This is a mirror of http://pweb.jps.net/~snowbum/Ignition.htm. It is provided here by cdegroot.com to increase availability of the original site. This mirror is updated daily so should be up-to-date at all times.

---

 Ignition module; ignition canister; coils; also information on Dyna & Boyer.  Sources & information on Modules & Hall elements; canister overhaul. Discussion, testing methods.  The "new coils and electronic module" changes.
Some of this Hall sensor information applies to the K and Oilhead bikes!!

©
Ignition.htm 
article #30

NOTE:   The three pin connector that connects the wire from the canister to the bike wiring has a thin wire bale clip around it, that MUST be removed before you try to unplug this connector.  Pry an end, or both, with a tiny tool, and then carefully remove that bale clip.  THEN unplug the connectors. Don't lose the bale wire, or, simply replace it until you are ready to reinstall the canister.

NOTE:   This article is on the ignition canister used with the Electronic Ignition, which means 1981+.   However, the canister was used with POINTS in 1979 and 1980.   Some information in this article applies.  

The 1979 and 1980 canisters with points did NOT have a felt pad to help keep the points cam lubricated, hence, if the lubrication on the cam dries out, the rubbing block and the cam, together, can make a very loud chirping noise.  In that case, remove the outer lid (2 screws) and lubricate the cam and sides of rubbing block....all sparingly....with a high temperature grease.   Put one drop of a decent oil on the outrigger bearing.  NOTE that the points can NOT be adjusted withOUT the outrigger bearing in place.

Before the 1979 introduction of the canister, the Airheads all had points ignitions with felts, and the felt AND cam required lubrication (both on the engine cam's shaft inside the ATU, and the ignition cam outer surface).  Since the automatic advance unit was at the cam tip (not buried as in the canisters), the automatic advance unit required cleaning and lubrication now and then could be done rather easily; with two types of Bosch greases used, one for the rubbing block, one for the area between automatic advance and cam....and the shaft.

UNDER NO CIRCUMSTANCES IS MY LISTING OF THESE SOURCES OF INFORMATION AND REPAIRS THROUGHOUT THIS ENTIRE ARTICLE TO BE CONSTRUED AS MY APPROVAL OF THEM, OR THEIR METHODS!!  They are listed for your information ONLY!  I specifically deny any responsibility for your use of these sources.    However, I do believe they were trustworthy at the time I put them here.

For further information, see Haynes, Clymer's, or other manuals, just be cautious for errors.

For canister mechanical and/or electrical overhaul, there are a number of sources for information, including, but not limited to the following, shown in various places throughout this entire article.

NOTE!:   One of the common problems, perhaps THE most common problem, is a sticky automatic advance unit, that results in a high idle speed, typically after a FULL warmup of the entire engine.  The PROPER fix is to disassemble the canister and very carefully clean and re-lubricate it.  This project is not easy for the novice.   Sometimes removing the oval side plate and squirting in a cleaner and then a fine oil (NOT WD40, it will harden) will help, but may or may not hold up over time.   When disassembling the canister, some of the parts can be burnished (sort of a more complete polishing) which DOES help.  Other common problems include failure to clean and renew the heat sink paste under the Module at reasonable intervals, it is located under the fuel tank.  If the heat sink paste is old and dried out, the Module will tend to overheat and cause ignition problems after the bike has been on the road awhile.   Another problem seen now and then is a Hall device failure.  Usually a complete failure, they have also been known to be intermittent.  USUALLY the Hall problems occur after a very thorough engine warmup.  Ignition gets intermittent, or dies.     The devices are replaceable, but BMW only sells the complete canister, VERY pricey indeed....but with this article YOU can repair them!

>>>>Some folks who are overhauling canisters: <<<<<

Motorrad Elektrik  (256) 442-8886     www.motoelekt.com

OAK Okleshen  AskOak@aol.com   Top quality work from our own Airhead Guru.

Dwight Small, 4800 Rim Rock Way, Fort Collins, CO  80526-4617   (970) 223-1541
     Rbike@excite.com  He advertises that he uses OEM sensors, and last known price was $110, exchange.

