Jump to content
Sign in to follow this  

'90 DR350s Fuel Injection Conversion

Recommended Posts

After reading other posts and having an increased interest in fuel injection lately, I've taken the plunge towards fuel injecting my '90 DR350s. This is not the first DR350 FI conversion, wildwestsidney did it some time ago on his later model e-start. My bike is kick start only, so this might be the first kick start DR FI conversion.

Why am I doing it? One reason: I want to. There are plenty of people who think this it's crazy to FI a nearly 20 year old dual sport, but it's my bike and I like working on it. It would be a lot easier to buy a newer fuel injected bike, but I just like the idea of doing it myself. What I really like is the ability to adjust air/fuel mixture exactly how I want it with just a bit of software. Carbs can do a great job, but they are what they are. One downside is that I will have to put a battery back on the bike, but that's a small sacrifice :banana:.

There are a lot more details to come, but here's what I've got:

  • 2003 Yamaha YZF R6 throttle body
  • 2002 NINJA ZX6R FUEL PUMP
  • 2000 HONDA CBR929RR FUEL INJECTOR RAIL
  • Innovate DB wide band O2 sensor/gauge (on the way)

The R6 throttle bodies were a great deal from eBay, one bore was missing a slide and cap, so no else wanted it. It's in great shape and has plenty of extra goodies such as TPS, fuel rail, injectors, pressure regulator, and a bunch of hoses/clamps. I got the CBR fuel rail for the injectors just in case the R6 injectors weren't enough. That came with a pressure regulator also. So now I have 8 injectors total, plenty of spares in case there's a problem.

IMG_0068.jpg

Probably the single biggest challenge is adapting the throttle body into my bike. It's hard to find something that will fit exactly. The length of each R6 TB bore is slightly shorter than the DR stock carb, the cylinder side port is 10mm larger in diameter, the air intake side is maybe 20mm smaller. A TT member was gracious enough to machine some adapters out of aluminum, which was a great help. I ordered the TB cylinder boot for the R6 just in case. Captain Midnight probably gave me the best advice of all: just get it mounted one way or another, duct tape if necessary, just to get it running. Once it's running there will be plenty of inspiration to keep going and improve things. I'm not making any permanent changes to the bike, so I can go back to the carb in case I run into trouble or just give up :banana:

The pump is a bit bigger than I would expect, about the size of your fist. I guess that's what it takes to generate 40+ psi at a decent flow rate. It's going to take maybe five amps to run the pump, I'll use a salvaged computer power supply (12 V/15 A) to run things until I'm ready to wire it into the bike (no battery yet).

I just ordered an Innovate DB wide band AFR gauge kit:

http://www.innovatemotorsports.com/products/db.php

DBkit_dual_700.jpg

That will go in first, I'm really curious to know the exact mixture I have now with the carb. At $190, that has been the most expensive part to date.

For those of you interested in FI, I strongly suggest reading the Megasquirt online manual:

http://www.megamanual.com/mtabcon.htm

I went through all the theory and verified the math (I have a degree in Physics, what do you expect? 👍). I've even gone as far as studying the embedded source code (I used to design embedded systems for satellite instruments, high end receivers, and a bunch of other stuff). You don't have to do all that, but the information is there if you are interested. At the end of the day, EFI is not nearly as complicated as you might think. Using the speed-density method, you basically measure air temperature and manifold pressure to compute injector pulse width for each cycle. It comes out to a few millisecond pulse each cycle, increasing with higher MAP, decreasing with higher temperature. It's basically a multiply and a divide, with some correction factors for warm up enrichment, acceleration enrichment, volumetric efficiency etc. Those values can be tuned as needed.

That's all for now. I'll post more details as things progress. Keep in mind that I am not independently wealthy, I've got a wife, and two small children to keep me busy. This is a long term project, but I've gone a lot further than I expected when I started researching the whole thing. I can see everything coming together, so I'm hopeful. The weather is getting colder here in MD, so I probably won't be riding much anyway :busted:.

  • Like 2

Share this post


Link to post
Share on other sites

Cool project. Keep track of the total cost as you go along, some other members might want to try the same thing if it is reasonable enough.

