The ads above are Google-sponsored, ......clicking on them at every visit helps support this website!  
Clicking on something in an advertisement helps even more!


Troubleshooting the Starting System
Starter doesn't engage or function otherwise.
Starts, or barely, won't stay running properly.
Faulty or otherwise noisy Valeo starter?
Load testing the circuitry AND BATTERY, using the STARTER!

Copyright, 2013, R. Fleischer


NOTE:  It is very important that the reader has already read section 16-A!
To do that NOW, click on:
I also suggest reading


Starter motor rotates the engine OK, but is difficult to fire up the engine, or to keep the engine running:

This is almost always a carburetor problem.  Always check valve clearances and if you have ignition points, check their gap, then check ignition timing,...always do these things FIRST.

Assuming it is a carburetor problem, there are only a few causes; these hints are for Bing CV carbs:

1.  If the engine will hardly fire up, and if it does so only one cylinder seems to be working, but might 'catch' after awhile,
    then the LIKELY cause is a blocked jet in the carburetor FLOAT BOWL.  There is a tiny jet in the bottom of a WELL in one
     corner of the carburetor.  Clean that jet.
2.  Be sure that the idle jet, which screws upwards into the carburetor body,  is not clogged, & that the carburetors are in
     some sort of reasonable adjustment.
3.  Enricheners were wrongly assembled, such as left and right carburetor parts exchanged; or, the more rare instance of
     factory wrongly marked enrichener shafts (lots of photos and information in the Bing CV carburetor articles on this
     website); or, enrichener cover at the carburetor has a sucked-in gasket or gasket failure.
4.   Check enrichener rotating discs to be sure THEIR jets are clean.

Starter motor load testing, including battery condition load and connecting wires testing
An edited & interpreted version of testing, slightly expanded-upon by me, from the BMW Factory Service Manual.

Check that battery terminals are clean and shiny and reattach them.  Check the the ground wire is in good appearance and tightly fastened at the battery terminal (-) and grounded to the transmission case reasonably tightly (do NOT overtighten the hollow bolt!) with the proper washer(s).  Check the battery positive wire to be in good appearance and firmly attached.  The battery must be fully charged.  Disconnect the charger.  Connect a digital voltmeter to the battery terminals themselves (NOT to any other place).  You will be doing TWO tests, within a couple of seconds, and in the second test the voltmeter POSITIVE lead will be moved to the starter motor itself (terminal 30).
Put the transmission into 3rd gear (4 speed transmission) or 4th gear (5 speed transmission).  Press the rear brake strongly, and hold it.
Turn ignition on and press the start button, holding it for about TWO seconds.  The voltage should not drop under 8 volts; the engine of course will not start.  Immediately move the positive lead of the voltmeter to the starter terminal.  The voltage should be not over 3/4 volt less.    If less, there is probably a cable/connection problem.

This test is BEST done with TWO digital meters at the same time, one at starter (to frame ground), and other to battery, as noted, that way only ONE test is needed, and you can compare the meters at the same time with a quick glance, and thus the test is MORE accurate.


Starting motor system and problems:

     The very basic starting system is the same on ALL Airheads; a starter motor, a starter motor solenoid (which is a HUGE
     switch), a smallish starter relay, the starter switch on the handlebars, a battery...and the necessary electrical wiring. 
     There are some variations in peripherals, including a special circuit in the starter relay of the /5 series; and, some
     variations in the circuitry which includes the clutch and neutral switches.   At its core, the Airhead starting system is very
     similar electrically and mechanically with what has been in use on cars and trucks for many decades. 

The starter motor is a powerful electrically operated motor.   To obtain the type of mechanical power needed from the
     starter motor, upwards of 1/2 to 1 horsepower may be needed if the engine is quite cold, has thick oil in it, or the engine
     has high compression ratio.  Thus, the starter motor requires a quite high electrical current.  Over 100 amperes is not

There needs to be a means of switching on and off the very large amounts of current to the starting motor.   The heavy
      duty starter solenoid switch, physically located on the starter, is fed by a large diameter copper-conductors-cored cable,
      directly from the battery.  The starter solenoid itself is activated by a modest current delivered through the pushbutton
      switch, itself caused by contacts closure on the small starter relay.  The 12 volt power travels from the starter relay via a
      thin BLACK wire that goes to the SPADE terminal on the starter solenoid unit. The starter relay is, therefore, a small relay
      that sends electricity to another much larger relay, called the solenoid (or, solenoid relay).   

