Originally Posted by Benzo_Mike

I have a 1996 CBR1000F that currently has 82,000 miles on it. I've completed three restorations/builds on it, including a COMPLETE nut and bolt disassembly/rebuild last year. Engine completely apart down to the last washer and nut, new EVERYTHING inside with a thorough blueprint/build. I have massaged every inch on the bike mechanically - suspension, brakes, induction, exhaust, motor, electrics, etc...

In the course of accumulating those miles I have encountered every possible issue/quirk that these bikes have, and I have VERY tight resolution to observe these quirks as the bike is bristling with aftermarket instrumentation, including a wideband. So I don't let anything go, if it is the SLIGHTEST little bit "off", I chase it down the rabbit hole and solve it.

The problem described in various ways through this thread is a common and ubiquitous one with this bike (and many of its contemporary cousins). EVERY one of these bikes (with the exception of perhaps a couple of very heavily modified ones) has this issue. It is an engineered-in flaw of sorts. It took me a while to figure it out and solve it (around 8 years ago) and it was driving me CRAZY, it is truly a genuine pain in the ***. The good news is that this issue is 100% correctible by fairly conventional means, if you decide to pursue it. And there is no practical downside associated with the solution.

The root of the problem has been correctly stated in this thread. The charging system (especially with everything stock and adjusted to factory specs) does not push NEARLY enough electrons at low RPM (most notably idle) to keep up with load. So yes, there is intermittent discharge of the battery in stop-and-go. But the real issue is not so much the effect on battery charging, but rather on live, real-time system voltage. A stock bike, with the idle set to the factory spec of 900-1100 RPM (for a 49 state US bike) will not produce nearly enough current to support the bike's electrical system. This is especially true in traffic, at a light, brake applied, turn signal on, etc. Recall this is old-school incandescent bulb technology, which eats a TON of current. The big failure comes when the bike heats up and the e-fan kicks on. That motor pulls a BIG surge of current when it kicks on and spools up and if you simulate this situation as I have described and watch system voltage with a multimeter, it will commonly drop below 10 volts at that instant, which is not enough to reliably power the CDI/old-school coils with HT leads and the bike will misfire, with an engine stall approximately 50% of the time.

Originally Posted by Benzo_Mike

So the 100% foolproof solution is a s follows...

First, raise the idle speed to ca. 1350 RPM. The alternator will produce a LOT more current than it will at ca. 1000 RPM. The alternator is pretty small/weak in the first place (rated at 28 A), so every little bit helps. This adjustment alone will not solve the issue, but it will get you a good bit of the way there.

Second, replace ALL of the bike's illumination with LED bulbs. I have done every single bulb (including all the little ones in the instrument cluster) with the exception of the headlights (I run dual 55W HID). In a typical stop-and-go situation, this will relieve ca.10-15A of current from the electrical system and in conjunction with the idle speed adjustment will solve the issue outright.

I have added several more layers of modifications to the solution on my bike, none of which are strictly necessary but all help. I have a full stick coil conversion and have upgraded the regulator/rectifier. It is VERY unfortunate (due to the wiring layout of the alternator) that we can not do a Shindingen MOSFET conversion, which would be VERY helpful. I also tuned up the cooling system as much as I could to help minimize the e-fan duty cycle. So thorough cleaning of the radiator and engine water passages, all new custom silicone hoses, all new fittings, complete water pump rebuild, HOAT coolant conversion, and a nice, cold 150 degree thermostat from an early-generation Accord. This all helps.

Then as a final layer of protection I put both the headlight and e-fan circuits on separate, switched relays, so I can shed even more electrical load with the flick of a switch (9 A alone for the headlights). This is a guaranteed backstop that I have never had to rely on.

My bike now idles rock-steady, even under the heaviest electrical load (fan cycling when hot). There is nothing that can be sensed when the fan kicks on, other than the surge of hot breeze out of the bodywork. I'm close to 13V now with the fan running.