fuel injection?

[Anonymous]

I hate carbs! I cant stand them. I am getting ready to rebuild completely honda cb 750 and I am almost considering trying to fit some kind of a fuel injection into it. Anybody ever tried that? Any ideas, or web sites?

 

[humhead]

I agree wholeheartedly with you. I found this website http://bgsoflex.com/megasquirt.html When I was looking for answers to this question. It looks very promising. Best of all it's significantly cheaper than any other system I've seen for sale. The newsgroup for users has at least a couple of posters who installed this system on motorcycle engines. Fitting an older motorcycle with fuel injection will take a lot of time & patience.

 

[lecroy]

One of the family told me about the megasquirt. It is a very simple design, but like you said, low cost. Seems like a fun project to do.

 

[Anonymous]

ms

ms

I have a megasquirt built on the bench ready to go onto my gs750 when I finish rebuilding the motor using a single throttle body from a small car on the intake to a blower and this time just used the car tps etc, previously I ran a kawasaki 750 turbo fuel injection system on it along with turbo and sensors and that gave me a std 5v tps on the end of the throttle bodies. The spacing is ever so slightly different, but you can get away with ovalizing the turbo inlet mounting rubber holes and doing some port matching between the (mismatched) ports. The kawasaki gpz11 fi system is similar to the turbo one (but has no map sensor and the ecu is fitted with a barometer inside its case) but has smaller (!) injectors.
The entire turbo fuel injection system is designed as a bolt on for the stock gpz's, so runs its own sub loom etc. Unfortunately the ecu isnt reprogramable, but you can get a degree of alteration by altering fuel pressure (adjustable fpr) and by using the racemode thats built into the turbo computer. Sorry dont know if the 11 has this also.
You'll need a return for the fuel injection to your petrol tank pretty much whovevers system you use and a fuel injection fuel pump (and relays to control it). Usually the return is just a bung with a small straw so the bottom of the tank doesnt get too aeriated by the fuel tap intake.
The megasquirt itself, I found easy to construct, but I am quite electrically minded. And its a LOT cheaper than the likes of a DTA or similar. And its nice to hook the laptop up and *twiddle*
I hate carbs too..

 

[first timer]

do you have any more info, i would love to see some pics, nomadic

-ryan

 

[lecroy]

do you have any more info, i would love to see some pics, nomadic

-ryan

Agree!!

Your running a blown 750??

 

[humhead]

I have a megasquirt built on the bench ready to go onto my gs750 when I finish rebuilding the motor using a single throttle body from a small car on the intake

I always wondered why motorcycle injection systems didn't use a single throttle body. Please keep us posted on your project. I live in an area where the air is less than pure, so the idea of less pollution, in addition to improved driveability & less maintenance is intriguing. If FI increases power too, great!

 

[Anonymous]

I always wondered why motorcycle injection systems didn't use a single throttle body.

Because individual throttles make more power, higher efficiencies, better mileage and generally provide sharper throttle response.

I live in an area where the air is less than pure, so the idea of less pollution, in addition to improved driveability & less maintenance is intriguing. If FI increases power too, great!

Just installing a home-built FI system will not do much of anything for emissions, driveability or power. You would be hard pressed to match a well tuned set of carbs in all of these areas. A race-only FI system is not that hard to do because it only functions at one basic setting (full throttle). A street system is very much non-trivial and all of the OEM's (car and bike) spend unbelievable man hours on fine tuning their systems to work well under all driving conditions.

Mark (who doesn't think carbs are really all that bad... )

 

[Anonymous]

So lets see here, what all involves something like this.

First I have to find injectors. I know these are rated by gph. Any other considerations when choosing one?

Then I need a way to mount them on. So some kind of fuel rail that conects to each one of them, and then I would need to add a fuel pump to build the pressure. Or can I just leave it at gravit feed.

Then the way of mounting them to the engine. i was just thinking about cutting some pipe, and welding few lips on one end so that it screwes in where the carb boots go now. On the other end closing it and making a small hole where the injectors would go. Hm... but how do I mix it with air? How do I add air at all?

then something like megasquirt to control the flow of injectors in combination with probaly 4 O2 sensors. I wonder if megasquirt will only take one sensor?

Also from what I undersood from their website, all injectors would squirt at the same time WTF? Should there be some way to time it in relation to intake valves opening?

