Making the Meter Mounting the Meter Bike Index
External Circuitry
Some may think that this modification is going a little too far but having spent some time devising an accurate voltmeter, it seems a shame to allow the bikes circuits to make the readings unreliable. (This circuit is worth consideration with any sort of meter movement as it will improve accuracy)
The wiring to the voltmeter as fitted by BMW is not ideal.
On the face of it, provided the meter is connected between the positive and negative terminals of the battery, all should be well and other circuitry should not affect the reading... but this ignores the phenomenon of circuit resistance.
It is conventional and convenient to regard the wires in the machines wiring loom as perfect conductors. This is, in fact, far from the case. All materials have a resistance and there is always, therefore, a voltage drop across the wire. In the case of a new copper wire of adequate size for the current carried, and with clean connectors, this resistance is negligible.
Over time, however, wire becomes corroded and starts to break due to flexing. This results in the resistance of a given length of wire becoming greater. In addition, each contact and connection in the circuit has a small resistance and, as the wiring ages, each one becomes dirty or corrodes and the resistance becomes higher.
If the voltmeter were to be connected direct to the battery, the result of this increased resistance would still be negligible. The voltmeter is, in automotive terms, a very low current device and the voltage drop is directly proportionate to the current drawn.
In the BMW boxer, however, the voltmeter is not connected direct to the battery. It grounds by way of the vehicle main earth point and the positive circuit is shared, in part, by all of the electrical circuits on the machine. This means that a comparatively high current will flow through parts of the wiring shared by the voltmeter.
A high current will also lead to a relatively high voltage drop and this will affect the meter reading. Worse, the amount by which it affects the reading will vary dependant on how much current is being drawn at the relevant time. To put it in simple terms, not only will the meter read low, it will also read even lower if the lights are switched on.
To illustrate:-
Say the actual battery voltage is 14 volts, the circuit resistance is 0.1 Ohms and the current drawn by the ignition system is 6 Amps. Applying Ohms Law (Volts equals Current in Amps multiplied by resistance in Ohms) The voltage drop in the system will be 0.6 Volts (6 x 0.1) and the voltmeter will read 13.4 Volts.
Now switch on the lights, say a further 6 Amps. The battery voltage remains at 14 volts and the circuit resistance at 0.1 Ohms. Now the voltage drop will be 1.2 volts (12 x 0.1) and the meter will read 12.8 volts.
A voltage drop of 1.2 volts will actually make little difference to the brightness of the lights but the implications for the meter readings are quite severe. 14 volts is an ideal voltage for a battery under charge but 13.4 volts would indicate a poor rate of charge and 12.8 volts, if correct, would indicate that the battery wasn't being charged at all. In both of the cases illustrated above, the charge circuit is working perfectly but the voltmeter indicates that there is a charging problem.
The answer has, of course, already been mentioned. If you provide the meter with separate wires running right back to the battery terminals, the meter will always accurately represent the voltage at the battery.
The schematic below shows the circuit I used. It uses a standard automotive relay activated by the original meter wire to switch a direct line to the meter. (In theory, the meter could probably be left permanently connected, the current drain - around 3 milliamps - should not represent a significant drain for an automotive battery - but I preferred to have the display blank when the ignition is off)