Sources of information, parts, ETC:

R-Bike Electronic Canisters, by Pete Serrino (serrinop@ehsct7.envmed.rochester.edu), see the following URL:    http://www.roadkill.com/~davet/moto/timeCanister.html

Fitting a second Hall Effect sensor inside a BMW Airhead Boxer's "Ignition Impulse Sending Unit" 1981 onwards, by Frank Warner.  Lots of canister information and a sketch.      Note: that article also lists sources for parts, including junked vehicles, somewhat Australian oriented, but useful. http://www.geocities.com/fwarner_au/mc_things/bmwr_ign_sensor.html

Troubleshooting Ignition Trigger, as posted to the BMW Airhead Boxer mailing LIST, on 19 April, 2001, by John W. Snider.

Ignition Trigger Repair Bulletin, by David A. Braun.   http://www.deathstar.org/~flash/ig_trig.html

Various postings over some time on the Airheads LIST; IBMWR; etc.

Hall devices:  Honeywell (MicroSwitch) part #2AV54. This is available from Newark Electronics as part 96F1986    Newark price: $13.57 each and their shipping was expensive, so check that.
Another source for the 2AV54 Hall Effect Device: http://www.onlinecomponents.com/  At the time of writing this, the cost was $11.77 each.  
Reported other sources:   Siemens HKZ101....available from Jacar in Australia as ZD-1900.

Note that the Honeywell device, in case you try to find it on the Honeywell website, is not all that easy, sometimes, to find.  That is usually because its official name on that site is a "Hall Effect Vane Position Sensor".  You can TRY using this:
http://www.honeywell.com/sensing/

Red, power; black, ground; green, signal

OLD units had the sensor plates screwed together and the rivets were hollow; later plates were a push fit, solid rivets.

  NOTE  FOR THE NERDY:   This is a schematic of the Hall Effect device.  Note that the output is an NPN transistor with open collector. 

VAG (VW and Audi) ignition module part nr. 191 905 351b  is a drop-in replacement for the airhead modules.
 Cars that you may find modules on in wrecking yards, or can get modules at autoparts stores:
     VW Golf and Passat 1979-1989
     VW T2 Transporter and Jetta 1979-1992
     VW Scirocco 1979-1995
     VW Corrado 1979-1995
     Audi 100  1980-1993
     Audi  80   1979-1993
     Audi Coupe  1981-1994

UNconfirmed information is that the Hall units may be found in 1979-85 VW; 1978-1983 Audi; 1979-1982 Porsche; and that Bosch 1237011052 may be the number.

Canister Ignition points:
Ignition points as used in canisters, that is, 1979 and 1980 models; unconfirmed data:  Bosch GB534...as on Mitsubishi Colt (69-80 6 cyl), some VW (Bus to 1972 for instance); Volvo, some Toyota, Ford Cortina TD and TF, 6 cyl Falcom XY and XD, ....etc.
NOTE:   Porsche, and some others, used Bosch GB752, which was identical, but with slightly higher spring tension, so somewhat less tendency to have floating problems at very high rpm; not important to most Airhead owners.

MORE NOTES:

The canister bikes automatic advance had 120° dwell angle for 1979-1980 (points models), and from 1981 (electronics models) the dwell was 104°.  

The original troublesome gray colored twin tower single coil, which tended to crack and fail intermittently, usually eventually totally failing, but initially usually failing when damp, was the Bosch 0221500200 and was replaced by Bosch 02210500203, BMW 12-13-1-244-426.  The troublesome coils were used in the 1980's, but NOT on all bikes until much later, and there were some on bikes such as the R45, R65, R80GS, R80ST.  DO see later and also above, herein, about the MODULE changes and how they apply, or do not apply, to the updated later coil.  The old coil was BMW 12-13-1-243-910.  The updated coil was black bodied.   In the typical scenario of the beginning failure,  Drying and epoxying the coil cracks could be done as a very temporary measure and may not work at all.    
 However, other substitutes will work:
1.  An Oilhead coil.
2.  TWO each 6 volt Bosch coils as used on SOME electronic module models in the early 1980's.   You MUST use the coils that have a LIGHTNING BOLT on the side, they have the correct electrical characteristics.  The part number is 12-13-1-244-142.  Mount one coil where the original twin-output coil was.  Use that coil for the LEFT spark plug.  Mount the other coil towards the rear, under the relay bracketry, and use it for the RIGHT spark plug.   Interconnect terminals 15 and 1 between the two coils.  The remaining terminal 15 goes to the original green wire; the remaining terminal 1 goes to the remaining black wire.   It is a good idea for the coils to be solidly mounted such that a good electrical contact is made to the metal body.  These Lightning bolt coils are generally HIGHLY reliable, so long as one does not open circuit the spark plug outputs by foolishly lifting the spark plug caps when the ignition is on (this caution applies to all coil ignitions).
3.  BMW updated the module with some internal timing changes inside that 'black box'.  Don't use the new black coils with a module with white or pink lettering.
4.  