Ok, so as you can see I am only at the start line, and have a ton of questions.
Any ideas here, any comments would be apriciated.

Thanks to everyone who hates carbs, and Mark I hate you for understanding carbs. :lol:

I asked a friend of mine who's been a harley rider for over 50 years, few questions about carbs when i was rebuilding them. He toled me that carburetor is a french word for don't fuck with me 8O
He was joking and he did help me but I still think of french translation for carbs every time I see them.

 

[Anonymous]

First I have to find injectors. I know these are rated by gph. Any other considerations when choosing one?

You need to find ones that you can get specs on relating to fuel flow/sec of open time, opening voltage, operating pressures, etc. You need to know almost EVERYTHING about your injectors if you hope to scientifically arrive at the correct set-up. Or you can get some from a wrecking yard and tinker a LOT.

Then I need a way to mount them on. So some kind of fuel rail that conects to each one of them, and then I would need to add a fuel pump to build the pressure. Or can I just leave it at gravit feed.

It must have a pump, most injection systems run in the range of 40psi on the rail.

Then the way of mounting them to the engine. i was just thinking about cutting some pipe, and welding few lips on one end so that it screwes in where the carb boots go now. On the other end closing it and making a small hole where the injectors would go. Hm... but how do I mix it with air? How do I add air at all?

You must have a throttle body somewhere to regulate the airflow into the motor. Take a look at some pics of current bikes with injection to see how it should look.

then something like megasquirt to control the flow of injectors in combination with probaly 4 O2 sensors. I wonder if megasquirt will only take one sensor?

You will only need one O2 sensor. The sensor does NOT set the mixture as you open/close the throttle, it is only used to tune for drift over time due to wear or slightly different fuels (or extra correction for altitude changes and/or minor tuning changes if the controller does a bad job of it). The fuel delivery is controlled by settings you program into the controller before hand. This is where it becomes a monster to do at home. The hard parts are dead easy to set-up and get working. All of the magic is in the controller programming. For the simple system you are describing, it will have an rpm sensor and a throttle position sensor (this is the minimum to make it actually run), then a stored map that tells the injectors how much fuel to dump based on the rpm and throttle opening. The problem is that your map has to have a setting for every combination of rpm/throttle that it will run at. Preferably set at about 1/8 throttle increments and 250rpm increments. The only good way to get these settings is to run it on a dyno and tweak them to find the best ones. Even then, in real world driving you may find combinations of load/speed/throttle that give weird responses (rich/lean/surging/bogging, etc.) andf then you need to program and tune those out. It takes LOTS of hours and resources to iron out all of these things to get the seamless drivability of carbs or the current OEM systems, let alone low emissions AND good driveability, which can be difficult for even the OEM's let alone one guy in his garage.

Also from what I undersood from their website, all injectors would squirt at the same time WTF? Should there be some way to time it in relation to intake valves opening?

This is not uncommon on basic systems and works OK for the most part. When the revs are up there is an almost continuous flow of air through the intake, so an untimed spray of fuel is OK, too. This is basically what carbs do, as well. In an optimal world, the injectors are timed to fire each cylinder when the intake valve is open and air is flowing in, but experience has shown it's not critical.

Ok, so as you can see I am only at the start line, and have a ton of questions.

Ask away. It is an interesting project, but very challenging, too.

Mark

 

[Anonymous]

Attempt at answers (big)

Attempt at answers (big)

So lets see here, what all involves something like this.
First I have to find injectors. I know these are rated by gph. Any other considerations when choosing one?

The driving impedance, fittings and how they mount to the fuel rail/throttle body , seal to tbi/rail etc, how the wiring plugs in etc.

Then I need a way to mount them on. So some kind of fuel rail that conects to each one of them, and then I would need to add a fuel pump to build the pressure. Or can I just leave it at gravit feed.

They wont work at gravity feed, they rely on positive fuel pressure to force the fuel through that tiny nozzle and atomise the fuel. Baseline my turbo ran 42psi, but I know other people that ran 50+ psi of fuel pressure in a bid to enrichen the stock sized injectors to cope with other mods. The pump doesnt set the pressure, the pump should be capable of far more than you run, and you have a pressure regulator on the end of the fuel rail set to whatever fuel pressure you want to run. OEM systems come with a preset one, but theres lots of info on the net on how to make a adjustable one for not much at all, or you could always buy one.

