Welcome to Stan's Web Site... NOT very pretty, but hey, it's my 1st web page, and a good start to document my XJ1100 fuel injection project. I'll work on "pretty" next week. I hope this info helps. slh Aug. 25, 2006
I'm still developing this site, and I still have lots to write and photos to insert; but here's the Bottom Line.
Here's a nice set of reference diagrams, Stan's MegaTune Displays and XJ1100 Reference Diagrams.

• Introduction
• The Little Things
• FAQ's
• Cost
• Hardware (Parts)
• EFI and MegaSquirt
  • Building MegaSquirt Stimulator
  • Building MegaSquirt Controller
  • Installing MegaSquirtNSpark
• Ignition
  • Reluctor Wheel
  • Mounting Shaft
• Assembly
  • Ignition System
  • EFI System
• Tuning the EFI System
• Final Report
• Stan's Bio
• Acknowledgements
• Links
• '06 TeXSive Rally
• '06 TeXSive Rally
• 2020 EFI Update

Introduction
In February '06, I decided to replace the carburetors on my '82 Yamaha XJ1100 Maxim motorcycle with an electronic fuel injection (EFI) system. I also replaced the ignition system.

The XJ1100 motorcycle is an inline 4 cylinder engine, producing approx. 90 hp at 5500 rpms. This project took about 7 months to complete, working nights and weekends. During this time, I left town for 5 weekends, including a wedding, two reunions, an offshore fishing trip, my spring bass fishing trip, and the birth of my grandchild, Katherine. When Houston Summer's arrived my motivation slacked; just too darn HOT to enjoy the garage. I also had a full time job and there's always home chores; remember it's a hobby; I also own a Honda CB900F Super Sport to "keep me satisfied". Stan's Photo Gallery

Stan and his '82 XJ1100 Maxim motorcycle.

EFI-ing a motorcycle is NOT easy, however, the results are very rewarding. If you decide to install EFI on your motorcycle, find a partner and make this project an adventure. It also makes it more fun and puts another set of brains working "the little things."

The Little Things
I found lots of "little things" tripping me up, and causing lots of delays. I'm laughing at them now, but during the "heat of battle", I found them "madding". I'd start a task, and then get stuck for several nights or weekends trying to solve the problem, finding the right part, or locating the wiring error. It drove be crazy and extended the project. I'm writing this to help others (and record my adventure.) I hope you learn from my mistakes, save some time, and enjoy the technical experience. Here's a list of some of my miscues:

1. When Building the MegaSquirt computer (CPU), be sure to install the capacitors correctly, some are polarized.
2. You need a 9 pin serial port on your computer to connect the MegaSquirt CPU. My wife, Deborah, picked one up at Best Buy and helped me install the board and software; she's a computer whiz.
3. The battery is grounded to the motorcycle frame, and NOT the engine. I struggled with a "weak starter" for months, before figuring this out. I recommend installing another ground wire from the battery to the starter motor.
4. Finding top-dead-center and making the reluctor plate became a "labor of love".
5. One of the GM DIS wiring diagrams has an error.
6. The GM DIS ignition system creates a ton of Radio Frequency Interference (RFI), resetting the MagaSquirt CPU; I corrected this problem by installing resistor spark plugs and RFI reducing spark plug leads.
7. The timing marks on the reluctor plate goes the to right of TDC.
8. If you are easily frustrated and lack patience, Do NOT start this project; spend some quality time with your dog and the Mrs.; If you like working on motorcycles and have good internet & computer skills, and a little money, then this is a GREAT project; one I'd do again.

FAQ's

• Why replace a perfectly "good" set of carburetors and ignition system?
  Two reasons ...climbing fuel prices and reliability issues; the carburetors were showing their age and the bike was difficult to start. The bike use to get 32 MPG.
  
• Why EFI? Gosh, Every engineer's dream ... to play with motors! The perfect project ...computers, chemistry, hardware, software, eBay, motorcycles, a loving wife...it doesn't get any better.
  
• Project scope? This project had two parts: a) replace the Mikuni carburetors, and; b) replace Yamaha's XJ1100 electronic ignition system.
  
• Why replace the ignition system? Yamaha's original '82 ignition system was fully electronic, but weak. I wanted something "hotter", and something where I could map the ignition timing for better performance.
  
• Do you need a computer? YES you do, its a mandatory requirement. A laptop is best, so you can take it on road trips, create logs, and tune your EFI controller. The computer needs to have a 9-pin serial interface port. Most newer computers only have USB ports. I was lucky, my sister had an old laptop. With a little research, you can find one. My desktop computer worked for months; everything except the final tuning. Note: Dan Mann made a USB to Serial Interface Cable work.

Hardware and Parts

* Ignition parts

Anyone driving a newer cars know their advantages; I like two reasons -- environmental (big deal) and performance (BIG Deal). The best reason is money; EFI systems are cheaper to manufacture and easier to maintain; and they can improve your fuel economy. Their disadvantage is complexity; electronic systems, especially on motorcycles, can reduce reliability and availability. Everyone knows this rule, “Keep It Simple”; Complex systems fail more often.

Electronic systems have come a long way…welcome MegaSquirt.

