This page will log all build progress from early january
2024. Before this the following work was done:
Oxygen sensor bungs welded to exhaust
Found a source for ECU board headers
(ignitech.cz) and designed + fabricated a breakout board sometime around august 2023. Huge thanks to Ignitech because the 22 pin sumitomo connectors for the original ECU are literally impossible to find anywhere else and don't even exist in any public catalogs
Designed a RusEFI hellen based ECU that would
connect to the original wiring harness and fit in the place of the original
ECU
Mostly wired up the breakout board to run the
bike on a MicrorusEFI with a igniter module from a Subaru SVX and rusEFI
dual wideband O2 sensor controller
Installed alpine linux on a raspberry pi 4 so I can run Tunerstudio's auto tune and logging functions on the ancient microrusEFI revision I'm using since it lacks bluetooth or onboard microSD logging
Block diagram of initial setup for running the bike on a microrusEFI. The final version will use a custom RusEFI based ECU that won't require any wiring besides two Bosch LSU 4.9 oxygen sensors
1-11-24: ECU connectors for new ECU design ordered from
Ignitech
1-14-24: SEE THIS? I HATE IT
1-15-24: All wiring done, ECU connectors arrived.
1-19-24, 2AM: Harness side connector going to microrusEFI replaced.
If you're buying a microrusEFI get the first style of connector shown
here with the removable front key. I have the side clip style which has been obnoxious to re-assemble after depinning and comes with flimsy pins that are very easily deformed. The model you want is Aptiv F765000
At some point before the next two entries there was a night of drinking and wiring where somehow a bunch of current went through the switched power for the wideband O2 sensors and CANBUS plug and fried the processor). This meant the ECU was completely dead until the STM32 was replaced, and after replacing and reflashing the STM32 there was no useful signal coming from the cam output of the TLE8888. I misdiagnosed this a a dead IC and replaced the TLE8888 (signal conditioners and IO for engine management), but the problem still remained. The actual problem was the VR+ pin on the ECU connector had no continuity which means I probably broke a trace when harvesting the microrusEFI from the board it originally lived on.
2-2-24: Turns out the TLE8888 wasn't dead, there was a broken trace or something on the PCB and the VR+ pin wasn't connected to anything, after adding a bodge wire there's now trigger events in the console and this should be running as soon as I can borrow a timing light.
2-11-24: Major progress since last update! I spent a lot of time trying to set the trigger offset with no success. The cam trigger signal for a VFR800 has three pulses, one low amplitude pulse followed by two quick pulses with double the amplitude 180 degrees later. TLE8888s have some kind of algorithm for filtering out noise that was causing trigger errors due to the amplitudes mismatching. After adding a VR sensor signal conditioner before the microrusEFI and changing the resistors on the MRE back to the hall effect configuration the trigger errors stopped and I was able to set trigger offset. As of right now the bike starts but isn't idling as the tune isn't set up outside of vague ballpark numbers.
2-20-24: OK I DON'T CARE IT RUNS ENOUGH I'M ORDERING THE NEW ECU DESIGN PATIENCE IS FOR COWARDS WHO FEAR CHANGE
3-23-24: it took this long for all the parts to be in stock lol
3-31-24: Replaced stock voltage regulator with a Shindengen FH020AA. The stock regulator has lasted 40005 miles and I don't want it to give up and cook my fancy new ECU when it gets here in another few weeks.
4-4-24:IT ARRIVED
4-11-24: Bike started on the new ECU for the first time! It doesn't idle yet and I'm still waiting for the rear connector for the oxygen sensors but the ECU works and only needed one minor hardware fix to make the tach work.
4-12-24 Killed the engine
4-18-24
Over the past six days I pulled the old engine and bought a used engine with 27000 miles. While everything was disassembled I got the centerstand powdercoated, replaced the stator in the new engine since it already looked pretty beat, replaced the swingarm bearings, and ordered a new rear sprocket since I forgot to do that when I replaced the chain and front sprocket in 2022.
The final parts for the ECU arrived while I was busy pulling the engine and once the bike is revived I'll actually be able to run the ECU with oxygen sensors and a proper case
5-3-24
Bike is back on the road, oxygen sensor harness for new ECU done, turns out almost all of the pull up resistors on the analog inputs of the ECU were wrong and upon fixing those all the sensors report reasonable data.
Bike running on new ECU for real soon
7-2-24
I've been busy and haven't worked on this much. The last two months were spent trying different schemes for cam/crank position sensing, and I suspect it's still not keeping accurate time at low RPMs as the most reliable configuration still is hard to start and doesn't like to run under ~2000 RPM. It looks like a hall effect cam sensor from a mid 2000s onward Honda/Acura has about the same dimensions as the VR cam sensor used in the VFR, so I've ordered one to hopefully solve any timing issues.