***Additional information will be found at ignitionsingleplug.htm  see #7 for instance.

Suppressors (Spark Plug Caps):  These were metal from 1970-1976, 1000 or 1200 ohms nominally, depending on which piece of literature you are looking at, and, in any event, the tolerance is  officially +-20% (if they are 800 to 1500 ohms, I would be satisfied), and from 1977 were 5000 ohms nominally.  ALL points bikes, can use 1000 ohm caps, the NGK are preferable to the metal types.   The electronic ignition bikes (1981+) MUST have 5000 ohm caps, using the BMW cable/caps...OR...NGK 5000 ohm caps are also OK.  NGK numbers for 1000 ohm caps are LB010F; for 5000 are LB05F.  They make 10,000 ohm caps too, so be SURE to ACTUALLY measure them, you MIGHT just confuse the part numbers.   Some R65 had metal caps molded to the cable, and were 5000 ohms.    

There is another article on this website with extensive information on spark plugs AND all the NGK caps information, etc.
sparkplugs.htm-32

Dyna:    Technical information on most models is available at:  http://www.dynaonline.com/english/instruct/index.htm
However, you MAY find SOME of that information confusing.

For the 70-80 points Airheads, Dyna had an Ignition Booster....and this could be used with SOME dual plugging coils.

Many years ago, the original Dyna D35-2 system was NOT a wasted-spark system.  They used DC1-1 green 3 ohm coils on the dual plugged conversions, and the sensor plate had THREE wires.  Each cylinder could be set separately with a timing light. Since this was NOT a wasted spark setup, each coil fed ONE cylinder (that is, each coil fed the top AND bottom spark plug on ONE particular cylinder).  Dyna has not offered this D35-2 in many years.   Now, what is available is the D35-1, wasted spark model.    The D35-1 has TWO wires at the sensor plate; and uses the 1.5 ohm coils...which are brown, or orangey looking, and are series connected.   Timing is set with an ohmmeter or test light, as you rotate the engine by hand, so to watch the magnet location, to know what side sensor you are adjusting.  

The popular Dyna III model D35-1 was for the 70-78 Bosch coils, or, Dyna coils....and the wiring is; to the Dyna box:  white, left cylinder; red, right cylinder; Black is ground. The other red goes to +12 at ignition switch and to the left coil +.  The coils are interconnected by a jumper, same as on original BMW, same numbers at coil primaries, if using those coils.  The right coil (-)is the one with the brown wire from the Dyna box.  
Unconfirmed:  Dyna III uses UGS3040T Hall devices
Dyna coils are colored.

blue coil is 0.7 ohms;...used on 92-95 airheads; secondary is 11.5K
reddish orangey or brown coil is 1.5 ohms, dual tower for dual plugging on the D35-1
gray coil is 2.2 ohms and 14K secondary

green coil is 3.0 ohms, for dual plugging on the D35-1.  

If using the Dyna TWO wire sensor setup, the two coils will be in series connection for the primary windings.  If the Dyna sensor setup is the three wire type, the coils are not connected that way, but are independent.  If either of the red or white hall effect wires are touched to ground, you should get an ignition spark output.

black coil is 5.0 ohms, 17K? 14K/
brown coil is 1.5 ohms, 14K secondary, for 81-92 stock airheads.

Watch out for improper sensors insulation, that is, broken wires at the sensor plate where they exit the epoxy, and for coils shorting a primary to ground when hot.  Use NGK 1000 ohm spark plug caps.

The BMW Ignition modules (the under-the-fuel-tank part):::
The information in the following paragraphs is believed accurate, but is not totally confirmed YET by my personal testing.   

The latest modules from BMW are not exactly the same as the originals, nor those substitutes shown in this article.  BMW had special modules for factory kickstarter equipped bikes, and also for the later bikes with the updated coils.  It appears that the very latest updated modules will handle higher current, and SOMEHOW detect synchronization problems between cylinders, and thereby SOMEHOW eliminate a prior problem with pulses cutting out now and then.    I have been told that these new modules do not have ANY primary coil current flowing until the canister unit sends a signal, as the original could, as key is turned on, have one spark.     Frankly, so far, I see no big indication that any of this is a big deal for earlier Airhead owners. 

