roger 04 rt said:
August 8, 2013
When I tested an R1200 GSW the other day, one of the improvements is that cold the bike seemed to start like, er, er, vroom. My 04RT is more like er, er er, er, er, er, vroom.
...
bike runs great except the slow first-start-of-the-day. I've GS-911ed all the sensors, looked at the LC-1 AFR charts, nothing seem to be wrong and the Motronic seems to be computing reasonable first injection times.
...
Any thoughts in how to get my bike to start faster 1st time in the morning? Er, Er Vroom!
August 24, 2015
Two years later, my '04 R1150RT (dual spark) now starts in about
1 second after sitting overnight and in about half to three-quarters of a second once the engine has warmed up (>40C). The final pieces of the puzzle were wiring the stick coils through a relay, directly to the battery (last month); and replacing the starter (last week). I’ve made it to Er, Er, Vroom!
Since it seems I'm at the end of this project finally and have learned a lot about what it takes to get the dual-spark 1150 to start quickly, I want to summarize the various improvements and repairs that I’ve made to reach the present state.
1) The Motronic, injectors and stick coils are voltage sensitive and the Motronic MA 2.4 doesn't fully compensate for low voltage. I experimented a lot by powering the electronics separately and the bike always started well when I did. A good battery, good starter and properly powered stick coils are all key.
--Odyssey PC-680: My battery had been chronically undercharged due to short trips, the 1150RT's alternator being a lower voltage than the PC680's spec, and my use of a Battery Tender Jr. with low voltage output. The battery was successfully rejuvenated by discharging and recharging 5 times, the BT Jr. was replaced with an AGM charger and I boosted the alternator output with the addition of an internal diode. I also modified the Kisan Signal Minder Flasher so that there was no key-off current drain on the battery.
--The stick coils were powered from the key switch and its contact resistance seems to have increased over the years. Later model dual-spark bikes power the stick coils with their own relay, called Load Relay II. Recently I added a relay and rewired the stick-coils (and while I was at it the secondary coil too) directly to +12V through the relay. It improved starting, idling quality and (surprisingly) top-end smoothness.
--The planetary gear cover inside my starter had fallen onto the armature a couple times causing serious internal arcing. I replaced the cover and the starter seemed fine but after fixing everything else in this summary, I got to the point where some starts were still slow and I knew it was related to the starting dip in the +12V. It seemed there must be excessive draw from the starter some of the time, causing the voltage dip and upsetting the Motronic, injectors and coils. After replacing it the motor spins faster, and quieter during starting and the bike starts very quickly whether cold or hot.
—To idle well in cold weather immediately after starting, the alternator needs to start charging the battery ASAP. I noticed from GS-911 logs that the alternator sometimes would take up to half a minute to start charging the battery. During that time, voltage at the injectors could be less than 12 volts, resulting in a lean, rough idle. This is caused by a much slower injector turn-on time at the lower voltage. The alternator starting current is supplied by the Batt/Alt bulb. By replacing the stock 1.7W annunciator bulb in Batt/Alt with the 3W bulb from the High Beam annunciator, the alternator always starts immediately and cold-idle is smoother.
2) All 4 spark plugs need to fire well for a quick start.
--Both stick coils were found to be sub-par and although the bike ran well, replacing them improved starting and idling quality, as well as top-end smoothness.
--The lower plugs seem to foul easily and although I didn't notice it while riding, fresh lower plugs made a difference to starting and cold idling. I've gone up in the heat range on the lower plugs to try and keep them cleaner.
3) The Hot Idle Speed, TPS and Fast Idle Lever need to be adjusted correctly for best starting.
—The BMW procedure for setting the Fast Idle Lever results in the throttle being opened about 4 degrees when the Fast Idle Lever is in the mid-detent position. Mine was at about 1.92 degrees, and many I’ve seen in GS-911 logs are lower, some as low as 0.64 degrees. Although the bike will start with that little throttle, 3-4 degrees open leads to faster starting. The adjuster for the fast idle lever ran to the end of the threads at a little over 2 degrees so a spacer was manufactured to sit in the bottom of the ferrule & adjuster to add adjustment range. After trial and error, 3-3.5 degrees when the Lever is in the mid-dent leads to a fast start but not too high a warm-up idle speed.
—Oddly, but confirmed on several bikes, if the Hot Idle Speed is set above 1100 RPM, the cold-start enrichment with the Fast Idle Lever OFF is too lean. This is because there is more air entering than expected at closed throttle while Open Loop. By reducing Hot Idle to 1100 RPM, the bike’s Cold Start AFR is slightly richer.
—TPS: Although a longer topic, I realized along the way that the closed throttle voltage expected by the Motronic is 340 mV. It was also discovered that the infamous zero=zero procedure results in the throttle being opened too far. The correct zero degree voltage is 250 mV.
Summary
Although everything on the list above contributed improvement, I found the most important were: 1) a fully charged, properly conditioned battery; 2) a starter that didn’t glitch the battery voltage badly, 3) separately powered stick coils and 4) good spark plugs. Although many in the forum helped I want to especially thank DR, GS Addict and terryofperry who provided valuable advice on debugging, starters and AGM batteries.
Borrowed a starter off a friends 1150 a well used one and she fired up within 