The sequence for starting is:  Ignition key ON, which supplies power to the START switch on the handlebars.  Pressing
      the handlebar switch button supplies power (completes the circuit) to the coil of the starter relay located under the fuel
      tank.  The starter relay contacts now close, sending a modest amount of current to the solenoid switch coils, the
      solenoid switch unit is located on the starter motor. This current causes a substantial magnetic field in the solenoid
      PULL-IN winding coil.   This magnetic field now moves a large magnetic steel slug.  One end of the steel slug has
      insulated contacts or a contacting strip which is pushed towards large copper contacts and the pressure causes those
      contacts to be connected.  Once connected, the large battery cable at the solenoid can now pass a very large flow of
      electricity into the starter motor. At the same time, the Pull-In winding inside the solenoid switch (there are TWO
      interconnected coils) disconnects and there is a secondary STAY-IN coil that draws less current, giving a bit more available
      current to the starter motor (a larger improvement than the headlight relay on some Airheads that turns off the headlight
      during starting), that keeps the solenoid piece pulled-in and the massive switch in contacting position.   The starter solenoid
      is nothing more than a mechanically & electrically operated high power electrical switch.   

     As the starter begins to rotate, a mechanical device called, generically, a 'Bendix' or Bendix drive, begins spinning.  It is
     located on the rear of the starter motor, on the starter motor shaft.   This device, via spinning forces and guide grooves,
     and with massive help from the plunger linkage to the Bendix drive, then moves rearward with quite considerable force,
     and as it moves rearwards it engages its teeth into the engine flywheel (called Clutch Carrier on later models) teeth, that
     causes the starter motor shaft to lock to the flywheel, causing the engine to rotate.  The teeth of the Bendix drive, and on the
     flywheel gear, are shape-designed to enable quick meshing at high levels of force.   It is entirely possible to design a Bendix
     unit that does not use a mechanical coupling to move the Bendix into the ring gear on the flywheel, and this has been
     popularly done on many vehicles.

     NOTE!   The actual operation of the solenoid, electrically, is a bit more complex, and I ran across nice videos that do
                   have a schematic and discussion, and here is the links. Be sure to watch the entire video, first is #2 of
                   a series of 3 to analyze a starter that won't stay engaged:
                   Here is #3:
You can probably access all of them from any of them, not sure, but it is what I had to do to get #3.  You can
                     probably just scan down the right side list for Moto Phoenix articles.
                     What is particularly good about #3 is that it clearly shows how to take the Bosch starter completely apart.


     Later model airheads may have a Valeo starter motor which has permanent magnets instead of field coils like the Bosch
     has.  The Valeo uses a planetary reduction gear inside its nose, which allows the starter to develop higher rpm
     and hence higher power.  

    There is a factory bulletin on the starter relay on 1985 and later bikes (most specifically, 1985-1988), information will be
    found the the Bosch-Valeo article:
    There is also a link in that article to another article on how to do general maintenance on the Valeo starter.


There are at least THREE types of Bosch starters used on the Airheads as original equipment. I say at least three, because there are some very rare instances of other models used on foreign-shipped bikes.  However, to my best knowledge, the number of teeth on the flywheel and starter is always either 8 teeth on the starter with 93 tooth flywheels; OR, 9 teeth on the starter with 111 tooth flywheels.

Bosch starters up through 1974 were 8 tooth, rated 0.5 Kw and 290 A.  The /6 bikes for 1975 and 1976 used an 8 tooth rated 0.6 Kw and 320 A.  The 8 tooth starters are used ONLY with the 93 tooth flywheels; and are for practical purposes, interchangeable.  BMW phased-in the Bosch starter at different times depending on the motorcycle model, so it is possible for some 8 tooth starters to be on somewhat later bikes...SO IT IS A GOOD IDEA TO COUNT THE NUMBER OF TEETH ON YOUR STARTER AND THE NUMBER OF TEETH ON YOUR FLYWHEEL.

For SOME 1976 and all 1977 and later, the starter has to be 9 tooth, for use with the 111 tooth flywheels  (more properly called the clutch carrier from 1981).  The starter is, rated 0.7 Kw and 320 A.

    Late bikes had Valeo starters installed, they are all 9 tooth.  The early ones were troublesome with
    glued-magnets coming loose.   See for fixes or new ones, without the

Starters can be repaired. 

   Starters do not always fit perfectly:
The fitting problem is slight, but can cause broken starter nose-cone's, so DO check.

