All I did was change the battery. It was idling normal before it lost battery power.
Is there any way you can get hold of a GS911 to do some diagnostics? I suspect as before, your TPS is mis-adjusted and it takes time for the system to try and compensate or the TVA is bad or out of adjustement, possibly a combination of things.
I believe I remember Sailor ( GS911 guru) stating that a bike when properly adjusted will start and idle with the TVA removed. If it won't do that, there is definitely something wrong. A GS911 might help isolate the issue.
Just to clarify and correct earlier post:
(1) By design, the TVA (idle actuator) is needed to idle at proper values of 950-1050 RPM by specs, but 1100 is very acceptable.
(2) By design, if TVA is removed, the somewhat hidden main idle screw would the new idle stop for throttle-bodies AND this would create a too low idle of approx 600 to 800 RPM (engine would die in most cases). Hence, based on above, you need a functioning TVA to idle properly.
(3) In addition to stable idle, the TVA also increases idle for a certain time in cold-start conditions. Idle may go between 1200 to 1400 for up to 30 seconds depending on how cold the engine is after being stopped for a full night.
(4) in addition, the TVA will stabilize or increase idle in very high load electrical demand like when using the reverser.
(5) the Oxygen-sensor is a not a critical item during start as its data is not used by Motronic until a waiting period (30 sec to 1:20) depending how cold the engine was.
HOWEVER, disconnecting the Battery (or removing MOTRONIC-EFIc fuse) for 10 minutes or more should NOT cause the problems he is seeing. BOSCH MOTRONIC 2.4 that we have on K1200 "brick-engine" will learn certain specific values because of assembly and wear tolerances for each bike. These learned values are lost when battery is disconnected, but ASSUMING all components (and engine state of tune) are withing acceptable tolerances, you should still idle fine after a memory reset caused by a battery removal.
This is a sign that something else is a bit out of normal tolerances. Unfortunately, these Motronic EFI are fairly old technology from 1990s and there is only so much the GS911 can tell you. You can check faults from TVA or TPS - you can also check TPS base setting, but the GS911 cannot know if you have low fuel pressure or a leaking fuel injectors. These 2 latter cases would affect cold start and can ONLY be checked with other tools or other methods.
Another common problem affected the "learned parameters" is the fact the main pulling throttle cables may have incorrect free play. This would affect and confuse the EFI when the ignition is turned ON as it is expecting a specific TPS position within a certain limited range when throttle is resting at idle. If cable free play is too small (too tight), the TPS is setting higher than normal AND by direct consequences the throttle butterflies would be opened more than they should.
One should NEVER troubleshoot these like you would with a simple single cylinder equipped with carburetor: the EFI system is easily confused when parameters are out of normal range during start.
A GS911 fault readout would help make a decision -OR- eliminate certain type of faults. To help further, GS911 can also give us a real-time log of the all parameters when engine is started. BUT... based on all tests I have done in the past (some very strange just to see EFI behavior) I continue to think that any K1200 "brick-engine" that is well tuned should not need anything when battery is unplugged for more than 10 minutes. Of course, there will be some "learning" when you ride but this is mainly to gather Oxygen-Sensor data to fine tune the base fueling Map for you specific engine (wear, tolerances, riding conditions....)