The GR650i

First Steps


The GR was our first attempt at building an injection system and bolting it to anything, so as you can imagine it was complete chaos. We had to make the whole thing up as we went along. Albert Einstein once said that “If we knew what we were doing, it wouldn’t be research!” By that definition, injecting the GR was Nobel-quality, cutting edge, revolutionary research of the highest possible calibre.

The Mk I GR Manifold. Amazingly enough, it fit perfectly. Richard threw a prototype ECU together and produced a manifold by welding some bits old gas pipe onto half a Corolla carburettor. Neil bashed out some code which would hopefully turn injectors on and off at the right time, and Steve supplied the bike. He also managed to arrive upright and sober. Impressive! (for him)

Neil had never set eyes on the GR before starting work on trying to inject it, and only identified it as a motorcycle on his third attempt. With his cry of “You want to inject THAT heap of crap?” ringing in our ears, we threw the bike, the software, the manifold and the prototype ECU together and hoped for the best. It wasn’t pretty.

The fuel supply for the VL Commodore fuel pump was pretty easy. As you can see from the photos, we just took the cap off the GR’s tank and threw in a couple of bits of fuel hose. Sorted. Not the safest or most professional solution, but very easy and extremely simple. Just like us, really.

Richard’s prototype ECU varied the mapping on a 2D map with intake manifold vacuum on one axis and RPM on the other axis. It was a loose collection of printed circuit boards, breadboards, wires and chips which took up half a workbench. Most of the connections were just wires twisted together. It didn’t look very impressive, but it worked. Barely.

While Steve unbolted the carbies and installed the manifold, Richard and Neil hooked the ECU up to the GR’s ignition pickups. By timing the signals from the pickups, the ECU could calculate where the engine was and fire the injectors accordingly. It’s not as accurate as the cam/crank angle sensors used on factory EFI systems, but it’s also a helluva lot easier and cheaper. Retrofitting a cam angle sensor to a cruddy old GR mill was most certainly not an option at this stage.

When we tried to crank the engine over to check the signal from the pickups, we ran into our first snag. The GR’s starter motor. Rumour has it that Suzuki bought a cheap batch of starter motors that had fallen off the back of a ship in the early ‘80s. Allegedly, they were recovered from the briny, re-crated, shuffled through a complex network of warehouses, given about three sets of forged Kanban cards, driven around Hammamatsu a few times on the back of unmarked semis in the dead of night and eventually ended up on the GR production line. When installed, they had enough grunt to start the new bikes once or twice in the showroom, so Suzuki was happy. This particular starter motor must have been dropped on its little windings a few too many times by kamikaze forklift jockeys during the parts laundering scam. Added to which, it wound up in a parking inspector’s bike. What do parking inspectors do lots of? Yup, stopping and starting. It had a hard life, and after 16 years of abuse it was absolutely shot. Turning over the GR for hours on end while we fiddled with oscilloscopes was completely out of the question.

The only way we could get the engine to turn over freely was by removing the sparkplugs. Without any compression, the starter wheezed away, the crank went round in little circles and the line on the oscilloscope jumped hither, thither and yon. Oops. It shouldn’t do that! Say hello to snag number two.

Richard and Neil's high-tech diode bridge. Refined, elegant, inspired design. Faultless craftsmanship. Words cannot describe. Truly exquisite. (etc etc.) Unfortunately, the signal from the ignition pickups was very noisy, and it immediately crashed the ECU. In order to filter the noise out of the signal, Richard and Neil bashed their heads together and produced a high-tech diode bridge. Again, not the highest quality workmanship you’d expect, but definitely no worse than the rest of the setup. Not much worse, anyway.

Steve administered CPR (and a spot of massage and aromatherapy) to the ailing starter motor and managed to revive it sufficiently to push one piston past TDC with the sparkplug installed. It would only turn over about one time in ten, but that was enough to give us hope. We wired up the corresponding injector, crossed our eyes, legs and fingers, and kept pressing the button until she fired. In a miracle we still don’t quite believe, she burst into life the first time she turned over. Our very first attempt at injection and we had succeeded. Truly, it is better to be lucky than good.

Yep, it ran like that. We couldn't believe it either.    

 The first test run
37 sec, ~625KB (mpeg)

Running on a single cylinder
and the street test


We had a go at doing some rough tuning, but without any load on the engine we were pretty much wasting our time. Of course, we didn’t have a dyno, so the logical next stage was to strap all the crap onto the bike and get out onto the road…

<<< The Saga                        On the Road >>>