EFI

[rubberneck]

I am curious about the possibilities of a Mass Air sensor....

 

I was just thinking about that myself. I had a guy tell me that it wont work well on a 2 stroke, but i dont see why. Hell people have said that injection wont work well unless direct injected, and i think that has been proven false so who knows. Who is going to be the first to try

[bbcmudtruck]

Very interesting thread and EXCELLENT write up! I wish more "out of the box" projects were documented as well as yours. Please keep us updated!

 

 

[Coupelx]

piss on mass air meters. with a megasquirt you have better options for a small engine, especially with ITB's.

 

 

 

Have you decided what strategy (Speed density, Alpha-N or Hybrid Alpha-N) you will run?

[mk2dubbin]

piss on mass air meters. with a megasquirt you have better options for a small engine, especially with ITB's.

 

 

 

Have you decided what strategy (Speed density, Alpha-N or Hybrid Alpha-N) you will run?

 

having used both strategies on my car; in my opinion - alpha-n is really the only way to go. speed density is nearly useless for an itb setup because the throttles open so quickly and nearly instantly take your engine to WOT. on top of that you will usually have a much higher idle vacuum pressure (higher in kpa, i.e. closer to atmospheric) so your effective control range is crap to begin with. i have no idea what the timing curve looks like on a banshee but if its a 3d curve, the speed density will miss a lot of the table.

 

hybrid is more useful but will mostly be effective on an engine with boost, again because the control range of the speed density part of the equation is so ineffective.

[Coupelx]

having used both strategies on my car; in my opinion - alpha-n is really the only way to go. speed density is nearly useless for an itb setup because the throttles open so quickly and nearly instantly take your engine to WOT. on top of that you will usually have a much higher idle vacuum pressure (higher in kpa, i.e. closer to atmospheric) so your effective control range is crap to begin with. i have no idea what the timing curve looks like on a banshee but if its a 3d curve, the speed density will miss a lot of the table.

 

hybrid is more useful but will mostly be effective on an engine with boost, again because the control range of the speed density part of the equation is so ineffective.

 

good to know. i figured map would be pretty useless on a NA engine. although there is the option of individual map sensors but i have never messed with it.

[THE358BANSH]

piss on mass air meters. with a megasquirt you have better options for a small engine, especially with ITB's.

 

 

 

Have you decided what strategy (Speed density, Alpha-N or Hybrid Alpha-N) you will run?

 

To start it will be speed density. All the different methods have been proven to work in 2-stroke and 4-stroke applications. Each just has their little qwerks... I think the speed density may provide a touch more cleanliness to the part throttle areas, but once it's WOT it doesn't matter much just get the fuel in and go!

 

 

having used both strategies on my car; in my opinion - alpha-n is really the only way to go. speed density is nearly useless for an itb setup because the throttles open so quickly and nearly instantly take your engine to WOT. on top of that you will usually have a much higher idle vacuum pressure (higher in kpa, i.e. closer to atmospheric) so your effective control range is crap to begin with. i have no idea what the timing curve looks like on a banshee but if its a 3d curve, the speed density will miss a lot of the table.

 

My only experience with a deal like this showed a useable range for MAP to be about 75-100kpa. In all reality that's only a range just under 4 inches of vacuum to work with. The throttle bodies have a map specific port on them with a smaller diameter hole than the syncing port. I'm thinking the R6 was using the small hole as sort of a oscillation damper, along with only being a signal port connected to both throttle bodie instead of a crossover tube. The stock ignition curve is fairly simple. 17 degrees up to 2000 rpm, advance to 22 degrees at 3500 rpm, then retard to 9 degrees at 10000 rpm. So far I will the pig started on the stock curve, then start playing from there...

