96-2010 DR650SE Fuel Injection Project....

[mxrob]

Yeah, the superior FI fuel atomization at lower RPM was what I was counting on for the larger bore TB to not hack up my low end. I'm still relieved to actually see it.

I was using the timing retard algorithm for the rev limiter but the bike was consistently making it past 8k on the dyno during testing... so I went in and turned on fuel cut instead. There is a limit the software will allow you to retard the engine... 12.1 degrees. Not sure where that particular number comes from. ?

Edited October 1, 2010 by mx_rob

[lowercasee]

Wow, fantastic results Rob!!! Makes me miss my DR even more, even though there's no way I would have the fortitude to install FI on it. LOL!

[msiddalingaiah]

Awesome!! How far off was your seat of the pants tuning? Can you post your fuel and spark tables?

[mxrob]

Awesome!! How far off was your seat of the pants tuning? Can you post your fuel and spark tables?

The shop felt I had a good AFR table built but my VE was not correct in all areas to match my AFR. I was rich on top end.... because I'm skeer'd to melt aluminum. ? But I was lean on some of the mid RPM high load roll-on areas. It's not an area where you get a lot of time in like cruise throttle or plain old WFO. Now that my VE tables match what my AFR tables are saying I can make all of my adjustments in the AFR table.... But, I still have an issue with too little resolution in the lower bins of my VE table. A one number change on my VE table at idle makes a big AFR change. I have to rescale my VE tables and adjust required fuel to get those lower numbers up into the low 30's instead of the low 20's so a one number change isn't such a drastic percentage. That way I can match my AFR table. I'm going to work on that this weekend. I doubt any of my numbers will be useful on your 350 but I'll post them up once I rescale the VE tables.

Oh yeah, the shop liked the secondary VE table set-up I'm using. They were showing me some of the different software that some of the manufacturers use for FI tuning. They have a bunch of different software for different manufacturer's FI. What I saw of the few they showed me is total cave man stuff compared to the Tuner Studio and MSExtra code. They were amazed at all the active windows I could have open at one time while monitoring the engine and then just make a change live on which ever window I wanted. They were also very impressed with the blue on screen trailing markers and color weighting of the active bins. Makes it much easier to make averaging adjustments to groups of bins. After seeing some of those very elementary tuning programs I don't feel so bad about not being an authority on all the stuff available with TS. ? I'll try to give you a call tonight when I get home....

[mxrob]

Tables of information...

This is my AFR table. Still to be tweaked in the cruise throttle ranges for best fuel economy.

This is my ignition table...

Same tweaking will be applied in the cruise throttle range.

And the volumetric efficiency tables...

I will be rescaling them both to allow for higher resolution on the lower bins. The plan is 1.5 X all the existing numbers... so, for example, the 23's in the lower two bins will be 35's and so forth. I actually thought it was going to be as easy as doing a table rescale (which is real easy) and then an equal reduction (33.3333...%) in the required fuel number. So I did that and the bike ran like sh*t. ? Obviously a tiny bit more to it than that cocky little exercise in futility. So I reloaded the final dyno MSQ back in the controller and I'll have to regroup on that issue.

Read from the bottom table up... lower table (VE1) handles up to 35% throttle... the upper table (VE2) handles 37.5% and up.

Edited October 2, 2010 by mx_rob

[msiddalingaiah]

Rob,

During the dyno run, what was the optimum AFR for best torque/power? Around 12.5 seems to be a good number, but I have read that it varies somewhat for different engines.

[mxrob]

Actually the lower RPM torque likes to be a bit leaner than HP. Not really sure why it works out that way... it's just what the engine liked. You can see it in the AFR table... we played around with numbers in all the areas... 12.5 was the best my bike could do for HP. Lost HP going in either direction from 12.5.

[el capitan]

mx_rob, you´re my guru! It cost me approx. half of yesterday night to read through this thread. After all that time you worked yourself through ignition-topics and had to get your combination of sensors, soft- and hardware running I was not too optimistic about you getting all this sorted out. And now that dyno-run and the vid with those clutch-free wheelies!

I just recently stepped down from a Suzuk SV 650 S (fairly well tuned) to an original old 1996 SE. And immediately went over it from shocksprings over brakes and bodywork to carb. I was so proud of having successfully managed to get the CV carb running properly on first attempt (Mikuni-jet, KTM-needle, drilling slide, opening airbox). After reading your FI-story I feel like a humble firstgrader watching those cool highschool-kids ...

