96-2010 DR650SE Fuel Injection Project....

[msiddalingaiah]

Unfortunately the space in the frame I am using (LC4 Adventure) is very limited, or I would have kept the twin throttle bodies that came with the 650 twin engine. As it is

they are way too wide and use up too much space.

Sorry, I was posting at the same time you were. Fitting a Y will work. As long as you have plenty of air and the mixture is right (which it will be with FI ), it will be fine. You will need a bigger injector if you go with a single.

It's best to squirt onto a hot intake if you can. It will work either way, but you might need more accel enrichment with a cold plenum to compensate for wall wetting.

Start a build thread and post the link, we'll all jump on and have more fun

[LukasM]

Wouldn't it be better to mount a separate bung and injector in each head inlet to ensure that both cylinders are getting the right mixture? I would imagine that you could tune the MS to do that after checking the pipes separately with an O2 meter.

There is actually a long build thread on Advrider already (lots of interest in a mid-size adventure bike with a twin engine), but it's currently on ice since I wasn't sure how to deal with a few issues of the engine swap. Got some other projects I am working on right now, but I'll pick it up again eventually.

http://www.advrider.com/forums/showthread.php?t=456542

[msiddalingaiah]

Wouldn't it be better to mount a separate bung and injector in each head inlet to ensure that both cylinders are getting the right mixture? I would imagine that you could tune the MS to do that after checking the pipes separately with an O2 meter.

Sure, it would be better to have two injectors, one at each inlet, but TBI with one injector is known to work. It all depends on how much you want to fabricate. You could check O2 on each pipe separately, but as long as everything is symmetrical, it should be OK. Can MS tune two injectors independently?

Look at it this way: single carbs feeding multi-cylinder engines have been used for a century or more, even on high performance muscle cars. A single injector feeding multiple cylinders is not much different. Of course, sequential multi-port injection is the ultimate, but it's more work.

It all depends on your goals and how much effort you want to put into it.

[HeadTrauma]

Wouldn't it be better to mount a separate bung and injector in each head inlet to ensure that both cylinders are getting the right mixture?

That depends. Some injectors are designed to spray down both sides of the intake port on a four valve engine. Some throttle bodies are equipped to use two injectors like the Suzuki TL1000 and possibly SV1000.

[mxrob]

Got it all installed and tested. What a PIA getting all the hoses and everything to fit properly. I had to take stuff out a couple times to get other stuff to fit... it's a routing marvel. I found out that the stock CA tank not only has the vent fitting and a different non-vented cap..... but the tank opening and cap engagement components at the top of the tank are different. Yeah, I found this out after taking the brand new 48 state cap out of the package I orderd months ago and tried to put it on my tank.? Well, the wife has an extra tank cap now. So I dissected the stock CA cap and drilled some porting to make it a vented cap using the outer edge of the stock gasket as the flapper valve.... works great.

I installed the whole system but left the fuel rail unattached from the TB so I could run a couple quarts of fuel through the system to clean it and remove trapped air. I was worried that maybe the stock petcock would have a flow issue... NOT! It filled up a 1 quart stainless shop bowl in 30 seconds or so. That was a relief to see. I also got the carbon fiber cover for the fuel pump in the mail as I was working on hosing. It like it a lot better with the cover. Nice and smooth so nothing will catch on the clamps and fittings... looks pretty cool too. I'll take some pics tomorrow. I took the tank back off to finish the wire extensions and harness to the handlebars. Bootloader switch, WUE led, accel enrichment led, tach signal (if I ever want to put one on) 02 sensor input wire. I'm in the process of installing the LC-1 wide band 02 sensor right now. The LC-1 and sensor grounds have to be connected in the exact same place as the Microsquirt grounds are connected. More darn wiring.... Sheesh.

It will certainly be cranked up on fuel tomorrow sometime. :usa:

[msiddalingaiah]

Yeah, there's all that little stuff you never expect

I think it's going to turn out great

BTW, does the DR650 CDI produce an output for the stock tach? That's how the DR350 is wired. If so, how are you driving your tach without the stock CDI? Or have you sidelined the stock instruments for now?

[mxrob]

It's done!?