MegaSquirt
Bruce Bowling and Al Grippo, developed a Build-It-Yourself, EFI system. Their web sites, MegaSquirt Electronic Fuel Injection Computer and MegaSquirt FAQ's, provide a great overview. I could say more, but basically, these two guys sell an electronics kit for EFI-ing your car or motorcycle. The kit comes complete with printed circuit (PC) cards, electrical components, software and documentation. And what's more amazing are the Success Stories.

Good things just get better...other people continue to develop software tools. And just in case you run into problems, there's an EFI Forum available to answer questions and provide assistance, where needed; cause things never go as planned, "As my old daddy use to say, It's always something."

Their MegaManual Index provides the latest web-based version of the manual for MegaSquirt. It's updated frequently, and contains the latest information on assembly, testing, troubleshooting, installing, and tuning your MegaSquirt.

In Feb.’06, I decided to modernize my ’82 XJ1100.

Building the MegaSquirt Stimulator
I started my adventure by assembling a MegaSquirt Stimulator (MegaStim). I purchase the version 2.0 board with 4 LED’s from DIYAutoTune.com. Note: In July '06, they started making a version V2.1 board with 5 LED's, so things have changed since I bought my hardware. I followed the instructions exactly as written in the Stimulator Assembly Manual. It was easy and very straight forward; and it gave me a chance to practice my soldering skills; cause the small successes are important.

9 volt power supply module and stimulator.	Stimulator

I tested the MegaSquirt controller by using the Stimulator; where you build a little, and then test a little; I liked this approach. It worked great, and provided confidence to continue. I used a 9 volt (120ac/9v DC) power supply module to power the stimulator, which also supplies power to the MegaSquirt controller when bench testing. The power supply is probably from an old phone or computer game; it worked great and sure saved on batteries.

Building the MegaSquirt Controller
I bought a MegaSquirt I EFI controller from DIYAutoTune.com , a version V3.0 board, . It came as a complete package, ready to assemble. I also bought a two-foot wiring harness, which is optional, but sure made things easier during the Assembly phase. The V3.0 board is a very professional looking product.

I bought 8 oz of solder from Radio Shack, which was about 4 times too much; I recommend buying the 2.5 oz., 0.032 dia. 60/40 formula spool, cat. no. 64-005. I also bought their Helping Hands Magnifying kit, which worked great for my tiring eyes. I'm an older guy with bifocals and this kit held the board rigid while applying the solder. The magnifying glass provided a better view of the work site. My biggest problem was lighting, or the lack of lighting. I tried a halogen shop light, but it produced to much heat and glare. I finally decided on a florescent desk lamp, which worked pretty well.

My brother Steve and a friend Walter looking at the MegaSquirt board.

Steve and I spent about 15 hours building the controller. We were slow but steady. The manual says 6 to 8 hours for the average person with average skills doing a first time assembly. I made a few mistakes during the assembly sequence and lost some time troubleshooting the problems. I can safely say that working a few hours each day and a little wrap-up time on the weekend, it's relatively easy to assemble the circuit board in a week.

1. Before I started soldering, I got organized ...I got a 3-ring binder with some tabs to hold my printouts. I generated a lot of printouts.
2. The MegaSquirt V3.0 Assembly Manual is 46 pages long. I printed the manual on both sides on the paper; print the 1st 24 odd pages, then flipped the paper, and print the remaining even pages. Be careful and insert the top of the paper correctly, or you'll be turning your notebook around after each page.
3. I used a 15-30 watt pencil soldering iron with a small tip. I laid out all the parts on a large table, keeping the individual parts in their baggies until needed. My partner, Steve, my brother, was real handy; he'd solder while I located the correct parts and checked off the completed steps in manual.
   
   The assembly proceeded in five functional blocks, plus the external wiring harness:
   - Power Supply Construction and Testing
   - Serial Communications Construction and Testing
   - Clock Circuit Construction and Testing
   - Input Section Construction and Testing
   - Output Section Construction and Testing
   - External Wiring Harness
   
   I bought some extra IC sockets from Radio Shack for all the IC's. There are a lot of pins on the IC's and I was concerned about applying to much heat while soldering. It turned out to be a GOOD decision, because much later, during my initial engine startup, I "smoked" an RS232 interface chip. The socket made things easy; I popped out the bad IC and inserted a new one.
   
   Power Supply Construction and Testing
4. The first module I built was the power supply module. In spite of reading the assembly instructions, I made mistakes. Be sure to install the diodes and capacitors in the right directions; some are polarized. Read steps 6 and 7 carefully and install the tantalum capacitors with the correct polarity. Ask me how I know?
5. In step 19, the manual discusses the Boot Header (H1) jumper requirement. I installed a small slide switch with enough extension wire to mount the switch to the case. I needed this boot switch later to install and load the MSns-E ignition software.
6. In step 23, I used a digital voltmeter and checked the six voltage points. Oops, troubleshooting time. I finally located the mis-installed capacitors, de-soldered, corrected the polarity, and re-tested; YES...everything checked out!
   
   Serial Communications Construction and Testing
7. The second module was the serial communications module. In step 25, I installed the U6 MAX232 serial communications chip. I decided to use IC sockets for all IC chips.
8. In step 26, I tested the serial link. I purchased a DB-9 serial interface card for my desk top computer, and later, my sister, Mary Helen, gave me a laptop. I installed the HyperTerminal software and set the comm port to Com4 and 9600 baud. I preformed steps 26g and 26h, plugged in the stimulator and tested the full loop-back circuit; it worked. Note: The MegaSquirt manual wants you to use Com1; for some reason, my serial card worked on Com4, probably due to the fact that I installed an after market card and the other communication ports were already assigned. Oh well, I got it to work, but I feel like my results were more "lucky" than talent.
   
