at redline. My filter clamps this voltage to 5v max as well as clean up the signal to give sharp edges for the micro to trigger correctly.


This is what it looks like in action on the C.R.O. The top signal is the one directly coming out of the inductive crank sensor with the bottom the output of my filtering circuit to the micro.


This shows the filtered input (bottom trace) and a test injector driver code (top) with a fixed pulse width of 1ms triggered to fire on the rising edge of the filtered input.


Another short vid of the injector test running on the bike.

Also handed over my spare engine block to Mr Fatzook to get a start on the intake and turbo manifold

Stay tuned..

[royce]

yeah but how bigs the coil