If the starter motor develops too much play in its bearings, it may tend to bog down, and might even start to rub its armature against the pole pieces.  Worn brushes, grunge in general, worn armature, and poor solenoid contacts, are all other things that greatly reduce starter power output.  Earliest Valeo starters on the Airheads had a habit of the glue letting go on the field magnets, jamming the starter internals.   Valeo replacement starters from such as are modified and don't seem to have these problems, or, are greatly minimized.   If you decide to install a Valeo, in place of a Bosch, be sure the Valeo fits exactly properly.  See:

It is a very good idea to disconnect the battery before removing (or replacing) the aluminum cover over the starter area. The reason for this is the potential (bad pun!) shorting of the cover against the battery terminal on the starter solenoid.  That connection needs proper orientation (to avoid shorts to the case/cover); and has to be TIGHT;....things to check on before you later re-install the cover.    You need only remove all the wires at just the negative terminal of the battery to ensure the power is disconnected.  If you have ONLY the one large wire at the negative terminal, you could disconnect at the speedometer cable bolt instead.   Yes, I do think it is acceptable to modify the battery cable lug, so it will just barely push onto the hollow speedometer cable bolt.  That way, no need to totally remove the hollow and somewhat fragile speedometer cable bolt.     Be SURE the TWO flat washers are on that bolt.  DO NOT OVERTIGHTEN!

If the starter does not work correctly, you need to determine if the problem is in the starter motor, and/or its solenoid; or, someplace else, such as cables or relay.   Just what your bike's symptoms are, may well determine your proper approach to your problem.  If the starter does not work at all, the tests may be a bit different.

It is vital that the battery be in good condition.   You can monitor its terminal voltage, and see what the terminal voltage does during cranking attempts AS THE ENGINE ROTATES.  If the starter is very bad, perhaps drawing HUGE amounts of current, that can drag the battery to under 11 volts.  This is a simple test, and is NOT the same as a 'locked' test, well above in this article, where 8 volts is mentioned.   A formal load test on the battery is done in a well-equipped shop; although Harbor Freight has two versions of decent testers available quite cheaply, and they ARE accurate enough.     Monitoring the battery voltage whilst you try cranking the engine (assuming the starter functions at all) might tell you quite a bit.   If the battery wires are poor, corroded, etc., the voltage at the battery during cranking may be much higher than at the starter.   Monitor the voltages at both places.  Is the battery OK during cranking attempts?   A battery CAN have good terminal voltage at rest, and fall dramatically during a cranking attempt (usually a failed inter-cell connection causes this).   You can also easily test for a bad + cable or bad - cable.

If there is no starter function, no loud starter motor solenoid noise, no tiny under tank click noise from the starter relay; then the starter relay, or its sometimes troublesome socket connections, is suspect.  A meter or test lamp will easily reveal the problem.  Monitor the battery, and also trace the power into and out of that under tank starter relay.  The wires for the input and output power are RED at the relay.  ALL the red wires must ALWAYS be at battery voltage, whether or not the ignition key or any other switch is on or off.  The only exception is if the relay is UNplugged.

If that starter relay does click
when the ignition key is turned on and the starter button pressed,  but there is no major click/clunk from the solenoid, then check the solenoid, be sure it is getting power to the SMALL black wire terminal DURING cranking attempts.  That black wire is the power coming from, you hope, the starter relay.

No cranking?  All seems OK?  Remove (pull off) the black small wire at the solenoid.  Run a temporary jumper wire (any small gauge wire is OK) very temporarily from that male spade terminal on the solenoid, to the battery terminal next to it.  If that causes the starter to operate, where it would not before, then your problem is NOT in the starter area; but more likely the thin black wire or the starter relay; or push button.  If the starter fails to operate, the starter or solenoid is at fault.

Certain years and models of BMW Airheads have various types of interconnections with the neutral switch and the clutch lever switch at the handlebars.  These can introduce complications in your testing.  You may need to find out if power is going through the starter relay.  The easiest way to begin this process is to leave the black small push-on wire DISconnected at the starter.  Key ON, Transmission in Neutral, pull in Clutch lever, emergency shut off switch centered (ON position).  Push starter button:  is there now power at the unfastened wire? If so, the primary problem is NOT in the starter relay.   I like to use an old headlight lamp for this test, saving the ones with one burned-out section, as using this lamp draws about the same current as the solenoid coil does, and is a better test on the starter relay and its contacts.  I connect one terminal of the lamp to the chassis as a ground connection, and the other lamp terminal to the black wire.

Release the clutch lever; does it still have voltage?    If so, the problem not in clutch lever switch.
Transmission to any gear but not in neutral.   Clutch lever pulled back.  Starts?  if so, problem not in transmission switch nor clutch switch.