Here's the left side of the engine with the cool carbon fuel pump cover. See the fuel filter/swirl tank buried in there? It has an upper side take off tube that I have pointing as upwards as possible to help remove trapped air. That line T's into the return line and the fuel pump flooding line (that's the big line coming off the top of the pump). That line t's directly behind the injector and goes down to the FP regulator by-pass with another big line and then comes out from under the injector and connects to the T that the filter vent connects to. You can see my extended fuel rail (black painted brass tube off the top of the TB) going forward to the stock R450 high pressure FI line. This curves around the TB and connects under it to one of the FP regulator pressure ports. The other pressure port line comes back from under the TB and connects to the outlet of the fuel pump. Sheesh... hose central. The FCR choke knob is plugging the hot start port which is cable operated on the R450. Suzuki wanted $57 just for the valve... then you'd have to buy the spring and the threaded plastic cable adapter just to hold the valve in place. Not! The FCR choke works fine. Just behind the knob is the idle air port adjuster. No having to move the butterly to adjust idle air flow and then having to re-calibrate the TPS.

The final layout of the µs installation. After my serial cable connector pulled apart I decided I needed easier access to the big wiring bundle so I ditched the zillion wraps of electrical tape and made a boot out of stuff from the R450 harness instead. I was also worried about water getting shot down the back of the seat (like pressure washing the bike) and getting into the back of the µs connector. This set up addresses both issues.

Right engine side. You can see the vent lines coming from the filter/swirl tank and the regulator return and FP flood line. The line going straight up connects to my return line on the CA tank. The 02 sensor is kinda close to the lower vent line so installed a protector spring just in case. You can see the end of the fuel pressure regulator behind the 02 sensor and above the starter. Yeah, it's amazing it all fit. The LC-1 oxygen sensor controller is installed behind the headlight cowling and the sensor is free air calibrated.

Right side µs view. It all just fit nicely. You can see the tip of the incoming air temperature sensor just above the top of the freshly cleaned Twinair filter.

The left view of the DR650FI.

The right view of the DR650FI.

So what's left? Varooom! I'm going to hook the big laptop up and sit down to go over every part of the msq file (engine specific tuning parameters) to make sure I haven't missed something. I need to turn off accel enrichment and a few other things for the initial start up and idle adjustments. I've got butterfly's in my stomach and two fire extinguishers at hand. The next posting should be it's running! Either that or my son notifying everyone that his dad was incinerated in the shop.

[mxrob]

I can't believe it... but it started right up!

It was way too rich right off the bat everywhere and my default lower VE table numbers were already in the mid to upper 20's. You normally want to keep them in the mid 20's to 30 in the first bins from what I've read then go up from there.

I shut it down and then figured out what my problem was.... the "required fuel" number (a millisecond fuel setting based on engine and injector size) was still set for sequential injection so the engine was getting twice the fuel it needed since I'm now running wasted spark. A quick keyboard stroke and a refire got me a whole lot closer. It was clean enough that I could adjust the idle air port to keep the engine running. A trip to the VE table and a few key strokes had it idling like a baby... I still had to drive the VE numbers down to 19 or 20 so I made another trip to required fuel and lessened the number a bit. A couple more required fuel adjustments got my initial VE settings right at 24/25 with a nice smooth 13.0 idle. Of course with accel enrichment turned off (to not have it influence my AFR numbers when adjusting the VE table) any quick movement of the throttle killed the engine instantly. Slowly rolling the throttle I set quite a few more bins so the bike is now at least ridable. That's next... strapping on the skid lid and going for a small ride around the block... ?

[mattheritage]

WOOHOOO!!!!!!!!!

[mxrob]

Yeah, there's all that little stuff you never expect

I think it's going to turn out great?

BTW, does the DR650 CDI produce an output for the stock tach? That's how the DR350 is wired. If so, how are you driving your tach without the stock CDI? Or have you sidelined the stock instruments for now?

The 96 and up DR650's don't have a tach... but the µs has a tach output. I went ahead and ran if you to the wiring bundle behind the headlight cowling in case I want to put one in some time. So for the older DR's with a tach you could still use it with the µs.

[mxrob]

WOOHOOO!!!!!!!!!

Yeah buddy! My "small ride around the block" ended up being a one hour ride-n-tune until it was getting too dark to be racing around on the Michigan back roads. Man, this mini laptop on the bars is so cool it's ridiculous. I've already got the bike running extremely well considering I still have not turned on any accel enrichment. I'm going to spend the day tomorrow logging as many consistent throttle positions as possible at varying loads. I see I have some AFR settings that I first thought would be good at a particular throttle position/RPM that are not..... it's the lower RPM settings with light throttle (in town cruise). Right now I do not have enough spark advance for as lean as I'm wanting it to run so it surges. Just need to add a bit of light-throttle spark advance to let me run the leaner AFR. How freakin cool is this anyway? I've entered tuning fantasy-land.