There is a bulletin from 1993 on early coils and modules in combination having misfiring.  Here is information from OAK, put here on 05/11/2005 from the posting he did to the AIRLIST (Airheads mailing LIST), and my chit chat on it from a long time ago follows that.  I have slightly edited Oak's comments for grammar, syntax, and ease of reading.  Any comments of MINE are in RED!

Subject: Re: Ignition Control Units (ICU) differences-mystery explained
Date: Sat, 6 Dec 2003 04:28:01 -0800
From: AskOak@aol.com
To: BMW Airhead Boxer mailing list <airheads@micapeak.com>

The following BMW bulletin 12 018 93 (2572)  was released on the list recently regarding the befuddlement of the ICU changes, the unusual wording, and if and how the changes cured any prior problems with the earlier ICU's. OK-read the following bulletin one more time, followed by the explanation.  
January, 1993:
Engine Difficult to Start, Temporarily Runs On Only One Cylinder -All Boxer Engines.
Problem: Under certain conditions (low temperature, poorly tuned or synchronized engine), initially only one of the engines' cylinders fire.
Remedy: 1) The ignition coils have been modified
             12 13 1 243 910 - old coil
             12 13 1 244 426 - new coil
              2) The timing control in the ignition control unit has been  modified so that no ignition sparks are suppressed even when the engine runs out of synchronization.   (confusing, EH!)

Control Unit identification: The part number of the modified control unit is the same as previous units, however, it can be recognized by the turquoise lettering. The modified control units can be used on all boxer engines with electronic ignition (from 1981 on).  However, control units with white or pink lettering should only be used with ignition coil 12 13 1 243 910 (all boxer engines before 1991), i.e., do not use the new coil with the white or pink control unit.  The new control unit, however, can be used with the old coil.

The bulletin also contain this wording:  "Models with Kickstarters:   The modified ignition control units (turquoise color) are only available with a switching off duration of 1.2 seconds. ie., to protect the ignition coils, the primary current is interrupted after 1.2 seconds if engine speed information is not received.   This has no effect on starting with the electric starter motor.  When kick starting, the engine should be turned over and the starting procedure completed within 1.2 seconds.  Engine speed information can be simulated by switching the ignition off and on again with the key or by using the emergency "off" switch if the starting procedure cannot be completed within 1.2 seconds" (you all, of course, understand what that means...NOT ME, from THOSE words!).

(Oak Explanation): The earliest ICU's used in production had TWO timing arrangements. When the key and kill switch are positioned ON, the spark coil current appears even before the engine is cranked. There was a 1 second timed unit used on electric start models, and  5 second timed unit used on kickstart models (to give the dude some time to get his kickstarting business in order...) .  If the engine was not cranked in either the 1 second or 5 second interval of time whatever applies, then the timer in the ICU would cause the unit to disarm and the coil current would cease. This was for the protection of the ICU and coil from overheating, as continuous current given enough time will exceed the permitted heatup limits of the components and potentially cause damage.  When the engine is running of course, the coil and ICU currents are switched and not on full time.

    There was another design limitation in the ICU.  The timing left to right tolerance was within fixed limits of each other.  If for whatever the reason (out of synchronization) the engine left to right speeds were not uniform, both the early units would or could miss a current pulse to the coil for one of the cylinders.  The new amended cure-all unit superceded both the early units as a replacement part. The internal design was changed so that the coil current would NEVER occur UNTIL the engine is cranked over. This prevents any possible overheating of the ICU and or coil.  The hang on time was extended and the problem with out of synchronization insofar as spark production was eliminated in the sense that anytime the trigger unit sent a switching signal to the ICU, the ICU would fire current to the coil, no matter what. This change eliminated the possible misfiring or lack of firing.

Also, the newer ICU has a slightly higher current carrying capability for the updated higher oomph coil. Hence the warning in the bulletin not to use the older units with the newer coil. The bulletin indicates the part number for the ICU is the same as the old one, but that is not so anymore. The part numbers have changed for the updated ICU and there is more than one version. See your dealer for the correct part number for your model. It will be 12-14-2-325-284 or 12-14-2-325-550, one or the other. The changes between the two updates appear to be strictly physical.