 

I called David at Noss Machine today and ordered domes and orings. Top end and clutch side of the engine is together except for the head. The Microquirt finally showed today, and I spent the better part of the night playing with it. I messed up since it has been so long since playing with one and accidentally corrupted the firmware. No big deal though, I just used it as an excuse to upgrade to the latest version. I also pulled apart the box and moved the jumpers on the ignition drivers to convert the box to "logic level outputs". The jumpers are the little black box shaped connectors located in between the heatsink and the AmpSeal connector. Basically, the EFI can control a standard two post coil when the positive terminal gets +12vdc and the box then grounds the negative terminal to discharge the coil. The only issue with a setup like that is it is extremely noisy and the chances of having a RFI related issue are large, especially in a compact unit like the MicroSquirt. In comes the LS2 coils with their built in ignitors. You provide grounds and +12vdc to the coils themselves, and then use the EFI to trigger a +5vdc low current signal to charge the coil to maximum dwell. When the +5vdc is removed the coil discharges. The coil is almost like a relay of sorts with the EFI just controlling the switching side of it. Anyways, all in all better because of the tight fight and close quarters in the Banshee frame.

 

I tossed in a few more photos just for everyone to see what's inside the magic black box. It's funny how all the electronics to run fuel and ignition are located on a board the size of a deck of cards. Technology is so cool... The last photo doesn't look like much, but it means huge. I rough wired everything together on the table to make sure all sensors were legitimate and functioning, along with the O2 sensor also. That gave me a chance to let everything communicate between my 'puter and the EFI. Other than my little firmware issue it went smooth, and I spent the rest of the night configuring and building a rough tune so when everything is installed I won't have to bother doing it then. I still have to verify a few settings and thoughts with JohnstonCountyShee about the coils, and then we will be in business. And as a side note, butt connectors and other nasty cheap electrical pieces located in the last picture won't be used in the final install. Cut, Strip, Wrap, Solder, Shrink, Repeat, Cut, Strip, Wrap, Solder, Shrink, Repeat, Cut, Strip, Wrap, Solder, Shrink, Repeat, Cut, Strip, Wrap, Solder, Shrink, Repeat...Later, Evan

 

[Trick2stroke]

This is a fantastic thread, very well documented, and great pictures. Also, it is inspiring, I don't know anything about efi but it is tempting me to learn more and attempt such a set up, as I tend to dislike carbs!

 

I can't wait to see the finished product.

 

Do you mind telling what the ballpark range of cost on this project is?

[jasons]

I recommend a Thread Pin to keep this at the top.

[J-Madd]

I know almost nothing about EFI, but have a few questions. Will an O2 sensor work on a 2-stroke? What are you gonna use the sensor for; I thought with a MAP the amount of fuel injected was dependant upon than, not on an O2 reading?? Is it a narrowband O2 sensor? Sorry for my ignorance, I know you have answers to all these questions.

[THE358BANSH]

This is a fantastic thread, very well documented, and great pictures. Also, it is inspiring, I don't know anything about efi but it is tempting me to learn more and attempt such a set up, as I tend to dislike carbs!

 

I can't wait to see the finished product.

 

Do you mind telling what the ballpark range of cost on this project is?

 

If you want to learn more about the Microsquirt controller, visit the homepage at MicroSquirt.info. There is a metric shit ton of information there and it is all good. Read the website front to back twice, and then you might be prepared to start asking questions. As for monetary outlay, I figured this would cost right around 1k for all the parts to do the conversion. Mind you, this doesn't include anything for labor time spent doing or researching the conversion. So far I am really close to being on budget. I have spent a little more on the rebuild I am doing at the same time, but not counting any of that junk since I had to do it weather it stayed carb'd or EFI.

 

 

I know almost nothing about EFI, but have a few questions.

 

Will an O2 sensor work on a 2-stroke?

Yes, but there are things to be mindful of. An O2 sensor measures the amount of oxygen in a system, not fuel. The sensor itself eventually will contaminate if left in the direct stream of air/fuel/oil. The solution for this is an extended bung for the sensor to mount in that moves the sensor out of the pipe a bit. It still reads fine, it just isn't subjected to the rigorous uses in 2 stroke or high heat situation like boosted motors. Steve Murphy, who did a similar conversion in 2007 on his RZ is still using his first sensor the he installed when he did the conversion.

 

What are you gonna use the sensor for; I thought with a MAP the amount of fuel injected was dependent upon than, not on an O2 reading??