You'll still need a return bung on your tank from the fuel rail too, and a car fuel filter (from a injected car) inline with the rail doesnt go amis either.
You can order injectors in barb top fitting then you can make a fuel line with t's off to each injector, but be sure to do a proper job with proper fuel clips and fittings, the slightest leak and your pumping fuel all over your motor. Better to get one made up, or use one off another vehicle (I run a kawa turbo rail).

Then the way of mounting them to the engine. i was just thinking about cutting some pipe, and welding few lips on one end so that it screwes in where the carb boots go now. On the other end closing it and making a small hole where the injectors would go. Hm... but how do I mix it with air? How do I add air at all?

Think of a set of carbs, but with no fuel components, air is regulated by butterflies still, and they come with somewhere handy to mount the injector to. Your now thinking of a set of throttle bodies.
The motor knows how wide you have the throttle (in a alpha-n system) because it can see the throttle position sensor output (which is really just a big volume control knob activated by the throttle butterfly pivot) and it takes these, along with barometric pressure, rpm and a few other factors (volumetric efficiency etc) and calculates a duration to hold the injector open for that cycle.
The megasquirt doc's are pretty good for reading up on how it all works. Ive said this before, but take a look at the kawasaki system, since thats a fi unit grafted onto what was originally a n/a carb'd motor by the factory.

then something like megasquirt to control the flow of injectors in combination with probaly 4 O2 sensors. I wonder if megasquirt will only take one sensor?

If your running a 4-1, I dont see the need for 4 sensors, if your doing this because you have a 4-2, then maybe your not at the extreme that something this involved is the next step.
If your running a 4-2 for other reasons then you will have to ask on the megasquirt list, im sure theres other people done the same thing on v8's and not combined the two banks into one or the like. I know for sure theres guys with aircooled vw's with the same issues due to their layout.

Also from what I undersood from their website, all injectors would squirt at the same time WTF? Should there be some way to time it in relation to intake valves opening?

Sounds horrendous doesnt it? except every motor I can think of uses batched fire injection, including most cars and the aforementioned 750 turbo and the gpz11 systems, both of which were pretty sucessfull implementations in their day. There really is very very little to be gained from all the complexity of timed injection on a per cylinder basis, and I believe even the likes of DTA and other commercial performance ecu's use batch fire methods. If you find a 2k ecu has batch fire, its pretty much done that its ok as a method.

Ok, so as you can see I am only at the start line, and have a ton of questions.
Any ideas here, any comments would be apriciated.

Its not a quick fix or a easy road, and you'll have some heartaches along the way, but it is possible. You will need a fuel pump, and associated control relays amongst other things and this is a big current draw if you run high pressures/volumes and have a marginal charging system.

Ask yourself this :-
Is this my only means of transport
Will I care if I cant ride my gs for a month (because it will always take you longer to fit the stuff than you think even if you have it all to hand and then you have tweaking time) (mine has been in bits for at least 6 of the past 10 years though some of that was cycle parts and some blowup repairs)
Am I going to get a big enough gain from all this to make it worth while (or would it be better to pay someone who is good with carbs to sort mine out)/ am I going to some extreme not covered by conventional methodologies/do I get a kick out of doing things unique.
Do I feel confident that my skills are up to this/have time to learn, or if not, that I can find the appropriate skills to do the bits I dont know about.
Be honest, if you have doubts now, at 2am when something goes wrong that means your off the road another week you'll curse the very thought of it.

My reasons for fitting fi was not purely because I had some bad carbs (but nearly being wiped out by my flatslides jamming on 3 occasions at wfo helped! ) more the fact that I was moving towards a more extreme motor step by step and carbs didnt fit too well for me. Plus Im impatient by nature and taking the carbs out to change out jets to fine tune it irritated the hell out of me! I dont care about emissions out of my '77 motor, I doubt theyre any worse than the original carbs were and I bet theyre a lot better than the worn out ones lots of people have now, but it wasnt my primary concern to put it mildly...

To the other posters:-
Multiple throttle bodies are indeed quicker to respond, and flow (marginally) better than one big throttle body (depending on intake/plenum design). Which is why mk1 used the 4 turbo individual tbi's when it ran the turbo too. However this time I have gone for a single because I have boost all the way from tickover, and Im limiting myself to 12psi anyway (well I say that now :twisted: ...) and with a blower its awkward to vent sufficient volumes of pressure when you shut the throttle. Having the tbi on the intake means it operates where there is a vacumn, a condition in which butterfly valves operate very well. No boost made, no excess inlet pressure no waste. Its why you see blown cars with bug catchers on. Also this time I have a intercooler and the pipework suits this implementation. Im not over worried about latency though, just have to suck it and see in the real world. Your mileage may vary, my bikes kind of a extreme example, though it is one of my streetbikes.

