Total Motorcycle Community Forums

In my ongoing attempt to start projects I don't have time to finish, I was looking at changing my Katana over to fuel injection via [url=http://www.megasquirt.info[/url]Megasquirt.[/url] For those of you not in the know, Megasquirt (MS) is a stand-alone ECU that controls fuel injectors. It can take a variety of sensors, including o2, MAF, TPS, etc... Infinitely tunable with a computer.

It seems to me I can convert my spare rank of carbs for the Kat into individual throttle bodies. I'd just gut the carbs, fill them so there is a nice smooth passage into the butterfly valve and bore angled holes into the carb past the butterfly valve for the injectors. After that, I would scavange a fuel rail from a car and some injectors. I'd also need to have a bung welded in the exhaust for an o2 sensor and figure out some kind of TPS setup.

I'm probably looking at about $500 if I get the pre-made MS kit and get my various other parts of ebay. I could sneak under that budget by talking to friends and friends of friends for parts or hitting up junkyards.

Does anyone have any thoughts on this? I just want something to tinker with over winter and this seems like it'd be pretty fun, plus I can move the MS setup to my next bike with little fabrication.

I think you should go for it.

You also need the MAF (Mass Air flow) Sensor. One for each throttle body.

It might be better to use one MAF and arrange for the carbs to pull all their air through it.

Seems like a lot of work though.

mydlyfkryzis wrote:You also need the MAF (Mass Air flow) Sensor. One for each throttle body.

It might be better to use one MAF and arrange for the carbs to pull all their air through it.

Seems like a lot of work though.

MAF is one of the options for a sensor. The MS kits look to have a MAP sensor built in already. It shouldn't be too hard to fab up a small airbox so I could use a common sensor for all four throttle bodies.

A simple plastic box about as wide as the carbs are now and three inches or so deep and four inches tall should be a sufficient airbox. I could then easily scavenge parts to mount a filter or four on the other side of it.

Skier wrote:In my ongoing attempt to start projects I don't have time to finish, I was looking at changing my Katana over to fuel injection via [url=http://www.megasquirt.info[/url]Megasquirt.[/url] For those of you not in the know, Megasquirt (MS) is a stand-alone ECU that controls fuel injectors. It can take a variety of sensors, including o2, MAF, TPS, etc... Infinitely tunable with a computer.

It seems to me I can convert my spare rank of carbs for the Kat into individual throttle bodies. I'd just gut the carbs, fill them so there is a nice smooth passage into the butterfly valve and bore angled holes into the carb past the butterfly valve for the injectors. After that, I would scavange a fuel rail from a car and some injectors. I'd also need to have a bung welded in the exhaust for an o2 sensor and figure out some kind of TPS setup.

I'm probably looking at about $500 if I get the pre-made MS kit and get my various other parts of ebay. I could sneak under that budget by talking to friends and friends of friends for parts or hitting up junkyards.

Does anyone have any thoughts on this? I just want something to tinker with over winter and this seems like it'd be pretty fun, plus I can move the MS setup to my next bike with little fabrication.

I'm actually an engine development engineer and I actually done injection systems.

Let me get this straight, you're going to buy an ECU and then throw together a bunch of spare parts, plug them in, and then map the engine.

And you've apparently never done this sort of thing before.

I can easily think of a million reasons you'll end up kissing the Katana goodbye.

Assuming you can get all the parts to plug in, do you have any idea how much dyno time you'll need to get the various engine maps?

As we say in the business, good luck with that.

Aquaduct wrote:

Skier wrote:In my ongoing attempt to start projects I don't have time to finish, I was looking at changing my Katana over to fuel injection via [url=http://www.megasquirt.info[/url]Megasquirt.[/url] For those of you not in the know, Megasquirt (MS) is a stand-alone ECU that controls fuel injectors. It can take a variety of sensors, including o2, MAF, TPS, etc... Infinitely tunable with a computer.

It seems to me I can convert my spare rank of carbs for the Kat into individual throttle bodies. I'd just gut the carbs, fill them so there is a nice smooth passage into the butterfly valve and bore angled holes into the carb past the butterfly valve for the injectors. After that, I would scavange a fuel rail from a car and some injectors. I'd also need to have a bung welded in the exhaust for an o2 sensor and figure out some kind of TPS setup.

I'm probably looking at about $500 if I get the pre-made MS kit and get my various other parts of ebay. I could sneak under that budget by talking to friends and friends of friends for parts or hitting up junkyards.

Does anyone have any thoughts on this? I just want something to tinker with over winter and this seems like it'd be pretty fun, plus I can move the MS setup to my next bike with little fabrication.

I'm actually an engine development engineer and I actually done injection systems.

Let me get this straight, you're going to buy an ECU and then throw together a bunch of spare parts, plug them in, and then map the engine.

And you've apparently never done this sort of thing before.

I can easily think of a million reasons you'll end up kissing the Katana goodbye.