Could you roughly estimate the time and money that all in all went into this project? Even with a good grip on used LT700-spares (unlikely to be available outside the US) the parts alone sure sum up right above 700 $, right? Not to mention the CNC-work. Plus the time invested to go from a box of loose parts to a properly running system. But even if you´d offer a kit ready to install at - say - 1.000 or 1400 this would still be a reasonable price compared to a new Keihin-carb alone. Any thoughts of assembling such a kit? I guess there´d be a bunch of buyers queing up immediately ...

Once again: My deepest respect for taking the challenge and my honest congratulations for the successful end!

[mxrob]

Thanks very much for the encouraging words..... I can't imagine me selling a kit with all the components... and we certainly could not talk about such a prospect on TT now could we? ? The biggest issue is finding the components at a reasonable cost. If new items were purchased it would be way too expensive even if you could do all of your own machine work. I did not use any CNC machines for what I have done so far... just a mill with a 4th axis indexer attachment and a lathe. I'm no extraordinary machinist by any means. Some of the more forward members over on ADV have pointed that out giving my project a score of about 3 out of 10 I think. Of course the score giver still has a BST on his DR650 I believe.? Touche'

Of course in doing a project like this you spend quite a bit on experimental stuff that blows up in your face... I've had my share of those so far. For the base LTR fueling components I'm using at this time I have about $250 wrapped up in them give or take a fifty dollar bill. I also have quite a few spares... even a full set from one of the TT members who started the project along with me and then had financial issues that did not let him continue. That's just the fueling components.... and if you can find them for that price. The Microsquirt FI controller is $435 with the harness. To do the wasted spark just like I'm doing now you'll need the Cherry hall gear tooth sensor... another $50. I also purchased a LTR wiring harness for $50 to get all the sealed water tight component connectors.... it happened to come with another MAP sensor and the IAT sensor... any one of the components there would cost way over $50 from Suzuki. Of course you have all of the wiring and misc stuff to consider. I had to build my own degree wheel and come up with a way to run both systems at the same time so I could get a base timing set-up. There are some very minor costs in electronic components for the ignition tach signal. Also the $75 Nology coil and the $25 Accel wires. Now if you want to run a wide band 02 snsor like I have... there is another $150. As you can see it adds up rather quickly.

I'm planning to work on a wasted spark set-up that will use the stock ignition VR so no additional machining or sensor wiring will be required. That will simplify the project and save some money. Still, not a cheap undertaking. I'm still testing, troubleshooting and working through issues as they arise. Thsi winter cr*p coming up shortly is going to really goof up forward motion on the project though. ?

[msiddalingaiah]

Some of the more forward members over on ADV have pointed that out giving my project a score of about 3 out of 10

WT*F??

I'm planning to work on a wasted spark set-up that will use the stock ignition VR so no additional machining or sensor wiring will be required.

Another bike?? Details! We need details!!

[Older and Slower]

How is your battery holding up to the additional demands?

[el capitan]

Hi Rob,

just to understand: Does the possibility to change ignition timing add so few to the power-gain in comparison to the effects of FI that your next project will leave the ignition-topic aside? Just concentrating on FI alone? Would this also mean that you don´t need the machined wheel of your current bike?

I just summed up the costs mentioned by you and ended up at $ 1150 to 1200 - quite some money, regarding the fact that there still was all the brain and work to be put into the project ...

Your bike already featured the je-piston and the cam. What would you estimate as a realistic rearwheel power-output for a stocker (with open airbox) and your FI-mods?

[mxrob]

How is your battery holding up to the additional demands?

Not as well as I'd first hoped. Voltage is a bit low. I was using one of Woodsryder's lithium batteries but I was fearing for it's life with the lower voltages I was recording. I switched back to the lead sled battery until I get the voltage sorted. I have some options like the stator rewire but I'm first going to try a mod that swaps out the stock SCR regulator/rectifier with a MOSFET type regulator/rectifier from a CB600RR. TC Tom has done this on his DR650 and said it brought voltage up quite a bit as there is a lot of waste in the SCR type unit. I can verify that since mine gets pretty darn warm. He said the MOSFET unit stays nice and cool by comparison. I have one sitting on the work bench.... for now I have the headlight unplugged. I still have the 35w HID option available as well and once I get all the tuning sorted I can drop the use of the wide band 02 sensor plus I won't have the handlebar mounted PC with it's power converter plugged into the electrical system either. Both use a fairly good amount of power.

