BMW motorcycles with flaky turn signals

Turn signal and grounding problems on BMW motorcycles

by Duane Ausherman

I wrote this piece for the /5 group to address a problem that one of the riders had with his turn signals.  They were acting very strangely.  This was not intended to be a “recipe” type of advice, but an explanation into why some of the anomalies occurred.

The description of the inconsistent turn signal flashing is typical of a grounding problem.  Grounding is simply a term for the electrical system on one “side” of a circuit.  It is made up of the same stuff as the “other” side of a circuit, namely resistance.  We tend to think of resistance as a fixed thing.  That is so often not the case and it is very easy to get confused.

I will use simple terminology, not the stuff that our engineers will like.  I have little choice but to use my “farmer” type explanations.  I hope that this helps my intended readers, but it may offend the technical types.

We will slowly build a simple turn signal circuit.  Lets start with a simple bulb as an example.  The bulb is a 12 v and rated at 12 watts.  The power law tells us that 12 watts at 12 volts will have a current flow of 1 ampere, or amp.  Volts (12) times amps (1) is the power in watts.  12 X 1=12.  If in fact we discovered that the circuit actually had 2 amps flowing then we would be consuming 24 watts of power.

The resistance that the bulb would actually have would be defined by Ohms law.  The current equals the voltage divided by the resistance, or I = E/R.  In this case the resistance would be one Ohm.

Here is where it starts to get interesting.  If you put your multimeter on the bulb to measure the resistance, you would be very surprised.  It wouldn’t measure anything close to one Ohm.  How can that be?

We incorrectly assumed that the resistance is linear.  It often isn’t.  The multimeter doesn’t use 12 volts to test resistance.  It uses much less.  At a lower voltage, the resistance is quite different.

For our next example we are going to use a more accurate type of multimeter called an oscilloscope.  This is nothing more than a voltmeter that uses a TV screen to show us what happens to the voltage and current each small part of a second, often dividing it much closer than a millionth of a second.

We would like to think that since electricity is fast (moving at the speed of light) that as soon as we throw the switch on our simple circuit with only one bulb in it, the current would instantly come up to one amp.  The TV screen would show zero volts and then rise up to 12 volts.  We would have a “square” jump at the bottom and again at the top of the 12 volts.

We would quickly find that something else happens.  It would start up quickly (not quite square, but close) and when arriving at the top it would take forever to finally get to 12 volts.  It is very rounded off at the top.  It takes 1/4 of a second to get there.  Observation of the bulb also confirms that the bulb doesn’t get to full brightness instantly, but there are more factors to explain that one.

The bulb actually starts off at a very low resistance and as current goes through the “business part” of the bulb, the resistance actually climbs up a lot.  The resistance is affected by the heat in the wire.  The resistance isn’t linear with respect to heat (voltage X current).

Now we add in a “flasher” to make the bulb turn on and off.  It is nothing more than a special type of switch.  The flasher is (the older original type) a metallic strip of metal that heats up and turns on or off a switch.  The higher the current, the faster it heats up and then turns off.  It cools down and makes contact again.  The “cycle” repeats itself.

So now we have a bulb that is nonlinear in operation.  We have attached it to a switch that we have wanted to be  nonlinear.  We now have a more complicated situation.  We then add in another bulb for the other end of the motorcycle.  The total current drawn by the two bulbs is what now controls the flasher.  A flasher that is correct for one amount of current, will be quite different for another amount.

Our next addition to the circuit is two more bulbs for the other side of the motorcycle to make up a “right and left” indicator for traffic to see.  This requires a “turnsignal” switch and associated wiring.  Our flasher will “see” either the right side circuit or the left side circuit, but not both at the same time.  Next we add in another bulb for the driver to see as an indicator.  We are starting to have a complicated system.  We haven’t even started talking about a 4 way flasher or cancel system.

To reduce the production costs, we can sort of “cheat” a bit.  We have to run copper (read expensive) wires “out” to the 4 turn signals.  We don’t need to run a return circuit back if we just “ground” them.  Just let the metal of the motorcycle carry the current back.  This is where we start getting into trouble.

At each connection to ground (metal of the frame) we have two metals in contact that aren’t of the same material.  This is an invitation for trouble.  With moisture and odd chemicals getting into the contact, we can end up with some totally unpredictable connections.  In extreme cases it can be a very simple radio frequency transmitter.  It is more common for the connection to be nonlinear.  That means that while the resistance of the connection should be zero, it will actually be some real amount and not fixed.  It will change, sort of like the bulb, but not in a predictable way.  It will seem to have a mind of its own.

Think of all of the places where we have connections on our 30 year old machines.  Each is potentially a problem, not just the return side.  That is why a special grease can do wonders for keeping out the moisture and oxygen that start the errant chemical processes.

With so many connections changing resistances so easily, it is quite challenging for a non electrical person to diagnose them by using a common multimeter.

The most common solution to turn signal defects is to run a separate ground wire from the turn signal housing back to a ground.