EFI conversion?

[V8Pinto]

Anybody on here ever convert to fuel injection?  I know a guy that is doing a Megasquirt install on his Falcon so the experience is nearby....

Just pushing around ideas.

[bbanjo]

I haven't heard of anyone doing it, but think that if someone had the time/spare bike/etc that a unit off of a new SV twin would be sweet to start with.

I don't know enough about them to say if they even have a chip that can be re-mapped to fit the bill.

I know a bunch about the units in H2O Vw's, but don't know beans about the units on any bikes. (pumps vs gravity fed fuel, injector types, PSI on the fuel rail and injector)

I'd imagine chopping a unit off of an inline four with EFI would be close to the beginning to setting one up. Maybe half of a V4? It would beat the crap out of re-jetting carbs.

Class?

[yamahonkawazuki]

if you can ifgure this out i will be interested in doing the same thing.

[Blueknyt]

im playing with that idea myself, nothing close to working yet.

[werase643]

I have the complete set-up for a TL1000R.....with instruments and some of the wiring.....had it for about 2 yrs now.  BTW, the throttle bodies are 52mm!!!!!!
I was trying to think how to trick the computer into thinking a DR650 was a twin and have some fun on my thumper...back burner now.....

[Blueknyt]

hmmm, 52mm throttle body, into a manifold, plenumed into both intake tracks? sound reasonable?

[JamesG]

subscribe to the www.mica peak.com  GS-TWINS or CHASIS email list.

Ask Dana Goulston about it. He just did a megasquirt conversion on his GS450 land speed bike.

[werase643]

hmmm, 52mm throttle body, into a manifold, plenumed into both intake tracks? sound reasonable?

a TL-R has TWO 52mm throttle bodies......

[Blueknyt]

a TL-R has TWO 52mm throttle bodies......

1 TLR cyl has the displacement of how many cc's? 1/2 of 1000 is ?

so 1 52mm throttle body may be too big for 2 smaller breaths through the airhorn. maybe something like a single 46mm throttle body would be the way to go.

[Rashad]

I would LOVE to do this if i had some more info. That would rock.. in fact my buddy and I were tossing around the idea, since hes a car genious anyhow... he even explained how simple it would be to do... simple, not easy.

[Rashad]

I would LOVE to do this if i had some more info. That would rock.. in fact my buddy and I were tossing around the idea, since hes a car genious anyhow... he even explained how simple it would be to do... simple, not easy.

[crazyfrog]

Hello. About suzuki GS500 EFI  you can visit my blog www.suzukiGS500efi.blogspot.com. The work it is in progress.

[Dr.Sparkie]

Hello. About suzuki GS500 EFI  you can visit my blog www.suzukiGS500efi.blogspot.com. The work it is in progress.

this is exactly like what I did. make sure you sample the MAP sensor synchronously with the spark. best bet is to use the MAP sensor on the #2 cylinder, sampled at the #1 spark. this will result in two readings for every otto cycle, discard the highest (because its not during the intake stroke) and you have an excellent MAP reading.
it took delco (or whoever, buti think it was delco) 2 years to get MAP injection on harleys because they couldnt solve the intake manifold pressure ripple problem.. i solved it independantly in 1 day, of which several hours was spent pissing and moaning about using a conditional branch statement in an interrupt... eventually i sucked up the 7 microsecond loss and it ran farkin awesome.

thread related: yeah, i injected my bike, but the generator couldnt support the fuel pump (honda civic pump), so i pulled it off. I grafted two dodge turbo minivan injectors onto the head, used the carbs as throttle bodies, a nissan FPR, honda civic pump, and homebrewed the controller out of an HCS12 and some driver electronics. it worked well, but required a car battery or a 12 power supply to run on the bike because of that damn fuel pump.

i wanted to try some snowmobile throttle bodies.. theyre usually cheaper than bike parts (on ebay) and theyre almost always twins, and they're always more than enough in terms of air and fuel flow.

[Chanse]

been asked  a hundred million times, attempted about four. When you figure it out let us all know...

[Kurlon]

I wouldn't be so worried about Alpha-N not being streetable, the vast VAST majority of EFI bikes use Alpha-N setups thanks to them being FAR simpler.

Otherwise, nice work so far!

[crazyfrog]

I wouldn't be so worried about Alpha-N not being streetable, the vast VAST majority of EFI bikes use Alpha-N setups thanks to them being FAR simpler.