The /5, ONLY, had a peculiarity; and, the circuitry arrangement needs understanding.   The battery feeds power to the ignition switch, as in all the airheads, but there was NO handlebars-mounted emergency cutoff switch.  From the ignition switch in the /5, the battery power fed a special metal can starter relay that has had various names, including Starter Lockout Relay.  It is located furthest forward on the left of the backbone under the fuel tank.  This is not just a simple relay.  It incorporates a simple transistor circuit; which, when stock (that is, UNmodified), is unfortunately overly-sensitive to temperature (possibly failing at even modestly cool temperatures) and it is also sensitive to even slightly decreased battery charge.    When in the 'problem' occurs, the relay may chatter, or sound like a cricket, and may even power the starter solenoid that way too; making a much louder cricket or rapid clickety-clackety noise.  To an old time mechanic, it sounds just exactly like a quite low battery. 

The transistor circuit's purpose was to eliminate possible starter relay operation (as in YOU trying to start an already running engine!) as soon as the alternator is spinning at engine idle or, actually, somewhat above.   There are low battery and low temperature problems with that starter relay circuit, and a modification is HIGHLY recommended, and there is an article on this website specifically dealing with that problem, and the fix.
The article is:

Other than the added safety anti-start function, the /5 starter relay coil is wired such that the grounding side of its coil connects to the starter pushbutton switch.  It is grounded when the pushbutton is pressed.   If the starter relay on the /5 fails to operate, it can be corroded inside, often fixable as it is in a metal can that can be carefully UNcrimped and things cleaned-up inside.  If the transistor fails the relay will not usually work at all.   The modification, see link just above, prevents problems at lower temperatures and/or slightly weak battery, is highly recommended.    There are NO complications in the /5, as to clutch switches, neutral switches in the starter circuitry, etc. 

***The slash 5 (/5) ONLY can also have another rare problem.   If the alternator rotor opens, or especially if the the voltage regulator open-circuits, it is actually possible for either of these faults to prevent the starter relay from working.

After the /5, things started becoming more & more complicated in the starter circuits.  In some ways, less complicated in not having the transistor anti-start relay!   The only thing that was simplified was the deletion of the /5 starter lockout relay, in favor of a more standard type of relay (but, not absolutely a standard type). SOME of the later starter relays contain one or two be sure to read #19.

Change #1 came in the early /6.   The neutral indicating lamp, which is green (and the GEN lamp which is red) both went through a fuse, before connection to the ignition switch.  The starter relay coil in the 1974 /6 models was wired to the same fuse as the INDICATOR if those lights are not working, chances are the fuse is blown! That fusing did NOT continue after the early /6.    The rest of the starter circuitry was the stone-simple starter relay, and starter assembly.
Beginning with EARLY 1975  /6 models, BMW incorporated an emergency kill switch....added between the ignition switch and the starter relay coil.  Sometimes these kill switches become corroded internally and act strangely intermittent.   

It was in the late 1975  /6 models, and then carried forward into 1977 and beyond (in some ways), that the beginning of complications came about that drive some mechanics or would-be mechanics crazy.  BMW added three things to complicate it all.   The starter button, when pressed, no longer just grounded the starter relay negative side of the coil to the chassis.  Whilst all the 5 speed transmissions, including the early 1975 /6, had a neutral switch, the early models were NOT wired into the starter section, like the later ones were. 
 Instead the path from that starter relay coil could take TWO directions:
        A.   If the neutral switch, located on the transmission, was in the ON position, that grounded
              the coil, but through a DIODE. 
B.   If the bars clutch lever mounted switch was closed by pulling in the clutch, the starter
              relay could be operated from the starter button, no matter if the transmission was in
              neutral OR NOT.  The diode was there to prevent the neutral lamp from indicating every
              time the clutch lever switch was activated.  These diodes have been known to short
              circuit.  If so, the neutral lamp will light up every time the clutch lever is used.  If the diode
              failed OPEN, the neutral switch will not allow starter operation UNLESS the clutch is
              pulled in.  The location of this diode VARIES!     For 1974-1980, except the R45/R65, it is
              UNDERneath the connection board in the headlight shell, where it can't be easily seen.
              For 1979-1980 R65, it is part of the wiring harness, near the Voltage Regulator.   For all
              the others from 1981, it is INSIDE the starter relay!  THUS, the starter relay from 1981 is a
              SPECIAL relay; although an aftermarket one can be substituted, and the diode added.

     Regarding that diode:

       The diode on all Airhead models that use it, must, in some conditions, pass the starter relay coil
             current; and absorb any high-voltage ‘kickback’ from that starter relay coil.