Assuming you can get all the parts to plug in, do you have any idea how much dyno time you'll need to get the various engine maps?

As we say in the business, good luck with that.

Your input is appreciated, especially with your background.

I'm aware this isn't going to be a plug and play operation. I look forward to tinkering with it a bit if I do go ahead with the system.

In a short time of doing research on the subject, it looks to me most, if not all of my questions have been answered on the Megasquirt forums. That was just in a few days' worth of research, if I dig more and do some reading I think I can have at least some kind of project to play around with for a while.

If you'd care to elaborate on a few of those myriad of reasons, I'd enjoy thinking about some of the finer points I'm sure I am overlooking.

I predict this project will reduce your recreational riding time greatly. And keep your cell phone battery recharged.

Well, to be fair, I did check out the website and it doesn't seem to be a recipe for complete disaster as I first thought. It's apparently one of those tuner ECU's mostly used for hopping up import racers. It appears to come with at least rudimentary software, both control strategy algorithms and basic calibration tables. As the basis for a hobby kit ala old diode radios, it's probably OK.

As for the rest of the system around it, they offer competent information on how to build a basic system and get adequate component specifications and electrical interfaces. Of course being able to read it and actually find componentry are 2 different things.

And, ultimately, calibration for one objective (power) is a heck of a lot easier than multiple conflicting objectives like an OE (power, driveability, emissions, fuel economy, durability, etc.).

You'll have mega issues with space and, I'm guessing, finding adequate components. These are designed to go in cars and, ideally, already fuel injected cars that have supporting systems. Try looking under your seat and figuring out where your fuel pump will go. Things like available fuel injectors, for instance, that can commonly be found in the tuner market may be bigger than you need or can use. And what about a camshaft position sensor? That usually involves adding to or modifying the camshaft and getting sensor wires out of the engine. I wouldn't embark until you've found the needed pieces and have a good idea where they will go and how much they will cost. That $500 may be quite optimistic.

The one reason I wouldn't attempt it is you'll turn a servicable, reliable machine into a potential basket case. You've got no guarantee of durability and you'll have no way of knowing until you've been riding. There could be a number of thermal issues with sensors, for instance, depending on where you mount them and you'll probably not be able to get real good information surrounding that. And how do you measure heat rejection under the fairing at full throttle anyways? My Triumph gets hotter than blazes under there. It could be particularly bad if you scavenged junkyard car parts to save a few bucks.

It's just tough to say. Fuel injection takes a relatively simple and straight forward mechanical process and digitizes it. When done right, it's great. But the devil is really in the details. One thing I did notice is that when I went to the "Success Stories" section of the website and pulled up "Motorcycles", there didn't seem to be any that got beyond "Well I got it started today and rode around the block. Still need some tweeks." And those posts were a year or 2 old.

On the other hand, if you've got the bike to spare or you've always got a trailer at the ready, go for it. It'll probably be fun.

I doubt that you'll see much benefit (None worthwhile after spending a TON of time on this.)

You could probably fit GSXR fuel components if the throttle bodies are the same sizes as your carbs. (This seems alot easier than messing around w/ the carbs at any rate - plus you'll have the carbs to go back to if you mess things up) There are myriad of wrecked gsxrs out there so you should have no trouble sourcing these parts.

Benefits - you'll have a full rack of throttle bodies and a TPS sensor already in place. If you are lucky, stock gsxr injectors will have the proper flow rates, etc for your application. Plus the spare carbs 'just in case'

You're going to have to get a sensor inside the engine. That's going to be a PIA.

Haltech makes a fuel injection kit you might want to look into. The even make a wiring harness where you can splice in what you need where you need it, and their PC software looks easy to use. Downside is the price for this system is in the $1500 range.

I'm considering doing this to my BMW r80. Some problems that have been mentioned to me - With a 2-valve head, chances are I won't see much improvement over the carb setup. I may find more power, but at the expense of economy (of course) Electrics are also an issue. you will need to power a few extra things such as a fuel pump, injectors, etc. Make sure your stock system has enough amps left over to give you wiggle room for the injection setup.

At any rate= - Even if you cobble a system together out of used parts, I bet you are still looking at a minimum of $1500 to do it. I'm really not sure I want to tackle the problem myself, and my bike only has 2 cylinders w/ a dry clutch/ flywheel setup (Easier to do the sensor and whatnot)

Aquaduct wrote:Well, to be fair, I did check out the website and it doesn't seem to be a recipe for complete disaster as I first thought. It's apparently one of those tuner ECU's mostly used for hopping up import racers. It appears to come with at least rudimentary software, both control strategy algorithms and basic calibration tables. As the basis for a hobby kit ala old diode radios, it's probably OK.

As for the rest of the system around it, they offer competent information on how to build a basic system and get adequate component specifications and electrical interfaces. Of course being able to read it and actually find componentry are 2 different things.