Share this post


Link to post
Share on other sites

I thought I would mention a TB database that HeadTrauma pointed me to:

As far as a throttlebody database, one was started a while back on the msefi.com forums here.

Most eBay sellers don't want to measure TBs, so it's hard to know what you are getting.

In terms of costs, here's the breakdown:

  • R6 TB: $10 + $15 shipping
  • R6 intake boot/clamp: $29 total
  • Fuel pump: $21 + $14 shipping
  • O2 sensor/gauge: $190 + $0 shipping

The Megasquirt controller is $140, haven't purchased that yet.

The grand total will come to something like $420 or so. It's more than a new carb, but it's hard to beat EFI flexibility. You could knock off $190 for the wideband O2 sensor/gauge and cut it down to $200. You can run without it, but I would really like to know what the mixture is.

Share this post


Link to post
Share on other sites

just another audience member. I recently purchased a '94 dr350 as a "toy" and one of the projects I was thinking of toying around with is fuel injecting it. I just wish I had to ability to weld as it would make fabricating stuff up so much easier ... but hey, maybe the duct tape is the best suggestion.

I'll be curious what you use for a signal/trigger wheel, that always seems to be the interesting fabrication part of these types or projects I think ...

Share this post


Link to post
Share on other sites

I've made more progress, but I'm tied up with work for a week or more. I'll post some of what I have soon. I think it will come together nicely.

In terms of trigger, I plan to inductively couple to the low voltage side of the ignition coil. I've got a bunch of ferrite toroids I can wind a few turns of wire and slip over the one of the low voltage coil wires. I'll measure the output voltage to make sure it's easy to detect, but I see no reason why this won't work.

Share this post


Link to post
Share on other sites

Is there a different model throttle body without a vacuum slide? Do you expect the throttle response on the R6 TB to be similar to the stock DR350SE (vacuum) carb?

Share this post


Link to post
Share on other sites

I suspect that in a few years CRF250R fuel injection parts will be floating around and will probably be ideal for such projects as these:)

There are many street bike TB's that don't use the goofy CV bits - the only ones that did are the first generation yamaha FI systems to my knowledge. Finding one that's small enough to work well with the small port on the DR head might take a bit of digging around - most nowdays are in the 40mm range, even on 600cc I4's - workable I'm sure but perhaps a bit larger than ideal. Many now use secondary butterflies which perform the same basic function as a CV diaphram, but are controlled by a the computer via a stepper motor. The bikes work fine with these removed and many folks do exactly that.

I know older GSXR600s used 38mm TB's as did the F4i. Crusiers tend to have much smaller TB's, for example a 750 shadow uses a single 34mm TB IIRC.

The FI bits off a WR250R might also work well (also 38mm I think).

Share this post


Link to post
Share on other sites

I've been looking for an '01 GSXR 600 TB. I think it might be 2mm too large for the stock DR intake port. Still, it looks like a good candidate. WR250 might work also, but I haven't seen any on eBay.

I got the intake boot for the R6 TB, the bolt holes appear to be about 2.75mm too far apart for the DR cylinder head, but it shouldn't be hard to open up the holes on the boot to make it fit. It cost only $15 new. I'll try to put it on next weekend and report back.

At first I would have preferred a TB without a slide, but now I think it might be an asset. After removing the slide, the cap makes a great mount for the intake air temperature (IAT) sensor. The cap is plastic and looks just like the DR CV carb cap, but a bit wider. I've got a couple extras in case I screw up. Also, the R6 TB has a plunger on the side that appears to be an air bleed for idle air. I blew air in there and it looks like it might work that way. If so, I could rig up some stop screws and maybe even a high idle solenoid for warmup. That would be really cool!

Share this post


Link to post
Share on other sites

I am interested with fuel injection on the dr. I think I have the mechanical and fabrication skills to perform this. But I have a few questions maybe somebody can answer.

  • Do you plan on using micro-squirt or mega-squirt?

  • Whats a good air/fuel ratio for the dr, did you check to see what it stock with your carb?

  • I know that there is a TPS sensor in the throttle body, But What sensors are required for this conversion?

  • Where did you source your intake air temp sensor? What about the MAP sensor?

  • Is there an injector in the YZF R6 throttle body?