             I recommend a diode rated at 400 volts or higher; and at 3 amperes. 3 ampere diodes have
             considerably more reliability in this usage, than 1, 2, or even 2-1/2 amp diodes, due to the
             internal construction of the diode. Be SURE to install the new diode so that the band-marked
             end (silver stripe) is in the original direction. I have seen these diodes installed wrongly, that is,
             backwards. For the diode when mounted on the underside of headlight bucket wiring board, the
             banded-end (silver stripe) of the diode is connected to terminal LKK.

In 1978, BMW changed things again.  This lasted into 1980.  BMW added 2 diodes...and a switch!  These models had an under-fuel-tank brake master cylinder.  BMW incorporated a low brake fluid level switch, which, when closed, turned on a brake failure RED light.  The switch was wired to that lamp, and via a diode to the starter relay coil.  This modification did NOT interfere with the starter operation, even if the diode failed in the open condition. If, however, that diode shorted (HIGHLY unusual), and at the same time one had a failed (closed) switch, or the level was low, the starter could fire up, mysteriously, all by itself, and NOT release!...unless the ignition or kill switch was used. 

BMW added a diode in the headlight relay.  The location of that relay varies with year and model.  The wiring and operation was such that during the time the starter motor was powered, the headlight relay coil was NOT energized, which turned off the headlight.  Depending on model, whether USA or European, this varied somewhat, and in most models... the instruments and rear running lamp WERE left ON during cranking; by use of the diode inside the headlight relay. Some had various combinations.  What is complicating the mess here, is that, at least for USA models, the diode in the headlight relay was so connected that if it SHORTED, then in some conditions of ignition switch setting, etc., it was possible for the starter relay to fire up, and NOT be releasable, until the battery was disconnected.    Yes, the same sort of thing like the 1978-1980 symptom of the shorted starter relay diode.  Thus the same SORT (nearly) fault could be had via a shorted headlight relay diode or shorted starter relay diode!

Theoretically this problem could only happen in the PARK position of the ignition switch.  The fuse was also now incorporated in the headlight relay section, and also fed the parking light.    Normally, if you used the starter button, even if the fluid level was fine,  the brake failure light would illuminate at every start...this tested that lamp at each startup.  It wasn't necessary to have the fluid be low, then, to turn on that lamp. 

With the 1981 models, no longer was there an under-tank master cylinder, so the operation is slightly simplified.  The wiring is exactly the same, but withOUT the master cylinder switch and, of course,  its diode was eliminated. This lasted through the 1984 models.

There were several more variations and changes until the end of production.  In 1985 and later, BMW put the starter button into the + power feed to the starter relay coil, and the negative side of that starter relay coil went to the same diode and the same neutral switch and clutch switch as before.   By moving the position of the starter button connection, BMW could incorporate the diode into the relay housing.  BMW also added another diode in the headlight relay area; this time in series with its coil.  The starter worked the same, however.    The GS and ST, and 1981-1985 R65 and R45 had slightly different headlight relay contact arrangements; which in other models cut other lights during starting besides the headlight lamp itself;...again, not a starter area to be concerned with.   [for the nerdy:  the power for the instruments and tail light comes from the ignition switch on these].  Earliest R65 models were different in this regards.

Versatile relays that can work fine for most functions in your motorcycle, such as switching lamps, running horns, starting, etc., is the Bosch (now Tyco) 330-073, rated at 30/40 amperes and 12 volts, SPDT, 5 pin, with tab for screw (tab area can be removed); or the Blazer DF005 or DF005W which also has a tab/screw mounting.  Be sure you understand how to use relays, and understand what is what on relays with diodes inside them, etc.


11/29/2006:  initial first draft and release
11/29/2006:  correct typo editing error in h.
11/16/2007:  re-number the items, combine certain sections.   Vastly improve clarity of descriptions.
10/03/2009:  clarify a few things and clean up some messiness.
11/22/2009:  go over entire article, clarify certain things
10/14/2012:  Add QR code, add language button, update Google Ad-Sense code
03/30/2013:  Clarify and add more information to #19
09/13/2014:  Clean up article considerably.  NO technical changes of any special note.
04/04/2015:  Improve description of the solenoid two coils operation, and copy the section to include it
                      in the boschvaleostarter article.
05/30/2015:  Add section on the enricheners, and clean up and revise article, fix links, etc.
07/21/2015:  Fix section(s) on starter models/teeth/years/numbers.
08/12/2015:  Add video link, some minor article cleanup, additional link and comment on 08/15/2015
08/29/2015:  Add section on load testing and cable testing

© copyright, 2013, R. Fleischer

Return to Technical Articles LIST Page

Return to HomePage