No pics at this stage Im afraid, but Ill see if I can take some soon, since I go back home in a few weeks along with my wife's birthday present a digi camera (I couldnt get away with buying her a gas analyser or some snap on spanners) Im working in another country to my toys :evil: , "running" a blower is at this moment in time probally the wrong term to use since the drives all over the workshop which is why I needed that crank/startermotor shaft dimension a bit back

 

[Anonymous]

Ask yourself this :-

Superb advice to anyone contemplating a serious project like this. Where I live our riding season is so short (6 months on a good year) I jealously guard my riding time. I am planning on doing a valve adjustment, jet kit install, new tires and getting the suspension rebuilt all in a one week period to allow riding for the rest of the summer with no other work aside from an oil change or two.

with a blower its awkward to vent sufficient volumes of pressure when you shut the throttle. Having the tbi on the intake means it operates where there is a vacumn, a condition in which butterfly valves operate very well.

Your operating conditions are very different from a naturally aspirated motor and using a single draw through throttle body makes a lot of sense with a turbo.

No pics at this stage Im afraid, but Ill see if I can take some soon,

You'd better, we are all waiting to see the monster. It sounds like a wild ride and a great project machine.

Mark

 

[humhead]

Just installing a home-built FI system will not do much of anything for emissions, driveability or power.

If this is true, why does virtually every auto sold in the U.S. have electronic fuel injection & a 3 way catalyst? Modern autos emit less than 1/20 the amount of pollutants compared to pre emission controlled autos.
Modern autos start easily, warm up quickly & run smoothly. Net power is at worst equivalent to pre emission engines.

A street system is very much non-trivial and all of the OEM's (car and bike) spend unbelievable man hours on fine tuning their systems to work well under all driving conditions.

Couldn't agree more. I understand fully this is an undertaking that requires lots of time and effort. Ther's no guarantee it will be successful. I have a second bike for experimental purposes, so the project won't leave me without a GS to enjoy.

Terry

 

[lecroy]

For a good time, try http://www.rceng.com
I was thinking something like a rag over the inlet and dripping fuel onto it would be more my taste!

 

[Anonymous]

If this is true, why does virtually every auto sold in the U.S. have electronic fuel injection & a 3 way catalyst? Modern autos emit less than 1/20 the amount of pollutants compared to pre emission controlled autos.
Modern autos start easily, warm up quickly & run smoothly. Net power is at worst equivalent to pre emission engines.

Cars have EFI for emissions regs, true enough. What I said is that simply bolting on a home-brewed system will do nothing. The OEM ones are very highly developed, with thousands of man-hours in them. Under those conditions, FI WILL give better performance, lower emissions, etc. I am not saying that FI is bad, I think it is one of the greatest advances made in cars in the last 25 years, but your home built system will not be equal to the one in my car. In power, driveability or emissions.

Mark

 

[humhead]

but your home built system will not be equal to the one in my car. In power, driveability or emissions.

I love it when someone says " It can't be done" Makes me work harder to prove them wrong. As I said before, I have a second bike for experimental use. I plan on getting everything else right first, then and only then add fuel injection. The megasquirt is currently based on an 8 bit motorola cpu. This is similar to mid to late 80's auto FI systems. By the time I'm ready to install, a 16 bit system will be available. Besides I'm not comparing whatever system I install to the one in your car. My bike has mikuni roundslides, not modern auto FI. It's a lot easier to improve on 25 year old carbs that require regular attention & still have shortcomings in all 3 areas. I have some experience tuning late eighties FWD Mopars.
They have a similar system. If I can match that, it will be an improvment in all 3 areas. Best of all, I enjoy working on challenges like this.

Terry

 

[Anonymous]

8 bit

8 bit

The kawa gpz's were 8 bit cpu's too. I cant see fuel map step resolution being uber critical enough to matter. Poster who said about the o2 sensor at cruise and correction only isnt quite right, the megasquirt can be placed in full closed loop operation if so desired.
The kawa system doesnt even run a o2 sensor though, they just have the maps on a rom (but including a secret race mode which when activated has a alternative fuel maps set up for a higher state of tune.)