Otherwise, nice work so far!

I didn't know that.


PS: sorry for my English skill (it is the second language for me).

[crazyfrog]

Hello. About suzuki GS500 EFI  you can visit my blog www.suzukiGS500efi.blogspot.com. The work it is in progress.

this is exactly like what I did. make sure you sample the MAP sensor synchronously with the spark. best bet is to use the MAP sensor on the #2 cylinder, sampled at the #1 spark. this will result in two readings for every otto cycle, discard the highest (because its not during the intake stroke) and you have an excellent MAP reading.
it took delco (or whoever, buti think it was delco) 2 years to get MAP injection on harleys because they couldnt solve the intake manifold pressure ripple problem.. i solved it independantly in 1 day, of which several hours was spent pissing and moaning about using a conditional branch statement in an interrupt... eventually i sucked up the 7 microsecond loss and it ran farkin awesome.

thread related: yeah, i injected my bike, but the generator couldnt support the fuel pump (honda civic pump), so i pulled it off. I grafted two dodge turbo minivan injectors onto the head, used the carbs as throttle bodies, a nissan FPR, honda civic pump, and homebrewed the controller out of an HCS12 and some driver electronics. it worked well, but required a car battery or a 12 power supply to run on the bike because of that damn fuel pump.

i wanted to try some snowmobile throttle bodies.. theyre usually cheaper than bike parts (on ebay) and theyre almost always twins, and they're always more than enough in terms of air and fuel flow.

hello.

So what you that to say it is that the MAP sensor return a fake value at "combustion process" ?

[Kurlon]

PS: sorry for my English skill (it is the second language for me).

Your english has been fine. : )

[Dr.Sparkie]

hello.

So what you that to say it is that the MAP sensor return a fake value at "combustion process" ?

sort of. because there is little volume behind the carb/throttle body in the "intake tube", the pressure read by the MAP sensor ripples alot, especially at low throttle applications. think of this volume as an "air capacitor", and its too small to smooth the pressure (voltage) ripples. if you continously sample the MAP sensor voltage  as a function of time, the time samples will show you this ripple signal.

but if you sample the MAP sensor synchronously with spark, you get the pressure signal at the same point every otto cycle. the reason I reccommend sampling #2 @ the #1 spark is because when the #1 cylinder sparks the number two cylinder is either coming to the end of the intake stroke, or coming to the end of the power stroke. The higher sensor reading will be at the end of the power stroke, because the air capacitor will have had time to "charge" through the throttle plate without being bled down by the open intake valve.

I have some sample code for the motorola HCS12... somewhere, just tried looking and the best i could come up with is some small assembler humor i put in another project:

SPRKMRK EQU     %100        * Sparkie gets 100%

you have to be creative with varible names with as11. (the percent sign indicates that the input is actually binary, so spark-mark is actually 4 decimal.... which isnt a bad GPA )

[crazyfrog]

hello.

So what you that to say it is that the MAP sensor return a fake value at "combustion process" ?

sort of. because there is little volume behind the carb/throttle body in the "intake tube", the pressure read by the MAP sensor ripples alot, especially at low throttle applications. think of this volume as an "air capacitor", and its too small to smooth the pressure (voltage) ripples. if you continously sample the MAP sensor voltage  as a function of time, the time samples will show you this ripple signal.

but if you sample the MAP sensor synchronously with spark, you get the pressure signal at the same point every otto cycle. the reason I reccommend sampling #2 @ the #1 spark is because when the #1 cylinder sparks the number two cylinder is either coming to the end of the intake stroke, or coming to the end of the power stroke. The higher sensor reading will be at the end of the power stroke, because the air capacitor will have had time to "charge" through the throttle plate without being bled down by the open intake valve.

I have some sample code for the motorola HCS12... somewhere, just tried looking and the best i could come up with is some small assembler humor i put in another project:

SPRKMRK EQU     %100        * Sparkie gets 100%

you have to be creative with varible names with as11. (the percent sign indicates that the input is actually binary, so spark-mark is actually 4 decimal.... which isnt a bad GPA )

wow. you really know a lot. Thanks man, and do not bother with the code. To save time i use mikroC to program the controller.

In fact you are right , the intake acting like a capacitor.

My plan was to read continuously the pressure sensor between two sparks and use the highest value, but probably is not very good.

thanks.

PS: i will update the blog soon.