   Clock Circuit Construction and Testing
9. The third module was the clock circuit and then cpu testing. In step 39, I installed the CPU U1 chip.
10. In step 40, I tested the processor. In this step, I downloaded and installed the MegaTune software. This step gave me the greatest concern, because I'm more of a hardware guy. However, loading the software was pretty easy. I learned the following lessons:
    

    Loading the MegaTune Software

    Installing the MegaSquirt software and the MegaNSpark extension was a little tricky; and it took me a little while to understand the process. It's like most things in life, "Once you've done it, it's easy." Hmm, I think they call this "experience."
    
    The 68HC908 processor I purchased already had the MegaSquirt software installed. However, I still needed to download the MegaTune program and install the MSnS-E ignition software (more on this later.) For now, I needed the MegaTune software for testing.
    
    I went to the MegaTune 2.25 site and downloaded the mt225 setup.exe - The standard installer. I chose the 'Run' option and all the defaults; the install program created all the directories and sub-directories; and located the files in the C:\Program Files\MegaSquirt directory. The installer also created a file called C:\Program Files\MegaSquirt\Car1; which is where specific motorcycle information goes. I copied the Car1 directory; pasted and re-named it C:\Program Files\MegaSquirt\StansXJ1100. This way I kept all the download files intact.
    
    I followed the instructions in step 40; ran the MegaTune application, selected the StansXJ1100 option in the popup window, and set the COM port to com4. (The MegaSquirt manual says com1, but I think the new serial interface card I installed set itself to com4.) Success...IT WORKED, my desktop was talking to the MegaSquirt controller.

11. In step 41, I connected the Stimulator, opened the "Realtime Display" under "Tuning" on the main menu and verified the incrementing time and battery voltage.
    
    Input Section Construction and Testing

12. I completed steps 42 through 49.
13. In step 50, I decided the GM DIS system produced a Hall optical input signal for the tach, so I performed steps 50a and skipped 50b.
14. I did all of tasks in steps 51, and the right side of step 52 (the Hall sensor items), and then in step 53, I installed the MAP sensor.
15. I chose AC Delco type temperature sensors, so in step 54, I installed the 2.49K ohm resistors.
16. In step 55, time for a little testing; I reinstalled the processor, hooked up the stimulator, connected the DB9 serial cable to the board and PC, started the MegaTune software with the "Realtime Display" running, and viola, on the first attempt, I saw my Megasquirt controller responding to inputs from the stimulator; it felt like Christmas. I verified the MAP sensor readings, it was 99 kPa on 3/18/06, on a very cloudy day ...perfect for Deer Park, TX at an elevation of 33 feet above sea level.
    
    Output Section Construction and Testing
17. I completed steps 56 through 61 as written. In steps 62, I installed the Q2 and Q4 transistors. The manual warns about "bridging the closely spaced pins" with solder. I confess, I'm guilty; I bridged the pins and had to start over. I carefully removed the excess solder and re-installed the transistors. There's a Note Block in this step, "For PWM Idle Valve Users", I ignore this note since it's not needed for my setup. I then completed steps 63 and 64.
18. In step 65, the manual totally left me confused; What is a high-current ignition driver circuit? After reading the GM DIS link, I decided to skip step 65 and jumped to step 66. During this step, I measured the resistance of the GSXR 600 fuel injectors; 12.3 ohms. i.e. high impedance injectors; I won't need the Pulse Width Modulated (PWM) circuits.
19. I carefully study the MegasquirtNSpark site.
20. I completed steps 66 through 68 as written.
21. In steps 69 and 70, I installed the PWM flyback damping circuits.
22. I completed steps 71 and omitted step 72.
23. I completed steps 73 through 77.
24. In step 78, I should have drilled the case, however, I saved a little assembly time and bought the case with the cutouts already CNC'd; well worth the extra dollars.
25. I completed steps 79 and 80.
26. It's FINISHED...well, almost ...I still needed to add the MegasquirtNSpark software and wiring modifications and clean the board with 99% isopropyl alcohol.
    
    External Wiring Harness
27. I purchased a 2 foot wiring bundle from DIYAutoTune.com for $12. The wiring bundle exactly matches the MegaSquirt V3.0 Assembly Manual (see page 45) for wire colors, and each wire is labeled. I followed the manual's External Wiring Diagram and soldered the wires to the following DB37 pin outs:
    
    

    External Device	DB37 Pin outs	Wire Color (Label)	Comments
    Ground Wires (5 wires total)	8, 9, 10, 11, 19	Black	I connected each sensor requiring a ground wire with one of these wires, including the TPS sensor.
    Intake Air Temp. Sensor	20 8	Orange (IAT) Black	IAT input Ground
    Coolant Temp. Sensor	21 9	Yellow (Coolant) Black	CT input Ground
    Throttle Position Sensor (TPS)	22 26 10	Blue (TPS SIG) Gray (5 VREF) Black	TPS SIG 5 V Reference Ground
    Main Relay	28 11	Red (12 V) Black	12 volts Ground
    Left Injectors	32, 33	Blue	Cylinders 1&2
    Right Injectors	34, 35	Green	Cylinders 3&4
    Fuel Pump Relay	37	Purple (The manual shows a greenish wire.)	Relay coil (85)

    
    
28. I connected four wires from the GM DIS module to the Megasquirt CPU:Stan, Verify this table, cause your notes are incomplete and the GM Manual has a documentation error.
    

