Fuel injection retrofit

[connie_rider]

You have a good point. I agree it's probably good enough for prototype.
My concern is your going to want the prototype to run smooth at all throttle settings.
  If it doesn't, you'll have to figure out why.
If the injector size is closer to the correct size, it would remove a variable.

Keep tinkering. Your doing great.

Ride safe, Ted

[gpineau]

This one looks a lot better than my first try and it is going to fit much better as well.
I will be mounting the injectors in the plenum directly across from the runner. They will be about 4 inches from the valve but will be pointy directly down the intake runners. .

[connie_rider]

Yepp, starting to look like it's possible....
I can't imagine how many measurement's were done to determine the fit.

Keep tinkering, we're watching...
  Ride safe, Ted

PS: BDF, I sent you an Email. Did it reach you?

[gpineau]

What?  Measure. What's that?     Just close my eyes and start cutting...lol

Yes it is a lot of geometry to get it right.  It is snug and solid even without the clamps.  Long way to go to complete.

[B.D.F.]

Yep, got it and replied yesterday. The e-mail address I have for you is questionable so you may have to shoot me your e-mail address and I will re-send the e-mail.

Brian

Yepp, starting to look like it's possible....
I can't imagine how many measurement's were done to determine the fit.

Keep tinkering, we're watching...
  Ride safe, Ted

PS: BDF, I sent you an Email. Did it reach you?

[gpineau]

Today activities.
Started building a fuel injector test/measure jig. I have a pump, regulator and a fuel rail. I will build a precise timer circuit that will turn on and off the injector so I can get accurate volume measurement  cc/min..

Got the vacuum/temperature sensor today and mounted it to the manifold next to the throttle body.
Made a clear window for looking inside the manifold while things are happening.

high res photos here. https://fuelinjectorproject.shutterfly.com/pictures/60

[gpineau]

Finished up the injector test jig.
Built a timer that will turn on the injector for 10 seconds.
Build a small gas tank of PVC pipe and mounted the fuel pump inside.
Marked the measurement vessel at 2.5, 5, and 10 cc increments.

Ran out of time to test it and I need to work tomorrow but I should be able to get some time to work on it this week.
Should be able to turn on the pump, set the pressure gauge, press the start button and when it stops measure the amount of fuel in the vessel and multiply by 6 to get the number of CC/Minute.  (excess fuel pressure is shunted  back to the tank)

I did my best to set the timer to exactly 10 seconds but it was tricky. I subed an LED for the injector and took a video then reviewed the video to see the exact time the LED turns on and off. I measured it at 10.004 seconds and called it good enough.

High res photos here., https://fuelinjectorproject.shutterfly.com/pictures/68

Timer measurement test.  https://youtu.be/-1rv5-9QzRw

[gpineau]

Sometimes I surprise myself. This tester actually worked! 

I was able to set the fuel pressure and fire up an injector and measure its output over 10 seconds.
I tested two different brands of injectors. One was way overkill it was 236 cc/min and I only needed 125 cc/min. The second one I tested came close to the specs for the ninja 635. it came in at 192 cc/min at 43 psi. So I will be using the ninja injectors. I have a lot more experiments in my head about measuring the spray pattern and volume delivered over a rang of lower pressures. The injector I will be using has plenty of margin. I can lower the pressure some and dial it into exactly what is needed at 80% duty cycle. I just need to make sure the spray pattern does not suffer from the lower pressure.

From the video I see that the spray pattern is about 60 degrees and will completely cover the diameter of the intake tubes at the distance they will be from the valve. Probably get some spray hitting the inside of the plenum unless I can get them closer to the intake tubes.

In the video I was testing the ninja injectors and i found one of the 4 was under performing.  Not sure what to do to fix it but I am hoping for some suggestions. Maybe just some carb cleaner will do the trick. The video ended short because of an operator error. I tested all the injectors but forgot to hit the record button. 3 measured 192 cc/min and one bad one measured 147 c/min.

the video is here. https://youtu.be/3bbu6AfAPQ8

 After reviewing the video again I can see the "bad" injector is sputtering. I am wondering that maybe I had air in the fuel rail that was gone at the completion of the test. I do need to retest that injector.

[gpineau]

Set the regulator at 38 psi, cleared the air and now all 4 injectors are delivering 162 cc/min.  So that works out that the needed 123 cc/minute can be supplied at 77.9% duty cycle. 

[gPink]

I gotta ask .... what do you do in real life? I find this work you're doing fascinating. 

[B.D.F.]

That is a reasonable size ratio and should work fine.

You can clean dirty injectors by soaking them in solvent, then flushing them in both directions (fuel in the normal way, then fuel in through the nozzle or 'backwards') while cycling the injector open and closed rapidly. They usually clean up very well and can be determined to be clean both through the volume delivered as well as the atomized spray pattern they put out. The testing is usually done in large, clear tubes using mineral spirits (though gasoline will work, it is just more flammable) and watching the spray pattern.