(Oak continued...) All in all, the newer ICU solved all the problems, with clear changes in design. You gets what you pays for. If you want to go out and buy older units that are obsolete, be my guest but you might end up with the same old problems haunting you, in the end when all over, at twice the cost.  Save yourself the grief and get your fresh updated parts that BMW attempted put forth to solve a certain problem. We sometimes criticize them for not fixing a problem then when they do, we turn around and buy old obsolete junk parts because it is cheaper.  Huh-- will someone please explain that logic to me??..
Without mentioning any names, there are non-BMW factory authorized distributors around who are selling supposed BMW parts when in fact they may not always be the genuine factory parts. (from Snowbum:  The modules from Stan at Rocky point SEEM to work OK or adequately).

  This is how Snowbum has it:

         The Ignition Modules (under the tank modules) were initially made in two versions.  On these early modules, a spark could occur when the key was first turned on.  One version was for the kickstart bikes, one version was for the no-kickstart bikes.  The original modules had a timing feature that cut off the current flow AFTER about 5 seconds for the kickstarter equipped bikes, and about 1 second or so for the NO kickstarter equipped bikes.    Later this was changed to about 1.5 seconds (UNclear to me, but perhaps this 1.5 was for both...or for the kickstarter equipped bikes...not sure).   The kickstart modules timings were originally somewhat longer to allow time to kickstart the bike.  Which is a bit strange, since engine rotation causes the modules to be triggered anyway.    The reasoning behind the shorter cutoff for later models is to lower heating of the coils, which supposedly was THE, or only one of the, causes for the original twin-tower gray coil to fail by cracking.    Apparently, some folks would turn on the key, and continue a conversation, without starting the motor.   To me this meant that perhaps some very early modules never had a timer at all, but I have never seen that...which does NOT mean it could not be true.    Pure conjecture on my part here.   The coil redesign for reliability (no cracking, etc.) did not result in any change in the electrical characteristics of the coil.    

There was also some sort of logic change that I knew nothing about.    ***NOTE that the PART NUMBER for the module was NOT changed, and a paint code used to identify the changed modules.  One is not supposed to use the replacement coil with a white or pink coded module.  But, you CAN, just don't walk away with the key on!   The updated module is paint coded turquoise.   The two part numbers for the updated modules are with turquoise lettering:
12-14-2-325-284
12-14-2-325-550

There is an interesting way to find out the 'time' for cutout, not that it means much to you, actually.    Watch the fairing voltmeter....or any voltmeter attached to the system.  You can also get fancy and watch the proper terminal at the module.  From key-on, the voltmeter drops a bit, and may continue to slowly drop....note the time from the initial small drop TO where the voltmeter gives a wee jump UP, ...it is quite noticeable. 

Some have used modules from various cars. OR, from Stan at Rocky Point Cycle.   I fail to see how this is a big problem, again, don't leave the key on without starting the engine.

***************************************************************************************

---

Some of the information and photos below, was conveniently copied from Thunderchild's website, for whom I hereby give credit.  It should be noted that I have personally done considerable editing and modifications to their information herein, and also added my own notes, and mine are NOT always clearly identified.

How to Test the Ignition Module

NOTE:  Ignition modules must have a reasonably fresh thin layer coating of electronic heat sink paste between the module and the mounting plate.  Failure to clean and reapply such paste every 2 years or so will cause the module to overheat, interrupt the ignition, and eventually fail permanently.
 