    GM DIS	GM Socket Pin outs	DB37 Pin outs	Wire Color (Label)	Comments
    3X Reference High	E	4 (SPR2)	Purple and white	LED 16
    IC Control (EST Timing Advance)	B	5 (SPR3)	White	LED 14
    Bypass Control	A	???	Brown w/White stripe	0 volts when cranking, +5 volts running
    Ground (Reference)	F	19	Black and red	Common ground

MegasquirtNSpark
(Software and Wiring Mod's)

The last thing I did (to the MegaSquirt controller) was to install the MSnS-E software and hardware modifications.

Installing the MSnS-E Software:
Stan, write this section.

Making the MSns-E Hardware Modifications:
This was easy, I took three (20 AWG) wires from a old piece of telephone cable and soldered them to the LED diodes installed in step 80; D14, D15, D16. The other end of the wires, I soldered to the SPR-2, SPR-3, SPR-4 pads on the circuit board, near the DB37 connector.

Stan, insert a picture.

Ignition System
Yamaha equipped the ’82 XJ1100 Maxim with a fully transistorized breakerless ignition system. The system uses magnetic pick-up coils and eliminates the need for breaker points.

I replaced the factory ignition system with a General Motors (GM) Distributorless Direct Ignition System (DIS). I found several articles describing GM's Distributorless Ignition Systems and GM's DIS Wasted-Spark ignition system. The real advantage is power; these systems produce high energy sparks, which improve starting and provides a modern replacement for the original Yamaha ignition system.

I found a bunch of GM DIS units at Best's Auto Salvage yard in Pasadena, TX. However, a quick search on eBay shows lots of people selling these systems. GM installed these units on lots of cars, so take your tools to the local salvage yard and "go shopping". You want a complete unit; including the coils and electrical connectors; especially the connectors, they are very expensive. GM wants $38 for just one connector and you need three. I had to look through a lot of harnesses to find a set. You also need to take your digital ohm meter and a set of small 1/4" metric sockets for checking the coils. I found every coil tagged (#1 - #4) bad. Well, I found 4 of 4 bad; so be sure to verify the coils with an digital ohm meter. The primary coil should read approx 4 ohms and the secondary 11K-14K ohms. You'll need to insert a small wire in the primary coil socket to read the resistance. I ended up with two coils labeled #2-#3, which is okay.Just be sure to blacken the numbers or use masking tape to remember which coil is #1-#4. The DIS unit has the coil numbers stamped in the plastic casing near the connector plugs, just in case you forget, but you need to look carefully for the engraving.

GM DIS control unit with mounting bracket and two coils showing the bottom and top views.	The GM DIS unit assembled . The unit has two coils; one tagged 1 & 4, and the other tagged 2 & 3, however, every coil I tested, with the 1 & 4 tag, was bad. (The coils above are both bad, the good ones are on the bike.)
The GM DIS Control Unit.

The MegaSquirt II CPU directly supports ignition systems. Their General Motors DIS Ignition Control site is excellent in explaining it's operations. However, I'm using a MegaSquirt I processor. I made this work by installing the MegasquirtNSpark Firmware software and making the MegasquirtNSpark hardware changes to the MegaSquirt V3.0 circuit board.

Making the Reluctor Trigger Wheel and Mounting Shaft
The GM DIS ignition system requires a slotted reluctor wheel. I also needed a small shaft for mounting the reluctor wheel to the engine crank shaft. I made both by using a bench grinder, a drill press, and some shop tools.

I also bought a 6" caliper for making some engineering measurements and checking my construction details.

Making these two parts took forever, and a good machine shop can make these parts in less than an hour. However, I had problems finding a shop willing to set-up for a "little" job. So I built them myself. (All my friends are shopless!)

I went to a local welding shop and bought a piece of 6" x 4" x 3/16" steel plate and a 1" diameter steel rod (shaft). I think Home Depot and most local hardware stores stock this material. I cut and grinded the reluctor wheel to a 3-3/16" diameter circle; and I cut the shaft (with a hacksaw) to 3/4" and then grinded it smooth to 0.70" long. The center hole is 6mm. The design is based on the Reluctor Trigger Wheel drawing near the bottom of their web page, However, I changed the wheel dimensions; the wheel diameter required to fit the XJ1100 timing case is 3-3/16".

Construction Details - Reluctor Wheel

1. I needed to make a 3-3/16" circle from a sheet of 3/16" plate steel. The first thing I did was to locate a center hole. I used a center punch to mark the center. Do NOT drill, yet.
2. Then I took a protractor and measured 60 degree angles around the center. I use a scribe and scratched lines in the plate. Be careful; accuracy is important.
3. I then scribed a 10 degree line. This is the timing pulse sync line.
4. I took a compass and made a 3-3/16" circle. Practice on a sheet of paper until you get the radius correct, then mark the circle. Be sure the mark is visible. I used a permanent marker.
5. I used a hacksaw to rough-cut the circle.
6. Then a bench grinder to get it "almost round".
   