Depending on how the EFI is configured will base how the O2 is used or not. In open loop mode, the O2 reads and outputs an air/fuel ratio but the EFI doesn't use the value to correct the base VE map. When the EFI is in closed loop fueling, the O2 reading is compared to a desired air/fuel ratio table and then uses a percentage to add or subtract for that specific map/rpm cell. As a last option, MicroSquirt offers a system called Automatic Mixture Control, which is basically the same thing as the new 4-strokes AutoTune feature. As the EFI is correcting for a value in closed loop, it modifies the base VE map to essentially tune the engine itself. There is a time and place for AMC, and used as a tool it is awesome. When used as a crutch because of lack of initial tuning, it can be disastrous.

 

As for the amount of fuel injected, using a speed density algorithm the main factors are manifold absolute pressure and rpm to determine a VE number. From there, a multitude of different corrections and/or modifications are made using all of the EFI's available sensors to calculate the next fuel injection event.

 

Is it a narrowband O2 sensor?

Negative Ghost Rider. The Innovate LC1 lamba cable and sensor is a wideband unit capable of reading 7.35-22.39AFR on gas, and then outputs a 0-5vdc signal that the EFI uses. It can be configured for narrowband usage and ouputs, but I won't be using them in my case.

 

Sorry for my ignorance, I know you have answers to all these questions.

 

See answers in red. No problems on asking questions, that's how people learn...Back to work on the quad boys!

[J-Madd]

[wfobasheeboy]

This post rocks. Keep us updated.

[THE358BANSH]

Glad everyone is enjoying this project as much as I am... I loaded the rolling chassis in the pickup and took it to the car wash to clean it up. I needed to get the motor and everything else mocked up one final time so I can decide where to mount the coils, ecu, and lambda cable. Needless to say, pressure washing with a wind chill of -1 degrees makes things a little chilly. Got back to the shop and decided I was sick of bending over or kneeling on the floor so I found a little stand in the back room to put the chassis on. it's kind of like the burgard cycle deal, just free so way cooler.

 

I noticed I never took any decent pictures of the intake boots so here they are below. The right one is ported to the size of the reed stuffer, I just have to finish the little notch on the bottom for the injector spray path. The engine went in smooth, just waiting for domes that should be here tomorrow. Next on the list was mounting the ecu and coils. Both of these have to be placed well, with a few important things to keep in mind. RFI, wire routing, moisture, etc. are some of the big ones. I found a place suitable for the coils that I was happy with and built a little tube bracket to hold them. It turned out sweet and simple.

 

Didn't get a chance to mount the ecu. So far I can't make up my mind as to exactly where I want to place it. Basically we've got three options, back by the battery, sandwiched in the area between the tank and seat above the throttle bodies, or in between the radiator and steering loop bracket. Back by the battery isn't bad, it's just the leads for the wiring get long and have to be in close quarters with the exhaust on the trip to the front. Mounted behind the gas tank makes all the leads nice and short, it just dumps the wiring right on top of the throttle bodies and makes it kind of cluttered. Option three by the radiator has probably the best pro/con ratio. The ecu would be out of the way of mud/dirt/moisture. The noisy leads for the coils would separate from the harness almost immediately so chances of RFI start to dwindle. Only real downside is space is at a premium up there. I played around for a while, and I figure I can build a spacer to shift the radiator forward about 1" on the top to give me adequate airlfow/clearance to the ecu. On the plus side my front Lakers have enough clearance to do this also. As it's rated the box can handle 105 degrees celcius~221 degrees farhenheit constant. I've never had a temp sensor in this engine so I am going by what I have heard, but most scenarios seems like 200 degrees is pretty hot for one of these engines. I have only heard of big motors getting that hot when putting around in first gear or so. What's the norm for a ported 4 mil motor mixing 50/50 fuels with a nice aluminum radiator? Anyone? As soon as I burn the coil mounts permanent and make a decision on the ecu mounting I'll get on starting wiring this pig. Hopefully by then the domes and HID kit should be here...

 

[Larry's Shee]

Nice coil bracket. Welding it in place? Coils fire at same time or alternately? Very interesting thread. How is the cost-time to benifit rotio going?

[Larry's Shee]

Also, what did you do to cases and jugs to get them that clean?