I did develop a small fuel seep at the fuel rail/injector junction. It's the o-rings on the rail fitting. I have them on order.... I guess they are soft Viton... $12.53 for two o-rings! Damn! Glad that's a one time purchase.

[NordieBoy]

Oh yeah

[mattheritage]

Yeah buddy! My "small ride around the block" ended up being a one hour ride-n-tune until it was getting too dark to be racing around on the Michigan back roads. Man, this mini laptop on the bars is so cool it's ridiculous. I've already got the bike running extremely well considering I still have not turned on any accel enrichment. I'm going to spend the day tomorrow logging as many consistent throttle positions as possible at varying loads. I see I have some AFR settings that I first thought would be good at a particular throttle position/RPM that are not..... it's the lower RPM settings with light throttle (in town cruise). Right now I do not have enough spark advance for as lean as I'm wanting it to run so it surges. Just need to add a bit of light-throttle spark advance to let me run the leaner AFR. How freakin cool is this anyway? I've entered tuning fantasy-land.

I did develop a small fuel seep at the fuel rail/injector junction. It's the o-rings on the rail fitting. I have them on order.... I guess they are soft Viton... $12.53 for two o-rings! Damn! Glad that's a one time purchase.

It's so cool to me that you can tune every throttle position for spark and fuel. What a beautiful thing. Makes me look at my FCR like a red headed step child. I wonder though... once you get the FI done and set and are so used to it you don't thing about it much anymore... What are you going to do then? Besides crank out FI kits of course.

Serioulsy though, congrats. I can imagine how stoked you are about it.

[mxrob]

It's so cool to me that you can tune every throttle position for spark and fuel. What a beautiful thing. Makes me look at my FCR like a red headed step child. I wonder though... once you get the FI done and set and are so used to it you don't thing about it much anymore... What are you going to do then? Besides crank out FI kits of course.

Serioulsy though, congrats. I can imagine how stoked you are about it.

Thanks Matt.... Yeah, ignition control is so important if you really want to get all the benefits of FI. You just can't run the lean cruise throttle settings if you don't have a way to advance spark so the burn completes at the optimal time. A less fuel dense charge takes longer to burn... this is why the stock DR650's head pipe glows so red hot at night on the freeway with it's leaner needle setting. A big portion of the fuel charge is still burning as it enters the head pipe instead of releasing it's energy in the combustion chamber where it belongs.

My time? I'm pretty anal about tuning so I'll be spending a lot of seat time perfecting the FI tune parameters. There are a ton of enrichment and cold temp adjustments to do...... some of which you only get one shot a day at adjusting. Bummer huh? I have to ride my bike. Then I'll take my bike over to Fastbikes USA and throw it on the dyno again. That will open a whole new tune parameter knowledge base for me. We already have a good torque/HP curve base so now we test timing and AFR combos at different loads and throttle positions to see what combo creates the best peak torque in the torque curve and the best peak HP in the HP curve. I'll probably do this in 5% throttle increments and interpolate the remaining values. I'm really geeked about getting this thing back on the dyno for this peak tuning exercise.

But a good point you bring up there... my wife was already asking the same thing. "You're going to go crazy (you'll drive me) if you don't have a motorcycle to work on." Not to worry... except for the intiminators in my forks my suspension is bone stock. I have a DRZ shock just waiting it's turn for some attention.... USD forks are on the agenda.... I don't have any SM wheels... yet. I have to send my head pipe out to have it ceramic coated inside and out...... If I'm not concentrating on one thing I'm thinking about a few others....?

[msiddalingaiah]

I knew you would get it running!!! Are you running straight alpha-N or a blend? How did you come up with the initial fuel table?

My lowest fuel table values are 55 (idle), so I can adjust in 2% steps or so. I found that when I had too high of a REQ_FUEL value (wrong injector flow rate), some of my number were down in the 20s, which meant 5% steps. Even one click would make a significant difference. Still, it was tunable.

Does your CLT vary much while riding? I need to improve my sensor mounting. Sometimes if the bike is not sufficiently warm, it will go into warm up when air flow cools the sensor and it runs way too rich. I can probably fix that by tweaking warm up enrichment settings also.

[mxrob]

And boy does it run! Yes, I'm running pure alpha-N to start. I took a default fuel table from TS and then modified it for my bike. My lowest VE idle numbers are around 25 at idle. I just set up some more resolution in the lower % of fuel load bins because this is much more important than the upper resolution... so I go 2,5,10,15,20,30,35,40,45,50,55,60,65,85,95,100. my goal is to be between 25 and 30 for the lowest bins. 25 seems to work nicely so far.

