deilenberger
A bozo on the bus
Also throttle sync is an issue at and just above idle, above that the ZFE adjusts mixture based on the O2 sensor readings.
I'm not understanding this too well... let me think it out out-loud:
At idle the sync is handled by the stepper motor bypasses. The throttle plates are actually fully closed (unlike oilheads) at idle. Right off idle is where the sync is most easily checked, since the amount of difference between the two TBs is a larger proportion of the total airflow - meaning it's easier to see. The actual sync should retain whatever offset there is (if any) throughout the range of throttle opening - if your cables are good and the junction box is good. Which is why we suggest adjusting the sync (if needed) just off idle (between 1,200-1,800 RPM.)
BMW's instructions - if you can decipher them on the REP-ROM are to note the amount of offset in an actual value at idle, and maintain that offset (not percentage of offset) throughout the throttle range. IE - if the right side offset is 0.01mBar at idle - then at 2,000 RPM the offset AMOUNT should remain the same. If the left side read a total of 1.00mBar, then the right side should be 1.00mBar + 0.01mBar for a total of 1.01Mbar. In my experience, if the valves are adjusted correctly - you can come fairly close to perfect throttle-body balance over the full range of engine RPM (mine is very close) at least on the hexheads. Dunno on camheads - haven't done one.
While the O2 sensors do correct mixture (most of the time) - this is really a separate function from the throttle body balance. There are two O2 sensors - one for each cylinder, and even if the throttle-body balance is out, the ECU will correctly adjust the mixture for EACH cylinder for whatever amount of air is in the combustion mix.
The O2 sensors DO adjust the mixture at idle - this is easily seen on the GS-911. The heated sensors used on the hex/cam-head are quite fast to start working (within about 15-20 seconds of when the engine is started), and you can then observe the output switching hi-low using the GS-911 plotting feature. The switching is an indication that the ECU is adjusting the mixture (if the O2 sensors saw a constant mixture at idle - they would have a constant output.)
The vacuum hose that goes to the left throttle body is the one from the purge valve for the charcoal canister. The electronic purge valve only opens under certain engine conditions. The amount of additional air introduced to the left side throttle body is small in comparison to the amount of air that is flowing through it at the RPMs in question. While it might seem to introduce a small imbalance - I suspect it is really not a major contributing factor to the vibration. It's simple enough to check - just disconnect it and cap off the input port on the throttle body. If the vibration is reduced enough to be noticeable - then you've found one smoking gun. FWIW - my charcoal canister was in the way of adjusting my front shock, so it's been on the shelf and I just haven't gotten around (in 76,000 miles) to putting it back on. I did cap off the throttle body, and did a bit of re-routing of the hoses so the tank vent now goes to a vent hose near the left footpeg.
The goal of balancing the TB's is to get an equal amount of power from each cylinder on each combustion cycle. This is when the engine runs the smoothest with the least vibration. The amount of air going into the cylinders should be balanced in order to help achieve this. The amount of fuel going into each cylinder should also be balanced, and the valve openings should be the same. All of these will have an effect on the combustion balance of the engine (as can other things such as bad ignition, or carbon buildup resulting in different compression in the cylinders.)
Dunno about the bottom coils/plugs only being used at lower RPMs, I've never seen anything about this. Just noodling in my head - since the purpose of the two plugs is to help start a larger flame front on the ignition of the compressed mixture, to my way of thinking, this would be equally important at higher RPMs as at lower RPMs. But that's just noodling - don't mistake it for any sort of factual data.