   

   Steve making the reluctor wheel.	Bench Grinder with a wooden grinding frame.

   
   
7. Then I drill a 6mm center hole.
8. I built a wooden grinding frame out of scrape lumber; some 2" x 4" material, a little piece of 1" x 4", and a small piece of plywood. The dimensions are NOT critical; all I was trying to do was get a stable platform so I could use my grinding wheel to create a round circle. I used screws to put it together, cause I had them handy and nails can slit the short pieces of wood. The important thing was to center the steel plate to the center of the grinding wheel. I cut my 2" x 4"s at a slight angle to improve the "fit".
9. I drilled a 6mm hole in the wooden frame and then inserted a 6mm bolt through the steel plate and frame. I was then able to turn the steel plate against the edge of the grinding wheel.
10. I put on some heavy gloves and started grinding. I turned the steel plate very slowly; letting the grinding wheel "chip" off the rough-cut edges. When I finished "chipping", the edges were rough, but the plate had a round look. Now for the finishing touch.
11. I took the "almost round" steel plate, used a 6mm bolt and nut and chucked it up on the drill press. Using a fine file, I dressed up the edges of the steel plate till it was shiny smooth. When I completed this turning, I tested the steel wheel by rolling it on a flat surface, it was perfectly round. I also used my calipers and measured several spots on the diameter of the wheel; "Close enough for government work."
12. I then took a cut off wheel and made six ??" x 1/4"deep cuts on the 60 degree marks and one on the 10 degree mark; seven total cuts. It took me several evenings to make this part, but I was able to do it using simple shop tools; a vise, a hacksaw, a bench grinder, some scrape lumber, a file, a cut off wheel, and a drill press.
    
    Construction Details - Mounting Shaft
    
13. I took a small scrap of lumber, 2"x4"x6", and drilled a 1" hole. You must use a drill press to make this hole, because it must be vertical.
14. I then used my Skill saw and cut the board in half, hole and all.
    

    My jig for holding the mounting shaft. I cut a 3/4 " piece of steel from the bar above to make the mounting shaft, and then grinded both ends flat till it measured 0.70" long. It will get hot, real hot...the blackened hole says it all, "Don't touch, you will get burned." Please use gloves and be careful.

    
    
15. Now the "fun" part ...I took my grinding wheel and chucked it into my drill press.
    
    

    My drill press "tool" for grinding the mounting shaft flat. I placed the mounting shaft near the center of the grinding wheel so it would be stable (less wobble.) Be careful and use standard safety precautions; you can break the grinding stone if you get too physical, and flying stone pieces are never a good thing.

    
    
16. I insert the steel shaft in the 1" hole and clamped everything in a drill press vise.
17. I turned ON the drill press and grinded the shaft "flat". I used a combination square to check the squareness. Be CAREFUL. The shaft will get very HOT and scorch the wood. Go slow. Ear plugs are nice.
18. Once flat, I flipped the shaft and ground the other end.
19. This took a while, but I got the shaft within 0.02 mils flat and square. Total length is 0.70".
20. Then I locate the center of the shaft and drill a 6mm hole; using my drill press.
21. The last step is to take a cut off wheel and make a small cut in the bottom on the mounting shaft so it fits into the slot on the side of the crankshaft; locking the mounting shaft to the crank. I used the old Yamaha's ignition plate as a guide.
    
    Making this part took several evenings, and I miss-drilled the center hole on my 1st attempt. Take your time, if I can do it, anyone can. However, a friend with a lathe could do this in an hour, where it took me days.

Assembly (Putting it All Together)
I installed the GM DIS Ignition system first, got the motorcycle running (using carburetors), and then installed the EFI system. I wanted to see if the GM ignition system would work before adding the EFI system.

Assembling the Ignition System
This wasn't hard; but it does require some welding. The GM DIS has three parts: the reluctor wheel, the VR sensor, and the electronic unit with coil pack.

The Reluctor Wheel and VR sensor replaces the Yamaha timing system. I removed the timing cover from the left side of the motorcycle, unscrewed and removed the timing plate and pickups. I removed the Yamaha ignition system (from under the seat), which was attached to the rear fender, and removed the associated wiring.

Yamaha's ignition parts: VR sensor plate, Vacuum transducer, and control module.

The new reluctor wheel bolted to the same place. However, there was one little problem, where do I mount the double slotted timing mark in relationship to top-dead-center (TDC). Where is TDC?

The reluctor wheel and VR sensor.

My approach was simple; I removed all the spark plugs, making it easier to turn the engine with an allen wrench. I put a thin screw driver in the number 1 cylinder and as I rotated the engine, my brother felt the piston come to TDC and start down. There's about 15 degrees of rotation where nothing seems to change. By rotating the engine clockwise and then counter clockwise, and holding the screw driver to the top of the rising piston, you can estimate TDC by marking a spot half way between the detected movement. This gets you real close; close enough for now. I took some white paint, applied it to the tip of a flat bladed screw driver and placed a marked on the inside of reluctor wheel mounting shaft, near the frame.

The reluctor wheel installed with the VR sensor. The far right timing make is TDC. Note: The reluctor wheel turns clockwise, so the timing marks (shown above in white) are After Top Dead Center.(ATDC), my bad! I should have put the timing marks to the right of TDC. As you accelerate, the timing marks advance or fire early, i.e. Before Top Dead Center (BTDC). I'll do better the next time.