I plan on getting the o2/air-fuel meter that you have for tuning to get it just right.

Share this post


Link to post
Share on other sites

Hows about a EXUP exhaust from a FZR400? for somthing totally tunable and exact torque delivery. Just need a trigger to open the servo. This may be only suited to multi cyilnder bikes but it would certainly make the bike a one off.

What ECU will you use?

Share this post


Link to post
Share on other sites
I've made more progress, but I'm tied up with work for a week or more. I'll post some of what I have soon. I think it will come together nicely.

In terms of trigger, I plan to inductively couple to the low voltage side of the ignition coil. I've got a bunch of ferrite toroids I can wind a few turns of wire and slip over the one of the low voltage coil wires. I'll measure the output voltage to make sure it's easy to detect, but I see no reason why this won't work.

based only on the reading of these types of projects I've done ... isn't one of the challenges interpolating the accleration/deceleration of the crankshaft as it cycles?

I know a couple of conversions I read about where there was only one signal per revolution ( or less ) the microsquirt had a hard time calculating when to fire because it couldn't take into account the angular accelerations from the piston power stroke, and counterweights, etc. With more signal points per crank revolution, it can calculate velocity before hand and make a better guess. Again, havn't done this.

My inclination would be to try the exact thing you are thinking of.

Share this post


Link to post
Share on other sites
Do you expect the throttle response on the R6 TB to be similar to the stock DR350SE (vacuum) carb?

I expect the R6 TB with slide removed to be much more responsive. With Megasquirt, you can adjust acceleration enrichment, so throttle response can be very quick. Wildwestsidney said something to this effect after FI conversion of his 350se.

Do you plan on using micro-squirt or mega-squirt?

Megasquirt, it's cheap(er) and so am I :busted:.

I have a plan to build my own ECU using AVR devices (I have a bunch in the man cave). I am confident I could do it, but it's hard to justify when MS is a known quantity and it has a great ecosystem built around it. I'll probably get just the bare board and the CPU, I've got almost all the other parts already.

Whats a good air/fuel ratio for the dr, did you check to see what it stock with your carb?

I would imagine that depends on your goals, e.g. best performance, best economy, or best emissions. I just got the Innovate wideband O2 sensor the other day, I'll get the bung welded on the exhaust pipe in the next couple of weeks. Anyone have opinions on the best placement of the O2 sensor bung?

I know that there is a TPS sensor in the throttle body, But What sensors are required for this conversion?

At the minimum, you need a MAP sensor, intake air temp. (IAT) sensor, and engine temp. (coolant temp or cylinder head temp will work) sensor. A wideband O2 sensor is nice to have for quantitative tuning, but not required. TPS is used for acceleration enrichment (similar to an accelerator pump on a carb), but you could leave that out. Throttle response will lag a bit without it.

Where did you source your intake air temp sensor? What about the MAP sensor?

I don't have the IAT sensor yet, I'll probably get a couple of discrete thermistors that are similar to the transfer function of the GM unit supported by MS out of the box. The MAP sensor comes with the MS kit, you can buy them seperately from Mouser or Digikey for about $10.00 each.

Is there an injector in the YZF R6 throttle body?

Yes, four of them, one for each cylinder. They are high impedance, about 13.5 Ohms each. They can handle 12V continuous with no problems. I read their flow rate is 22.9 lb/hr, which is bit on the high side for a 350, but Captain Midnight tells me it's not all that critical.

I plan on getting the o2/air-fuel meter that you have for tuning to get it just right.

I went back and forth on that, but I think it's worth $190 if you want to get the best performance. I'm sure you can get away without it, that's what the MS manual says. Many street FI bikes in the US don't have O2 sensors as emissions are not required for bikes in many states.

As mentioned above, wildwestsidney had good success on his 350se. That was a few years ago. I tried to PM him to get his VE tables and tuning parameters. He didn't respond and hasn't posted for a while. He's in MA, does anyone have contact with him?

Share this post


Link to post
Share on other sites

Cool project, I love DIY EFI conversions!

Got some questions regarding your fuel system:

- Isn't the pump designed to be placed in a tank and rely on fuel to cool it?

- I am assuming the pressure regulator is part of the pump assembly. Does it need a return line to the tank or can it be run in a "dead-head" configuration?

