Great Bug Hunt
Being the infuriatingly anal perfectionist that he is,
Neil decided that getting the CBRi running well was his cue to completely
re-write the ECU software. He claimed that the first version was written
while he was pissed, which certainly explained some of the problems we’d
been having. Version 2.0 of the software was produced without the aid of
ethanol, and Neil proudly declared that the routines for controlling the
injectors and calculating engine speed from the ignition pulses were much
faster than they had been before.
While Neil was screaming for vodka and bashing his head against
his laptop in an effort to improve the code, Richard was pushing the
CBRi back from the local servo. He’d taken it down there in search of more
go-juice, and a power spike in the ignition system reset the ECU and
nuked the fuel map. Richard didn’t have his laptop or mobile phone, so
he didn’t have much choice but to push the bike home. It’s 1.68km from
the servo to Richard’s place, and most of it is uphill. When he recovered
from the exertion, he bodgied up a flash card addition to the ECU. The
Mk IIa ECU could remember it’s fuel map through resets, power loss or
acts of extreme stupidity. Richard also wrote a program to store and modify
maps in his Palm Pilot and upload them to the ECU.
We loaded Neil’s V2.0 software with high hopes, and were disappointed
to find that it controlled the injectors so well that the old maps didn’t
work anymore. We tried to tweak them, but without much success. Sometimes
the bike would start without any problems, but sometimes it wouldn’t start
at all. A bit of stuffing about with an oscilloscope revealed some of the
problems. The new code was much more susceptible to noise in the ignition
circuit than the old code, which played havoc with the injector timing.
V2.0 handled the injectors much faster and more accurately when it was
getting the right signal from the ignition pickups, but when it wasn’t
getting the right signal it was squirting fuel onto a closed intake valve
and the bike ran like a dog. This was most noticeable at low RPMs when
the engine was cold, so the CBRi was now almost impossible to start cold
and run for any length of time. We were keen to try the new code at high
RPMs where any fuelling improvements would be magnified (with a corresponding
magnification of horsepower), so we bolted the carburettors back on to get
it running and warm. The plan was to quickly convert it back to injection
and get back on the road, but our haste had tragic consequences.
Richard got a tad too enthusiastic about seating the intake stubs
in the carb rubbers with his beloved “knockometer” (hammer), and the aluminium
manifold literally came apart at the seams. Various repairs were attempted;
hot glue, Steve’s tube of Knead-It and finally the MIG welder. None were
particularly successful, so it was back to the drawing board for a change
of material and some revised geometry. Aluminium, although light, had
been a hassle to work with and so the Mk II manifold was welded up out
of chunky 2 and 3mm steel plate. It was specifically designed to withstand
a psycho belting the crap out of it with a hammer. The main changes from
the Mk I were a thinner body profile to bring the injectors closer to the
ports and internal strengthening ribs.
The new manifold was fully airtight, unlike the old one, and that
made cold starts and low-speed running much easier. In fact, the CBRi
now started better than it ever did with carbies, and needed much less
fuel than before. However, it also had a tendency to flood, which meant
more plug extractions, more skinned knuckles and lots and lots of bad
language. It wasn’t long before we discovered that the ECU was sometimes
leaving an injector turned on continuously. Occasionally it would flip
into a schizophrenic frenzy and trigger the injectors at random. The result
was always an oversupply of fuel and often, complete engine hydrolock.
More plug extractions, but no bent valves. Thank Saint Soichiro-san for
a tough little motor!
While Neil searched the code for the injector control bug and
tried to find a solution to the noise problem from the ignition pickups,
Richard disappeared into the gloom of his electronics workshop, intent
on mayhem and destruction. He emerged some time later proudly displaying
the Mk III ECU, a box about twice the size of a pack of smokes. The Mk
III does everything the Mk IIa did except take up as much room. Of course,
the Mk III still has a blinky LED. We checked with Richard’s psychotherapist,
but apparently his fixation is harmless. We still try to keep him away from
sharp objects, though. We’d like to confiscate his knockometer as well,
but we’re afraid of what he might do if we try.
The fuel rail was also working perfectly well as it was, but that
didn’t stop Richard from tinkering with it. We keep telling him that
if it’s not broken, he shouldn’t play with it, but he doesn’t listen.
Anyway, Richard decided that the short sections of fuel line connecting
the injectors to the rail were too bodgy. It was probably those hose clamps.
Richard hates hose clamps. Steve was in favour of just ditching the clamps
and hoping for the best (after all, we’ve always mixed fuel and electrics
before and nothing has exploded yet), but he was outvoted. A new fuel rail
was soldered up from 10mm copper pipe such that the injectors would plug
straight into it. A sprinkling of rubber gaskets to keep everything sealed,
a couple of springs to hold the whole shebang together, et voila – the Mk
II fuel rail. It only leaks a little bit. The main advantage of the Mk II
is the fact that it is solidly mounted onto the manifold rather than being
a press fit.
At time of writing, Neil is embroiled in a fight to the death
with that great nemesis of all post-grad students called Thesis. He is
also incapacitated by a mystery virus, which may be unknown to science
but is far more likely to be an extreme variety of procrastination. Richard
is slaving away in the salt mines of academia, working 18-hour days for
very little money and absolutely no recognition, and Steve is being paid
by the Government to sit on his bum and pick his nose. It’s a tough job,
but someone’s gotta do it, and let’s face it, he’s not qualified for much
The CBRi remains temporarily bug-ridden and off the road, but
as soon as Neil stops coughing up green slime, when Richard is dragged
from his office like a white and pasty grub being plucked from a rotten
log, and by the time the Queensland Government can function without Steve’s
invaluable (cough cough) assistance, progress will continue.