You'll never match a modern auto FI system for emissions, but only because they tune for emissions at the expense of driveability and power because they HAVE to. They use the o2 sensor to run the motor lean to achieve this, not at stoich as you'd think. Thus to these ends, their hands are tied in a way that a amateur tuner's arent. My bike certainly wont have to pass any sort of emissions testing to be legal on the road (it has to pass a "smoke" test, that is if it doesnt make too much blue smoke visually its ok) , so I can tune for rideability and power. This goes way back too, the gpz turbo fi system was made over lean for the us market to pass the emissions legislation there (and we're talking 1985 here), and tweaking them to run more fuel pressure and a couple of mods to get round the changes results in a sweeter running and idling and better behaved bike more akin to the euro version bikes with actually a net gain in road manners and power (and less likely to blow the back out the surge tank in extreme conditions).
Manufacturers can be improved upon in respect of the specific areas you are interested in, despite the odds because although they have these huge R&D budgets and expertise, they also have a whole bunch of compromises foisted on them by budget constraints, legaslatitive, market perception and other factors.

Besides, im hoping its not going to stay at one throttle opening long enough to worry about anything but hanging on :twisted:

 

[Anonymous]

I love it when someone says " It can't be done" Makes me work harder to prove them wrong.

Go to it. I would be very interested to see your results. It's a very cool project, just a HUGE very cool project.

If I can match that, it will be an improvment in all 3 areas. Best of all, I enjoy working on challenges like this.

If you enjoy the work and tinkering, it will not matter if it is totally successful, because you had a good time getting there anyway.

The kawa gpz's were 8 bit cpu's too. I cant see fuel map step resolution being uber critical enough to matter. Poster who said about the o2 sensor at cruise and correction only isnt quite right, the megasquirt can be placed in full closed loop operation if so desired.

The map resolution can be critical to driveability issues. Otherwise, it is not that important. For your turbo set up, you could probably map it every 500rpm in the high torque area and it would be fine.

Even full closed loop operation only uses the O2 sensor for small corrections, the sensors do not respond fast enough to have them correct for throttle opening changes. Full closed loop systems still run from a map for fuel settings and then the O2 sensor will tweak it a bit if it sees lean or rich at steady cruise.

True enough about the tuning of our bikes compared to OEM autos. The OEM's are hamstrung with regs from the govenrment, but my comments about emissions were based on the original post that said emissions were an issue. As I mentioned previously, a full race FI system can be done much more easily than a fully refined street engine version. But just don't expect to match a well set up set of flatslides for power, or driveability in a street vehicle with a home-brewed injection system. It's all of the light-load, part throttle stuff that can be a bitch to get right. The full throttle settings are usually easy.

There is a reason why Yamaha ran carbs the first season of MotoGP - because they couldn't get the same smooth response from FI as well tuned carbs.

I expect Terry will get a system to run OK if he sticks with it, but it's the final refinement and polishing that is so hard. I know we got a turbo'd 600 Ninja engine to run on the Formula SAE car when I was in university and we ballparked the initial FI settings from calculated airflows, with no bench testing at all. It never ran completely right while I saw it, but it DID run OK with a bit of tuning on a dyno. I expect that is where this will end up.

Good luck with it, it should be a fun, educational project.

Mark

 

[humhead]

Mark
I do appreciate your words of caution. It sounds like you've done this sort of thing more than once. I like to challenge myself. Exercise for the brain. Even if I'm not completely successful, I'll learn a lot.
[/quote]but it DID run OK with a bit of tuning on a dyno

This is encouraging....... There's a motorcycle shop with a dynojet less than ten miles from home.
My benchmark isn't a late model auto. If I can match a mid to late eighties FI system, that will be a big improvment. Remember, I'm starting with 25 year old roundslides (ugh!)

 

[Pete's GS]

EFI works well in cars because it has had a lot of R & D under the pressure of the EPA regs, which I'm all for. Motorcycle manufacturers seem to have taken longer to get EFI dialed in, though. There were complaints of "digital" feel and dead spots. It appears to be getting better throughout the industry now but when you think an engineer has to make EFI work from 0-15,000 rpm in all sorts of weather and air densities and engine loads, you have to be impressed. It's still hard to beat a good set of CV carbs though.