During riding the CLT doesn't vary more than about 15 to 20 degrees.. runs around 175 to 195.... on the hotter side during high speed/load freeway riding and then if I get into heavy stop and go it runs around 200-205. Not bad at all IMO. I'm well pleased with my Rube Goldberg copper widgit.

[mxrob]

I started this morning working on setting up my warm up enrichments. Really nice to watch the AFR, RPM and pulse width while adjusting some enrichment points and numbers. I had it warming up nicely in a few minutes and it stayed really smooth as it tapered off. No need to touch the throttle. I also have some timing advance I can use but I have that turned off right now until I get a good handle on fuel. After a few more runs out and about I decided to turn on accel enrichment. Boy did that wake things up.... I have some reading to do on the affects of different parts of the code but just a blind adjust this and adjust that has the bike nearly perfect.... I just have a tiny bit of right off of idle hesitation. I'm wondering if it might be my start from 0 TPS calibration... Gotta do some more reading on the algorithm.

I also turned on "over-run fuel cut". This is cool beyond words. Right now I have it set to cut fuel anytime I'm higher than 2,000 rpm but lower than 50kpa on MAP and lower than 3% TPS and I have it wait 1 second before paying attention to EGO. Basically it shuts fuel right off on decel until the engine gets down to 2k RPM. No sense in wasting fuel being pumped through the engine on decel. Obviously with no fuel there is no chance of any decel popping... but I noticed another really cool result from using this feature. Engine braking is much better than before... I guess that's logical since no combustion is taking place at all until I get to 2k RPM... you can feel it kick back in just as you approach a stop. Decel sounds meaner with the total fuel cut. Man this is fun!

[mattheritage]

I also turned on "over-run fuel cut". This is cool beyond words. Right now I have it set to cut fuel anytime I'm higher than 2,000 rpm but lower than 50kpa on MAP and lower than 3% TPS and I have it wait 1 second before paying attention to EGO. Basically it shuts fuel right off on decel until the engine gets down to 2k RPM. No sense in wasting fuel being pumped through the engine on decel. Obviously with no fuel there is no chance of any decel popping... but I noticed another really cool result from using this feature. Engine braking is much better than before... I guess that's logical since no combustion is taking place at all until I get to 2k RPM... you can feel it kick back in just as you approach a stop. Decel sounds meaner with the total fuel cut. Man this is fun!

That's a benefit I never really thought about with FI. I wonder how much fuel is wasted on decel? I guess with the FCR its better than with stock, but it's still wasting some. A few miles to the gallon maybe?

[mxrob]

That's a benefit I never really thought about with FI. I wonder how much fuel is wasted on decel? I guess with the FCR its better than with stock, but it's still wasting some. A few miles to the gallon maybe?

I doubt it would make a ton of mileage difference over the FCR-MX's air cut system but I still got a tiny bit of decel popping on my FCR.... which is far better than trying to drowned decel popping on the BST with a sloppy rich pilot setting. There is absolutely no decel popping with the FI system over-run fuel cut... there's nothing to burn. I have to study the data logs to make sure the EGO response is delayed long enough so the controller doesn't go "wow, we are lean I better crank up the fuel" that is what the delay EGO timing is for. It allows the AFR to get back down to my 13.0 idle setting before it takes a look at the oxygen sensor reading to make adjustments.

[msiddalingaiah]

After a few more runs out and about I decided to turn on accel enrichment. Boy did that wake things up.... I have some reading to do on the affects of different parts of the code but just a blind adjust this and adjust that has the bike nearly perfect.... I just have a tiny bit of right off of idle hesitation. I'm wondering if it might be my start from 0 TPS calibration... Gotta do some more reading on the algorithm.

It's amazing how much of a difference accel enrichment makes, even though I don't think I ever had it dialed in perfectly. I'm just about at the same point as you are now with fuel. I don't think the TPS calibration matters much, at least not on MS-I. It only looks at the rate of TPS (TPSdot) in terms of Volts/sec, so an offset should have no effect. Of course the newer code with X-tau enrichment might be more sophisticated.

You can add TPSdot to the calculated fields in MegaLogViewer, but you need to give it the low and high TPS ADC counts. It's no big deal. It's good to see that because accel enrichment settings depend on it.

Post your msq file once you have it stable. I'm curious to see how it compares with mine.