The ignition system of a 1981+ BMW motorcycle equipped with electronic ignition consists of three major parts: the coil (or coils), the ignition trigger and automatic advance unit (in a canister) and the ignition control unit (often called the ignition amplifier or module). The coil(s) are usually easy to test, but verifying whether the trigger or the control unit has failed can be a bit more tricky.  A VERY RARE failure is a coil that fails to work properly when it gets hot.  A not so rare failure is the early ONE coil models, in which the coil is gray-bodied.  Those tend to crack, often visibly, and fail, most often when exposed to high humidity or moisture. Referring to the wiring diagram, note that three wires go from the control unit to the ignition module. The Canister Hall device has a + input from Module pin 3; and the trigger signal is developed in the Hall device and sent to the module pins 5 and 6.   Each time the trigger wire, that is the center terminal of the three-terminal cable of the canister, goes to ground, a spark should be generated from the high voltage produced by the coil.    Testing the control unit therefore should be as simple as powering the unit up and grounding the trigger wire. The procedure is as follows (be sure to read ALL of this!!!): 1) Unplug the connector that connects the harness from the trigger unit to the harness from the control unit. Be careful as there is a small diameter round wire spring used as a holding clip, holding the connector together, and that 'bale wire' MUST be removed first.  For access to this plug, simply disconnect the battery FIRST (to protect the diode board from inadvertently shorting it during cover removal), then remove the outer cover over the front of the engine.  The plug is roughly centrally located.  2) Remove one plug from the engine, re-attach the spark plug wire and cap, and fasten the spark plug on the cylinder such that the plug metal body is grounded and you can see the spark gap.  You MUST be sure that the plug is very well grounded...do NOT just lay it on the cylinder,...rather... clamp or otherwise hold it there!  You can otherwise injure the canister Hall device or the module. 3) Stick a pin or fine wire into the center connector of the 3 conductor white connector coming from engine harness (NOT the canister harness).   This is the terminal of the Ignition Trigger that is shown in the diagram as 0....a brown wire.

4) Re-connect the battery.  With the ignition switched on, and the bars KILL switch ON,  momentarily ground the pin/wire you just added coming from the center connector. Each time you do this you should get a spark at the spark plug. This will indicate that the control unit (module under the tank)  is operational. ***5):   An even simpler method that can be tried first, but is not 100% foolproof, is to just turn on the ignition switch, have the Kill switch in the middle ON position.  Turn the kill switch OFF, then back ON, then OFF.  Every time you turn it OFF from the ON position, you should get a spark.   This tests, as does the previous test, only the coils and module, NOT the Hall element in the canister. Do remember that it is not unheard of for the amplifier (or any other piece of electronics for that matter), to operate perfectly when cold and fail when warm/hot.    NOTE:  corresponding Ignition Control Unit   numbers to Bosch motorcycle standard wiring numbers are:  Ignition Module                  Wiring 1                                           1, coil 2                                          31, earth (ground) 3                                              , Ignition trigger 4                                          15, Ignition, light switch 5                                            + trigger 6                                            0 trigger 7                                             , tachometer output, if used Information on the pin-outs, etc., are in Service Information Bulletins 12-007-83 (2075) of May 1983, and in SI 12-006-82 (2061). In the diagram below, the Ignition Control Unit (Module) pin numbers are NOT in the actual order they are on the part.  ON THE PART ITSELF, with the unit laying flat in front of you, pins towards you, the pins are numbered from RIGHT to left, first on right is pin #1, last on left is pin #7.  DO NOT!! use an OHMMETER function on the module pins. The above is copied from a full wiring diagram. Although not drawn as such in the diagram, the trigger wire (to pin 6 of the unit) is the center connector.   The ignition control unit pin 6 has a brown wire; pin 3 has a blue wire, pin 5 has a green-yellow wire; pin 4 has a green-glue wire, pin 1 has a black wire, pin 2 has a heavier brown wire. The ignition trigger plug green/yellow (pin 5) may be violet; the blue may be violet/white; and the black may be violet/black.   Inside the canister unit, the sensor has a red wire which connects to the violet or green/yellow; a green wire which connects to the violet/white or blue; and a black wire which connects to the violet/black or black wire.....in the plug.

---

Testing the Ignition Trigger in the Canister

>>>Do NOT use the ohmmeter functions of your test meter to test the Hall canister device.<<<

NOTE:   The three pin connector that connects the wire from the canister to the bike wiring has a thin wire bale clip around it, that MUST be removed before you try to unplug this connector.  Pry an end, or both, with a tiny tool, and then carefully remove that bale clip.  THEN unplug the connectors. Don't lose the bale wire, or, simply replace it until you are ready to reinstall the canister.

 

This takes a bit more equipment and knowledge.

The ignition canister contains a Hall Effect (magnetically sensitive) sensor.  A rotating metal part provides the magnetic source.  Rotation causes a voltage on the output pin to vary thereby triggering the input pin of the ignition control unit. Note the diagram showing how to wire the connector for testing. The unit is powered from a 9 to 12 volt power supply (battery), and the output from the sensor (center pin) is pulled 'high' (that means positive)  through a 10,000 ohm resistor. As the shaft on the canister is rotated, voltage from the center output pin will fluctuate between a few millivolts and V+ (9 to12 volts). If this makes no sense to you,  then it might be wise not to invite trouble. And remember that this will not show up intermittent or heat related faults.