Next, I installed the VR sensor. I took a thin piece of sheet metal and cut it about 1" wide and rolled around the sensor. I placed it close to the old ignition plate screw hole, marked the location, I drilled a hole in the metal holder and used the old ignition plate screw to attached it to the frame. The picture has a good view of my installation. I used a little RTV (silicon) to make the installation "stable" and lock it in place.

Now to set the timing. I used a 1/16" roll pin to lock the reluctor wheel to the mounting shaft. Here's my approach; First, I aligned the mounting shaft to timing mark I made earlier with white paint. Then I loosened the center bolt so the reluctor wheel would turned with effort, but not turn the crank. This was a little tricky because everything wants to turn. I then moved the reluctor wheel so slot 6; the slot just before slot 1 and the timing sync pulse, was just touching the VR sensor. Note: The GM DIS unit "fires" when it detects the slot. Slot 6 should be set just left of the VR sensor. I then took a very small drill bit and drilled a pilot hole through the reluctor wheel and into the mounting shaft. I only drilled the mounting shaft a little so I could get a mark. I removed the reluctor wheel and mounting shaft and used my drill press to drill a 1/16" hole into the mounting shaft to fit the roll pin. I also drilled the the inside of the reluctor wheel, just deep enough for the roll pin to fit, but be captured when I reassembled everything. If you look at the picture above, you can see a small pilot hole near the 3:00 o'clock position, near the center bolt.

One final comment; all this effort gets the timing close enough for starting. Once the motorcycle is running, the Megasqrirt controller electrically sets the timing. I'm off by at least 5 degrees; no big deal. I just re-programmed my timing curves with a 5 degree offset so that everything stayed in sync.

Last step: I made a timing pointer. It's barely visible at the 11:00 o'clock position. I added some timing marks to the reluctor wheel to the left of top-dead center. Don't make this mistake...the timing marks should go to the right of top-dead-center. Boy, what a surprised; I started the motor and my timing light "flashed" to the right on my marks. Oh well, things happen.

The GM DIS Unit
The DIS control unit fits under the tank near the yoke where Yamaha mounted the original coils. However, I made a few modifications: First, I removed the integral braking line (more on this later), then I chiseled off the front and back, coil mounting brackets. The control unit fits snugly, with the coils turned to the right side of the frame and the electrical plugs to the left side of the frame. I made two brackets out of 1" steel flat bar. I took the bar, drilled and tapped the mounting holes using the control unit as a template, and then dry-fitted the unit in place. Be sure to insert and remove the DIS connectors before welding the bracket; I got mine to close to the frame (the first time) and had the cut it off an start over. When I was satisfied with the dry-fit location, I marked the bend locations, removed the control unit, and bent and cut the brackets to fit. (cut to fit, paint to match.). A little welding, grinding, and paint.

Welding the front DIS bracket to the frame. The rear bracket is also visible.

Now add the wiring. The GM DIS control unit has three connectors; one for power, another for the VR sensor, and the last goes to the MegaSquirt controller. Get this, You DON'T need the MegaSquirt controller to run the engine; the DIS unit will operate in a "limp" mode and not advance the timing. .

I ran the power for the GM DIS unit from the master relay. And the ground came from the DB37 connector on the MegaSquirt controller.

The VR sensor wiring is a two conductor 20 awg shielded cable. I ran the sensor wiring down the lower part of the left frame, came up the frame just in front of the battery, and then forward on the right side of frame to the steering yoke. I soldered the sensor wiring to a GM DIS connector and used heat shrink tubing to seal the splice. I then plugged the sensor into the DIS control unit.

The wiring to the MegaSquirt cpu is a four conductor 20 awg cable. I ran this wiring down the right side on the frame with the VR sensor wiring and attached this cable to the External Wiring Harness.

Before connecting a fuel source, I used starting fluid. I bought this from an auto parts store and it comes in a spray can . I took several towels, wet them, and made sure my fire extinguisher was handy. I turned on the key, hit the starter, and it fired. Wow. Note: I did get several backfires, and a flash fire. Be careful, do NOT put your face near the carbs, It burned the hair off my right arm.

I made a temporary fuel tank out of a Jim Beam bottle ... and restarted the engine.
Here's Steve holding the temporary fuel tank!!!!!

Hey, it's legal, my XJ1100 was born in '82, so it's over the 21 year old drinking restriction. Please notice the outboard motor bulb in the right picture near Steve's left hand, I used this bulb to start the fuel flow.

When I was sure everything was working, I took a hacksaw and cut the front of the timing case cover to fit over the VR sensor. As they say in construction, "cut to fit". I used RTV to seal the hole in the case and the VR sensor.

Now that I had the GM DIS ignition system working, I was ready to install the MegaSquirt controller and the EFI System.

Assembling the EFI System
Installing the CPU, Fuel Pump, Throttle Body & Cable, Temperature Sensors & Narrow Band O2 Sensor

The first thing I did was removed the carburetors and air box; I threw these in a recycle box. Then I removed the enrichment cable, also called a choke cable.

CPU
I mounted the CPU directly under the seat just above the battery. I use a chisel and a cut off wheel to remove a few brackets attached to the seat support frame. I took a 1" flat steel bar, bent the ends approx. 90 degrees, and made a CPU mounting bracket. Steve tack welded it to the frame.