Brian

Set the regulator at 38 psi, cleared the air and now all 4 injectors are delivering 162 cc/min.  So that works out that the needed 123 cc/minute can be supplied at 77.9% duty cycle.

[connie_rider]

Sounds like your making great progress.
I used to test injectors during fuel tests I did at Shell. Your set up is similar to what we had.
I have a question on your pump and regulator.
 Are you bypassing extra fuel back to the tank, or dead heading, and controlling pressure after your regulator?

What I'm wondering is; will you be able to maintain the pressure/volumn as requirements change during run cycles?

Ride safe, Ted

[gpineau]

I gotta ask .... what do you do in real life? I find this work you're doing fascinating. 

I was a Electrical engineer/ manager most of my career.  I designed computer system and wrote software at the hardware level.  I retired in 2009 due to health. I tried to stay busy flipping old motorcycle and trucks but it didn't keep me from sitting on my butt in front of a computer most of the day.  So recently I started back to work at Home depot. (part time)....

 connie_rider,  the regulator is set up to bypass the excess pressure back to the tank and that seems to be working. I have tested the pump full on and it gets  up  to 80 psi so I think it is enough to keep all 4 injectors running. . Its not a very good regulator though. After adjusting it and then turning off the pump, the next time the pump starts the pressure is different. This happens every time I cycle the pump on and off.  So it's time to get a more expensive regulator.

I still have a third set of injectors to test. Maybe they will get closer to the size I need for the 700cc magna. I hope in the end that this works well enough to retrofit my 1999 connie.  I learn something new every day.


B.D.F. Thanks for the info about back flushing. What sort of solvent do you use. I use  Berrymans B12 cleaner for just about everything.

[gpineau]

Help!

Do we have any material science guys out there in Connie land?

I need to find a flexible sealant that is impervious to gasoline.  I am trying an experiment with different types of glue/sealers/epoxies to see which ones are going to last.
The reason is that I used silicone rubber on my measurement flask yesterday and by the third test it began leaking at the bottom.

This is really bad news because the air plenum and the runner tubes are sealed with large quantities of silicone rubber. If they soften and come loose big pieces are going to be sucked into the intake port. For sure I need to disassemble them and reassemble with a more resistant sealer.  I think the JB-Kwik weld will do the trick but it is very rigid and not forgiving of mistakes. I am looking for a flexible sealant with the correct properties.


Help ! 

[gPink]

How about an RTV used for valve covers?

[gpineau]

Thanks, gPink.

searching other forums it looks like Permatex high temp red gasket maker is resistant to gasoline. Wish I had know the regular silicone sealant was not before I coated the inside of the plenum with it... Well now I know.  Lesson learned.

What time I have left before going to work I will use to strip the silicone rubber out of the plenum and order some Permatex red from amazon.

https://www.permatex.com/products/gasketing/gasket-makers/permatex-high-temp-red-rtv-silicone-gasket/

Tested a third set of injectors (secondary injectors off a ninjx zx600) they are putting out 144 cc/min @43 psi which is closer to what I need for this motor.

[gpineau]

Cleaning the plenum of silicone rubber is not possible. It is bound for the trash can.  I am getting better at making these things. The first one took days, the second 4 hours. I made a new one in about 30 minutes.  Cant put it all together until the new gasoline resistant sealant arrives in 3 days.

Meanwhile I have chosen the 147 cc/minute injectors and I will be making a fuel rail to hook them up on the plenum.

[connie_rider]

I can feel your frustration..
     Keep tinkering your getting there.   

Are you planning to use an ECU from another bike to control the injectors etc?
If no, what do you intend to use?


Ride safe, Ted

[gpineau]

I can feel your frustration..
     Keep tinkering your getting there.   

Are you planning to use an ECU from another bike to control the injectors etc?
If no, what do you intend to use?


Ride safe, Ted

No, I think trying to make a preexisting solution to fit my particular problem may be more work than just making one from scratch. I would have to reverse engineer an existing system and figure out how to make it behave in my setup. Its possible but I would rather start from scratch and build something where I have complete control of the hardware. 

Like everything else I have done so far I will role my own. There are  multiple single board processors to choose from. Almost every semiconductor manufacturer has development boards to support their controller chips. 

I am looking at a 32 bit micro-controller that is supported by the latest offerings from Arduino.  I have made a preliminary design with this.

I am also looking at a controller chip from Texas Instruments.  They also offer development kits and support to get it up and running and so i can easily make changes.

Which ever controller I choose it must support C and python programming languages.

As far as the hardware for driving the injectors I am experimenting with a couple of chips from Texas Instruments a  TPIC46L02,  low-side predrivers, and a TPIC2601  a power DMOS array. I have samples of each. Just looking at the power driver it seems too small to dissipate the heat I think it will generate. But TI specs it to drive up to 6 injectors. The predriver is also tiny it will be a challenge to solder to those tiny little legs. 

I haven't given this a lot of attention yet because i want to get a solid mechanical fuel delivery system before I change focus to the electrical hardware.

[connie_rider]

Your way over my head on this....
Anxious to see if it all works out!!

Ride safe, Ted