Overhauling the Ignition canisters:   THIS SECTION IS IN PROCESS, IT IS NOT COMPLETED NOR CHECKED FOR ACCURACY.  Furthermore, things may be out of order.

NOTE!!!  It appears that there is more than one style of plates/sensor setups.  The mounting plate may have a second plate affixed, and then you can't drill one rivet and use screws and nuts.   One rivet can be ground down from the top plate.   The new sensor comes with the mounting rivets imbedded in the plastic housing.  You won't likely find a way to flatten the blind rivet BETWEEN the plates.  You can remove those rivets, and use the holes for your own rivets.   Make sure the sensor is flat and flush with the plate...it is a bit tricky, clearance is tight for the rivet tool, although you can modify one.

Disassembly (I highly suggest you make notes on what goes and fits where, as you disassemble the canister):

NOTE:   Some screws may be quite tight.  You need a PROPERLY fitting screwdriver. Don't bugger up things!
Remove the pin that holds the coupling to the shaft....at the end that goes into the camshaft.   You will need a small diameter round drift.   Be careful doing this, some are in very tightly, some are not.  If quite tight, you may want to do this with a press, or make up something so as to use your vise.  Do NOT bugger things up. In any event, support the shaft.    You will find some shims and a washer. 
Remove the cap (Phillips type screws).  That cap normally has a spongy material on the inside.  It tends to deteriorate.  Remove it all.  That spongy material can make a mess inside the canister once it moves all about.   Remove screws for the bearing plate, and remove the plate; noting how the plate fits.   Note the big snap ring, and how it fits in a groove.  

Remove the C clip.  Remove the three outside screws.  Looking inside the canister, notice the 3 tabs that those 3 flat screws fit into.  Mark a tab and can, so you replace them later in the same position...(if you want to), but the tabs are not equally spaced, so this is moot.  Now remove the screws.

The flat screws hold the retainer.  That fits the snap ring only in one direction....a keyhole in the retainer matches a protruding stud on the snap ring.   Note that how the big snap ring fits, where its open end is.

Remove the plastic rivet from the wire strain relief.   Note that the plastic strain relief has a tit, you have to pry or pull it out....once out, the strain relief moves in the slot easily.

The coupler has a spring thing wound around a groove.  Pry an end loose, and remove from coupler.....you now need to remove the pin which is now visible.   You need to improvise something that will not injure anything, to remove that pin.  Do that now.

 Use a soft hammer to move the shaft through and enable removing the contents, do this a bit at a time with the plastic wire strain relief...back and forth.  Remove slowly, don't use a lot of force, and write down what you see about ....ALL about the many washers, etc., you find.

NOTE!   BMW does not sell canister parts.  I think Motobins in UK sells the springs for the advance weights.

It will be clear to you to remove an E-clip, and another snap ring, which hold the 'shield' in place.   Note the pin, locating the advance and shield.  Don't mess up that shield!  Don't loose the small location pin either.  This stuff is expanded upon in the next paragraph.

NOTE regarding the Sensor, if yours has failed:   Once the canister is basically disassembled, you need to remove the 'vane' without injuring it.  The 'vane' is between TWO C rings.   Remove the top C ring, and remove the pin (very small).  Be careful.  You will have to improvise something for this vane removal.  You can now remove the other C ring, and the 2 flat head screws on the sensor bottom.  The sensor is riveted and you have to drill that out....replace with a steel pop-rivet or nut/screw..  Use a Dremel or similar or?? to grind off the rivet.  See note at beginning of this canister section, about the dualplate and blind rivet models.   Cut the wires on the old and new sensors, so you can connect them and insulate them (shrink tubing).     Be cautious about the re-riveting; and keep things in proper shape.    If the wires are long enough from the sensor, splicing can be done outside the canister.   I prefer them connected inside, and the wires kept short enough to avoid fouling on moving parts.The triggering magnet part of the sensor and is retained by an E-clip and a snap-ring.  The snap-ring does not hold the rotor in place, the PIN does.  The snap ring keeps the pin in place.   The PIN locates the magnet, with respect to the shaft.  Don't lose that pin. Don't mess up here.   Put the unit over a common socket of correct size, all on your workbench, hit the shaft with a SOFT hammer (plastic?).   