	(Front of bike)
MegaSquirt CPU and Mounting bracket

Fuel Pump
I bought a '99 Suzuki Hayabusa in-line fuel pump on eBay for $81, plus $12 shipping. Don't get confused, the later model Suzuki's have internally mounted pumps in the tank.

The Hayabusa fuel pump fit nicely on the rear tank frame bracket. I took a piece of 1" x 3/16" flat aluminum bar stock and bent it into a bracket to hold the pump. Then I drilled two small holes in rear tank bracket and used some small metric bolts to mount the pump. Shown below in the middle of the picture are two shinny bolt heads. Note, I locate the pump bracket as far forward as possible on the tank bracket to keep it from interfering with the seat tab.

The Hayabusa fuel pump fits nicely under the rear tank bracket. Unfortunately, all you can see is the fuel pressure regulator, with a vacuum hose attached, in the middle of the picture, near the electrical connector.	The fuel pump (under the rear tank bracket) & CPU

I used some heavy duty, fuel pump rated, rubber hose for the fuel lines. And I installed a fuel pressure gauge in the high pressure line, just in-front of the fuel rail. (Stan, insert a picture.) The fuel pump produces approx. 38 lbs.of regulated pressure. The high pressure fuel line is less than a foot in length, so it looks good. However, I might come back later and install a steel line, or a braded fuel line. Maybe it's over-engineering, but a broken fuel line could be serious.

The XS/XJ1100's has two petcocks, mounted on the left and right side of the gas tank. I used the right petcock to feed the fuel pump, and the left petcock as the fuel pump return. I installed a metal can type fuel filter in the feed line, cause we all know how gas tanks rust and pickup "garbage". I removed the fuel line octopus system and set the petcocks to "PRI", for Prime. Using this setting allows the fuel to flow straight through each petcock. Note: The petcocks have two connections per petcock, I plugged the connection, closest to the front of the bike, with a plug made from polyurethane fuel line and a metal bolt. I cut off the threads of the bolt and pushed the bolt into the fuel line. I then pushed the fuel line plug onto the connection. Note: The polyutherthane fuel line doesn't need a hose clamp; it fits tight.

At the top of the fuel pump is a regulator, which is visible in the picture. I ran a vacuum line from the throttle body to the regulator. At the bottom of the fuel pump, coming out the side are two fuel line connections. The one nearest the bottom is the pump suction (or inlet) tube. The one slightly higher is the return. This took me a while to figure this out.

Here's a mechanical drawing of the Hayabusa fuel pump. There are two fuel line connections, side by side, coming out the side. The one nearest the bottom, the one that has the 1 (material number) pointing at it is the fuel inlet.

Wiring the pump was easy, I followed the wiring diagram in the MegaSquirt V3.0 Assembly Manual.I mounted the fuel pump relay on the left side of the pump bracket I built and used in-line fuses to protect the circuit. If I do it again, I would buy a fuse bloc; all those in-line fuses are "tacky looking".

Throttle Body
The Suzuki GSXR 600 Throttle Body (TB) is basically a one for one replacement for the Mikuni carbs; they just slip into the intake manifold boots.. I used a complete 2004 Suzuki GSXR-600 throttle body; which includes the Bosche injectors, fuel rail, and throttle position sensor. I spent $62 on this eBay item. It did NOT come with a cable harness. which was a mistake on my part. I could NOT find a harness for a reasonable price, so I made one using some 18 AWG wire, some small electrical clips, and heat shrink tubing. I did find a 2003 wiring harness on eBay auction, but alias, the 2003 throttle bodies are very different from the 2004 and later models; the fuel injector connectors are different and the throttle sprocket is on opposite side of the motorcycle.

The GSXR 2004 throttle cable attachment lever and UNI (foam) air filters.

Installing the throttle body was easy, I used a little teflon spray oil, "slimmed-up" the manifold boots, and made a firm push; the TB's slipped right in.

I had to replace the throttle cable; because the attachment hardware was different and the cable was way to short. I was able to keep the original throttle grip, but had a tough time finding the new throttle cable. I finally found a motorcycle shop in Texas City (about 25 miles away) that had the parts. I purchased the cable with a black cable sheath, two barrel ends and attachment hardware for about $30. In searching the web, I found a web site called the Flanders Company where you can order the parts.

The throttle cable and attachment hardware.

The throttle body has two butterfly valves; the one nearest the injectors (front of the bike) is the throttle plates, and I didn't use the butterfly towards the back of the bike. I took some thin wire, drilled a small hole in the rear butterfly cam, and tied it open. Some people remove the rear butterfly plates and then seal the shaft holes with JB Weld; but I decided to leave them on.

One final comment, After I got the engine running, the idle speed was very high; over 2500 rpms. I loosened a throttle plate adjustment locknut (near the throttle plate cam) and slightly adjusted the plates. It only takes a small adjustment, so go slow.

Temperature Sensors
I mounted two temperature sensors; one for the "coolant" (CLT) and one for the air temp. Mounting was easy, especially since the XJ1100 is an air cooled engine and doesn't have a coolant loop. I wedged the CLT sensor between two fins near the middle on the engine just below the throttle bodies. I haven't secured this sensor yet, but I think a little Permatex high temp RTV silicon should work nicely. MegaSquirt uses the CLT sensor for warm up settings.