Thanks,

Lukas

Share this post


Link to post
Share on other sites
TPS is used for acceleration enrichment (similar to an accelerator pump on a carb), but you could leave that out. Throttle response will lag a bit without it.

MAP-only works better on multi-cylinder engines with some plenum volume. A TPS will probably be necessary on a small-plenum single because the MAP signal will be so rough at idle. The ECU will interpret the big sweeps as accel/decel and make the mixture go all over the place. It's even worse when it's controlling spark as well. Hybrid alpha-N is handy in situations like that.

I went back and forth on that, but I think it's worth $190 if you want to get the best performance. I'm sure you can get away without it, that's what the MS manual says. Many street FI bikes in the US don't have O2 sensors as emissions are not required for bikes in many states.

That's true. Once the tune is finalized, the engine should run fine without EGO sensor feedback; it just omits mixture trimming. As well as bikes, many OEM cars are programmed to ignore EGO feedback under high load/WOT too. The common narrowband sensors are not useful for anything but lean cruise or idle anyway.

Whats a good air/fuel ratio for the dr

There is no one answer to a question like this. The short explanation is that it depends on a lot of factors and will vary largely based on throttle opening and RPM. Generally speaking, airhead engines prefer richer AFRs than water-cooled engines in order to keep cylinder head temps in check. It's not uncommon for "lean" cruise AFRs to be richer than stoich whereas water heads can often be leaned out to the brink of misfiring.

Fun EFI trivia question #3,901,645; How can you tell the difference between a lean misfire, a rich misfire, and an ignition fault with a wideband O2?

  • Like 1

Share this post


Link to post
Share on other sites

Is a crank position sensor required?

I saw somebody else EFI their bike a long while ago. and he had to machine some bits on to the flywheel. If i remeber correctly you can't have too many pegs on the crank or megasquirt won't be able to count it fast enough.

Anybody else know what I'm talking about?

Share this post


Link to post
Share on other sites

- Isn't the pump designed to be placed in a tank and rely on fuel to cool it?

- I am assuming the pressure regulator is part of the pump assembly. Does it need a return line to the tank or can it be run in a "dead-head" configuration?

Lukas

There are in-tank pumps, I specifically looked for an external pump. It's a cylinder about the size of your fist with in/out port on one end. The pressure regulator is on the fuel rail, I'll post pictures in a week or so that makes it clear. There is a fuel return line from the regulator back to the tank as well as a vacuum port.

Share this post


Link to post
Share on other sites
Is a crank position sensor required?

I saw somebody else EFI their bike a long while ago. and he had to machine some bits on to the flywheel. If i remeber correctly you can't have too many pegs on the crank or megasquirt won't be able to count it fast enough.

Anybody else know what I'm talking about?

Yes, that was the DRZ EFI project, not sure what happened to that. I think he was trying to add both fuel and spark, I'm going to use the stock CDI for spark.

For fuel only, ignition timing is all you need, that seems to work fine and doesn't require a lot of fancy machining. All bikes have crank position sensors for spark, so it's easy enough to trigger on that.

Share this post


Link to post
Share on other sites
If i remeber correctly you can't have too many pegs on the crank or megasquirt won't be able to count it fast enough.

Anybody else know what I'm talking about?

A 60-2 wheel is about the limit for MS's processing ability, but the only engines I know of with that type are out of Europe and have somewhat newer Bosch systems. Most carbureted bikes just have a single trigger tooth for the ignition, which is fine for fuel-only EFI. Controlling ignition needs a lot more accuracy, especially at low rpm on a single cylinder engine.

Something else worth noting: it's not a big deal if the fuel is injected into a closed port, better known as batch fire(all injectors open at once) and bank-to-bank(each half of the injectors take turns). MS has two injector channels, so it can do either one with two or more injectors. Staged injection too...that's a fun one. :busted:

Share this post


Link to post
Share on other sites

Something else worth noting: it's not a big deal if the fuel is injected into a closed port

That makes sense. I would imagine a hot intake would evaporate/boil any liquid gas in short order.

BTW, I assume MS can run hybrid alpha-N. Is it hard to tune?

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Reply with:

Sign in to follow this  

×
×
  • Create New...