It is very important that the parts be very well cleaned, and if need be the shaft burnished in the ATU area.  You need to use some judgment and have some experience here.  Sticky ATU problems show up as increasing idle speed USUALLY when the engine is quite hot....as the parts expand from heat.   The weights are on a Teflon thing, don't oil the Teflon unless you have an oil that is exceptionally slow in evaporation and probably contains Teflon;....I use a Teflon grease in that area.   It is this paragraph that you need to do if you have sticky ATU function...but if the canister is apart, do it anyway.

When doing the final assembly of the canister, pay attention that the shaft continues to rotate freely, in particular as you tighten the side screws.  Ken Lee had one that "dragged the edge of the rotor with the bottom of the sensor housing between sensor and magnet" (his words).  He suggested cutting a small relief slot with a Dremel tool for clearance.
When mounting the canister to the engine, install a new O-ring at the base.

NOTE:   The sensors vary somewhat in how they look or are mounted. Pay attention to what you are doing.  The canisters that have a lower plate and upper plate assembly that has the blind area, can be worked on, with some thought.   You could drill a couple of access holes for a riveting tool.   The Classic K bikes use two of the same type of sensors, just not in a canister!    Oilheads are similar

Modifying the canister for dual-plug conversions:
This is, or is not necessary, depending on who you talk to.   One point of view is to use an IDLE ignition timing of OT, or slightly advance from OT; and leave the canister advance unit stock.   Some who have this point of view may recommend bending the advance ears outward just shy of touching the case inside wall.  That adds a bit of maximum advance.  Another point of view has it that the advance should have its total amount shortened a bit.    One article on posted by Pete Serrino recommended using one or more layers of shrink tubing over the advance weight stops.   He did not specify the type of tubing, but said that one layer of tubing shortened the curve by 2° crankshaft; and he uses 2 layers for dual-plugged 336 cammed bikes, adn 3 or 4 layers for 1000 cc 308 cammed bikes; and one layer less on 800 cc bikes.   I honestly do not thing that restricting the advance range is really necessary, and, in any event, I feel it depends on many factors, including the compression ratio in use.   Certainly for very high performance engine modifications, one might want to delay the maximum advance.   I agree with what Pete said about one modification:   Drilling the weights with a 1/4" hole about 2/3 up from the pivot point, and, a 1/8" hole close to the first hole, will move the rpm up 500.   I prefer stronger springs, but, like Pete, never found any, although some Porsche ignition springs were quite good enough.

Revisions:
09/16/2002:  wire color code, notes, links fixed and verified; red emphasis areas. Additional changes to clarify meanings.
01/25/2003:  additional minor clarifications and cautions; add Newark part number for the Hall element; note on canister/high idle; minor clarifications and grammar and typos of little importance.
02/01/2003:  points parts number; gray coil information; information on Dyna and Boyer
to 02/05/2003:  module timing and color coding information; some rearranging and editing of little consequence; other modules from cars; add suppressors information; information on dwell; canister shaft swelling; Bosch module versus SYSTEM, which I've colored, just above the sketch, as green
05/15/2003:   minor clarifications, more information on coils; fix hyperlinks operation
05/15/2003:  correct .htm, I upper case must be i lower case
07/13/2003:  hyperlink for #2.
12/05/2003:  greatly expand information on substitute modules
02/03/2004:  expand information on modules and coils slightly in preparation for information to be obtained in future
03/09/2004:  update slightly, not uploaded
10/31/2004:  Add several URL's; edited entire article for better clarity.
12/30/2004:  expand information on alternate sources for canister points
04/09/2005:  add note at very top
05/11/2005:  Electronic module and coils information from OAK, from Airlist; slight revision here and there due to that.
03/22/2006:  Add Stan Smith, Rocky Point cycle as source, and Transpo number
04/13/2006:  Correct hyperlink to ...oldbritts....
05/30/2006:  updated things on the new modules and coils, for clarity
11/16/2006:  Minor editing, and updating/expanding information on RockyPointCycle
11/23/2006:  Add another source for the Hall elements
11/27/2007:  Technical details updating.  Modify testing and wiring notes.  Stopped major updating, awaiting when I can personally overhaul a canister, step-by-step, making sure that section of this article is clear to anyone.

Return to Technical Article Index

Return to HomePage