I mounted the air temp sensor on the right side of the bike just behind the side panel using an existing brake line bracket.

The XJ1000 originally had integral braking; where the foot peddle controlled both the rear brake and front left caliper. I hated this design, so I replaced the old braking system with stainless steel brake lines, and re-routed the lines. The hand brake now operates both the front calipers and the foot peddle operates the rear brake. I inserted the air temp sensor into a rubber-grommet bracket originally used for the forward running integral brake line. Note: I love the new braking system, this baby can stop.

Narrow Band O2 Sensor
I mounted a narrow O2 sensor in the right exhaust pipe. My bike has dual exhaust; cylinders 1&2 combine go down the left side and cylinders 3&4 go down the right . On the right side where the pipes combine and just before the muffler, I had the local muffler shop weld on an O2 bung. They did a great job; but I'm unhappy with the mounting angle. I wish I had taken extra care in marking the location and slope; the sensor is a little to vertical. If I hit a deep hole or jump a curve, I could damage the sensor. Oh well, Life's never perfect, I'll live with the imperfections. It works just fine. The next time, I'll take some clay, stick it to the pipe and insert a wooden stick to mark the angle.

Tuning the EFI System
I found this on the MegaSquirt User Forum in October 2006, "So, basically, get your AFR consistent, adjust the AFR on the dyno up or down for maximum power, then adjust timing. You may find that you will loose power at some RPM points and gain at others, but a little analysis will help you optimize the advance at all RPM points." This made a lot of sense; I decided to follow his advice.

I need to define some terms (before we get started):

Let's get started...here's a set of my MegaTune Displays; they're handy for setting up your initial values and for reference. First we need to configure the basic settings: the Engine Constants Table, the EGO Control, Fuel VE Table, Accel Decel Mode, RPM Based Accel, and the Code Based & Output Functions.

The Engine Constants 1 table is dependent on several factors; your fuel pressure and the fuel injector ratings are the most important. I used a '99 Hayabusa in-line fuel pump creating 38 lbs. of fuel pressure and fuel injectors from a 2004 GSXR600 Suzuki. I adjusted the Required Fuel value and found that an injector pulse width of 1.6 ms at idle was about right. I set the Injector Staging to "Simultaneous".

Configuring the fuel Volumetric Efficiency (VE) Fuel VE Table was a little harder. I guessed at the Y-coordinates values (the MAP axis) and selected 20, 30, 35, 45, 50, 60, 65, 70, 80, 85, 95, 100. I set the X-coordinates values (the RPM axis) to 1000, 1400, 1800, 2600, 3400, 4200, 5000, 5800, 6600, 7400, 8200, 9000. This table is a great start, assuming your Engine Constants 1 table closely matches my setting.

Configuring the Spark Table was easy; I took the Yamaha timing charts and typed them into the table. The final tuning requires a Dynometer; so for now, I’m running with my original configuration values. If you compare my ignition table with the Yamaha timing chart, you'll see a little differences; I made a few tweaks as I gained experience riding and listening to the engine.

Crank Setting and Warmup Enrichment

Tuning the Engine:
I knew NOTHING about tuning an EFI engines. I know more now and continue to learn more as I ride and study my data logs. Reading date logs are important, so follow these steps:

• “Read the book.” MegaSquirt’s manual is excellent.
• Every time I accelerated (from a stop), the engine bogged; slowed down. The VE table is TOO lean.
  The Acceleration Table: I tried un-successfully to set the acceleration table (add a link). I finally tried the RPM based Accel Table (add a link) and things got better.
  Set up your
  MS saves the log files in a spreadsheet (.xls) format. The MSTweak2000 can NOT read this file if it has words in the header, “…29q”. Use Notepad and delete the bogus entries.
  Insert pictures with my motorcycle, and battery.
  

later, I installed the MegaTune software to my laptop, as well as some MegaLogViewer software

It's running ...Walter and Steve looking at the Megatune display.

Final Report
After all this work (and money), how's it running? I'm glad you asked...the Bottom Line .

Stan's Biography
I was born Stan Hutchison on June 30, 1954; I attended High School in San Antonio, TX and graduated from Texas A&M University '76 with a BS in Electrical Engineering. I married Deborah in '73 and we have two kids, Scott and Robyn, and one grandchild, Katherine. I work for NASA in the Mission Control Center at Johnson Space Center, Houston, TX.

Send me a note if you have questions or need help.

It does get any better ...

Fishing, more fishing (offshore), and motorcycles. Stan's other ride, yea, it's old, but hey, it's a HONDA.

Acknowledgements (The Good Guys):

• Deborah, my wife; she rocks, lets me play with men toys.
• Steve Hutchison (Hutch), my twin brother, kept me going and provided time and skills. My brother Steve helped me build and install the ignition system and also the EFI system. He also welded the DIS and MegaSquirt mounting brackets on the bike.
• Mary Helen, my sister, provided the laptop. Way Cool.
• Walter, a motorcycle friend, "Check'in us out."
• Best Auto Parts (auto salvage yard) in Pasadena, Texas donated the GM DIS unit with coils, very nice guys.
• Dan Mann, in Michigan, provided excellent insight and encouragement, a great sounding board; converted his XS1100 in '04.

Links to some of my favorite Web Sites

Back to Top