Edited October 27, 2010 by mx_rob

[mxrob]

Hi Rob,

just to understand: Does the possibility to change ignition timing add so few to the power-gain in comparison to the effects of FI that your next project will leave the ignition-topic aside? Just concentrating on FI alone? Would this also mean that you don´t need the machined wheel of your current bike?

I just summed up the costs mentioned by you and ended up at $ 1150 to 1200 - quite some money, regarding the fact that there still was all the brain and work to be put into the project ...

Your bike already featured the je-piston and the cam. What would you estimate as a realistic rearwheel power-output for a stocker (with open airbox) and your FI-mods?

After seeing the improvement in how my bike ran just on the ignition part of the FI controller alone I was further sold on the ignition portion being a requirement. That extra timing resolution and 3 dimensional timing control is a big part of the improvement. It also will allow a sequential injection algorithm with a very precise injection point. This will be a help in attempting to use MAP for a more efficient fueling control. I haven't even finished Alpha-N fueling... still have a few bugs to work out.

Yeah, like I said... not a cheap endeavor. Lots of fun though.

It would likely be a bit higher than the stock bike with an FCR... just like my bike is a bit higher than with the FCR. The dyno charts on the FCR thread show the power my bike made after the dyno testing with the stock motor... don't remeber off the top of my head. It's more than just HP though. The smoothness and throttle control is pretty amazing... even going above the FCR in many ways.

[ProCycle]

I'm first going to try a mod that swaps out the stock SCR regulator/rectifier with a MOSFET type regulator/rectifier from a CB600RR. TC Tom has done this on his DR650 and said it brought voltage up quite a bit as there is a lot of waste in the SCR type unit. I can verify that since mine gets pretty darn warm. He said the MOSFET unit stays nice and cool by comparison. I have one sitting on the work bench....

Hey msiddalingaiah

Can you comment on this? How does the MOSFET type regulator manage run cooler? Since the permanent magnet generator still has to be running wide open all the time where does the excess energy go?

[Older and Slower]

The only thing i can think of is that an SCR is an either on or off type of device and a MOSFET is actually a transistor thereby allowing a variable amount of shunt current depending on the base drive to the transistor. Maybe it allows a much finer amount of control in the battery charging scheme of things. (shrug)

[mxrob]

I can only make assumptions here... but since his battery voltage is higher using the MOSFET unit over the SCR unit I take it more of what the stator is producing is making it to the battery instead of being turned into waste heat. It will be interesting to see what difference it makes....

[msiddalingaiah]

Hey msiddalingaiah

Can you comment on this? How does the MOSFET type regulator manage run cooler? Since the permanent magnet generator still has to be running wide open all the time where does the excess energy go?

That caught my eye as well.

As switches go, here's the difference between SCRs and MOSFETs: SCRs are like diodes that can be switched on (they switch off by themselves when current drops below some threshold). They have a voltage drop that's somewhere around 0.7V or more depending on the current (more current, higher drop). The power dissipated is current * voltage drop, so 25 Amps can waste 20 Watts or more in the SCR.

Modern high current MOSFETs are more like resistors at low voltages. They have very low "on" resistance, some as low as a just a few milliOhms. The power loss there is I^2 * R, so 25 Amps wastes less than 1 Watt in the MOSFET.

That's part of the story, but I don't know for sure how these newer MOSFET regulators work. If all they did was shunt excess voltage to ground like SCR type regulators, there's still a lot of power wasted in the stator coils (more in the case of MOSFETs). If they are efficient switching regulators like the ones we talked about, very little power would be wasted.

Does anyone have a link to technical data on MOSFET regulators?

[Hectoralis]

So, off subject, but MX Rob, I recently checked out your website. Very legit. Appreciate the detail with which you document your mechanical adventures. I will certainly be borrowing your knowledge and experience in the future.

[mxrob]

I love doing/seeing creative solutions to problems apparent and, in many cases, not so apparent. Everyone sees things in a little different light. Glad to be of help/inspiration.