And, ultimately, calibration for one objective (power) is a heck of a lot easier than multiple conflicting objectives like an OE (power, driveability, emissions, fuel economy, durability, etc.).

You'll have mega issues with space and, I'm guessing, finding adequate components. These are designed to go in cars and, ideally, already fuel injected cars that have supporting systems. Try looking under your seat and figuring out where your fuel pump will go. Things like available fuel injectors, for instance, that can commonly be found in the tuner market may be bigger than you need or can use. And what about a camshaft position sensor? That usually involves adding to or modifying the camshaft and getting sensor wires out of the engine. I wouldn't embark until you've found the needed pieces and have a good idea where they will go and how much they will cost. That $500 may be quite optimistic.

I'm not sure if the TPS is on the rank of throttle bodies on an R6, but I can get them to my door with injectors, fuel rail and other misc. sensors on it for $66 via ebay. I'd wager the injectors are small enough to feed my 600cc Kat.

I don't believe I need a camshaft position sensor for simple fuel injection. Most setups seem to use a TPS, wideband o2 sensor and a MAP sensor.

Aquaduct wrote:The one reason I wouldn't attempt it is you'll turn a servicable, reliable machine into a potential basket case. You've got no guarantee of durability and you'll have no way of knowing until you've been riding. There could be a number of thermal issues with sensors, for instance, depending on where you mount them and you'll probably not be able to get real good information surrounding that. And how do you measure heat rejection under the fairing at full throttle anyways? My Triumph gets hotter than blazes under there. It could be particularly bad if you scavenged junkyard car parts to save a few bucks.

It's just tough to say. Fuel injection takes a relatively simple and straight forward mechanical process and digitizes it. When done right, it's great. But the devil is really in the details. One thing I did notice is that when I went to the "Success Stories" section of the website and pulled up "Motorcycles", there didn't seem to be any that got beyond "Well I got it started today and rode around the block. Still need some tweeks." And those posts were a year or 2 old.

On the other hand, if you've got the bike to spare or you've always got a trailer at the ready, go for it. It'll probably be fun.

While I do currently have a bike to spare, it's a 20 year old UJM that has its own issues. Still runs and rides great, but I'd rather upgrade to something a bit newer.

TechBMW wrote:I doubt that you'll see much benefit (None worthwhile after spending a TON of time on this.)

You could probably fit GSXR fuel components if the throttle bodies are the same sizes as your carbs. (This seems alot easier than messing around w/ the carbs at any rate - plus you'll have the carbs to go back to if you mess things up) There are myriad of wrecked gsxrs out there so you should have no trouble sourcing these parts.

Benefits - you'll have a full rack of throttle bodies and a TPS sensor already in place. If you are lucky, stock gsxr injectors will have the proper flow rates, etc for your application. Plus the spare carbs 'just in case'

I can get some R6 throttle bodies with injectors, fuel rail and misc. sensor for $66 shipped to my door. As stated above, I think they would work alright for my 600.

TechBMW wrote:You're going to have to get a sensor inside the engine. That's going to be a PIA.

What is that for, now? No sensor needed inside the engine at all for just EFI.

TechBMW wrote:Haltech makes a fuel injection kit you might want to look into. The even make a wiring harness where you can splice in what you need where you need it, and their PC software looks easy to use. Downside is the price for this system is in the $1500 range.

I'm considering doing this to my BMW r80. Some problems that have been mentioned to me - With a 2-valve head, chances are I won't see much improvement over the carb setup. I may find more power, but at the expense of economy (of course) Electrics are also an issue. you will need to power a few extra things such as a fuel pump, injectors, etc. Make sure your stock system has enough amps left over to give you wiggle room for the injection setup.

At any rate= - Even if you cobble a system together out of used parts, I bet you are still looking at a minimum of $1500 to do it. I'm really not sure I want to tackle the problem myself, and my bike only has 2 cylinders w/ a dry clutch/ flywheel setup (Easier to do the sensor and whatnot)

$1500 is a bit steep for me, especially if I can get this done on the "cheap." Here's what I'm currently looking at (some shipping charges not included):

Megasquirt v2.2 ECU kit (all parts, case): $140
Stimulator kit (make sure the board works): $40
LC-1 wideband o2 sensor w/controller: $200
'04 R6 throttle bodies, injectors, fuel rail, sensors: $66

I then need to acquire a fuel pump, pressure regulator and possibly a TPS. Budgeting $70 for these items brings the grand total to $456, plus shipping from two different stores. Add in another $100 of misc. parts (solder, soldering iron tips, fuel line, fittings, wire, etc...) and I'm looking at $556. Better make the slush fund another $100 and bring the total to $656.

After that, I can apparently borrow my friend's PDA for datalogging, analyze it with the help of the Megasquirt forum and tune it with his laptop.

I really appreciate all the input, guys. I always get carried away looking at projects like these and don't see the pitfalls and whatnot. Please give me some more feedback about the response, thanks!