Note:I'm sorry, but I can't answer individual PMs about issues with ET systems. If you have a question that isn't answered below please post it to this thread and I'll try to answer it if it's of general enough interest.
Troubleshooting, Repairing and Timing a Lucas ET (E.T.) Ignition System
Questions about ET systems come up often enough that a number of people should find a "project" thread on the subject a useful reference, and the first day of a new year seems like a good time to get started. What will follow is based on a document I wrote about ten years ago, although I'm still revising and adding some new material for use here. Also, I'll incorporate answers to any questions that come up.
Although it's popular to criticize the ET system, my experience is that much of that criticism is undeserved. That said, it is more sensitive to sloppy maintenance than is a "regular" ignition system, and it only produces 35 Watts for the headlight. On the other hand an ET system completely eliminates the need for a battery along with the cost and maintenance issues that go along with that. However, my purpose in writing this isn't to try to convince anyone to use an ET system, it's to explain how to troubleshoot and repair a malfunctioning system.
First, a bit of history. Magnetos on airplanes have difficulty at high altitude due to internal arcing that increases as the pressure gets lower. To overcome this, just before WWII Bendix invented (or implemented; I don't know who actually devised it) a "low tension" version of the magneto that separated the components, placed the high tension coils adjacent to the spark plugs, and eliminated the arcing problem. This magneto was used on all Double Wasp radial engines during WWII (Corsair, Thunderbolt, Hellcat, B-26, etc.) proving the intrinsic reliability of the design.
The next composite shows the circuit of the Bendix "low tension" magneto along with that of the ET electrical system, with the corresponding components highlighted in color (and the electrically separate lighting circuit blacked out).
As can be seen, the two systems are electrically identical with each other as well as with a conventional magneto, and thus the functioning of the ET system needs to be understood as the magneto that it actually is.
The ET system is the basis of most modern self generating systems. The points are replaced by electronic switching. One way to assist a good spark when kickstarting is to turn the lights off, this frees up magnetism for the ignition side. A few years ago I wound some coils for a AJS stormer. The ignition used 2 coils, I fitted 2 coils for the lights and 2 for the brake lights and horn. Braking or sounding the horn dimmed the already poor lights even though there was no electrical connection
The brilliance of a motorcycle ignition-magneto is that it is A self-contained auto-tranformer spinning on a shaft In a permanently magnetic field That can produce Primary and secondary voltage ... and snap the spark .
The Lucas ET ignition took this simple idea and made it as complicated as posible . Both mechanically and electricly . Primary voltage is from one location , the stator ( a magneto ). Secondary voltage from another , a coil. The Points are not opperated from near the primary , fleeting AC voltage source , But through shafts and gears at a 3rd. Location ... thereby complicating the signals potential capture . This separation of magneto parts with wires and gears unduly complicates the simplicity of the concept . The ET ignition is not an improvement , but an aboration in engineering. You will not find any thing like it on any aeroplanes .
Want to run without a battery ? Run a normal dc ignition with a capacitor. Its a magneto that's primary spark is now dc controled.... no need to time the points-spark to an AC wave form .
The Lucas ET ignitions true brilliance is that it can be easily maintained by anyone with 3 hands . .
I disagree, the ET setup from a production costing point of view at that time is OK. The HT coils are separated from the heat and a low cost stator is used as the generator. Many japanese and chinese and indian made marques use/used a similar setup. It's far more related to cost than anything else, engineering is a secondary thing. Proper magneto's are expensive!
Most small integral CDI type systems work in a similar fashion too, putting Thyristors etc on a bike in those days would have been another expensive development exercise. They wouldn't even spend the money on a decent voltage regulator!
The brilliance of a motorcycle ignition-magneto is that it is A self-contained auto-tranformer spinning on a shaft In a permanently magnetic field That can produce Primary and secondary voltage ... and snap the spark .
I completely agree with you. On the above. However, on everything else you wrote, I disagree.
Not that many people change their minds on opinions as strongly held as yours, but please hang in there as the thread unfolds. Perhaps something might cause you to decide there may be a tiny bit of redeeming value to the Lucas ET system.
I am looking forward to this ,Magneto Man,its already off to a great start. Thanks in anticipation.
My own limited experience of ET ignition is based on the field bikes my friends acquired for tattie money. Bantams, C15s , mopeds. Most had a mag of some sorts, the C15 S, with ET was always a runner, didnt foul plugs and was a mystery to us all, we never fiddled with it and it kept running, it like nearly all of the field bikes needed to be bump started, once warmed up a kick was possible. Most stuff that ran with a normal coil ignition had failed or faulty gennies and needed fresh batteries to run. This was the mid 70s when a field bike was 5 - 20quid, money for spares was short, petrol was harvested by the pint.
71 Devimead A65 750 56 Norbsa 68 Longstroke A65 Cagiva Raptor 650 MZ TS 250 The poster formerly known as Pod
It's easy to see from many posts on this Forum that a high percentage of motorcyclists have an aversion to anything electrical. I believe that, plus the fact an ET system isn't as tolerant of sloppy maintenance as is a battery system, accounts for many (if not most) of the knee-jerk negative comments about it. However, I hope with this thread to bring the E.T. system out of darkness by shining the full 35 W of headlamp illumination on it...
Originally Posted By gavin eisler
the C15 S, with ET was always a runner, didnt foul plugs and was a mystery to us all, we never fiddled with it and it kept running,
Thanks for your comments. The ET system is ideal for this sort of use. When it was designed in the late '50s it was to replace the magneto on sporting machines where both weight of a battery and acid leaking from it when a bike was on its side on a scrambles track were significant considerations. What killed it wasn't any sort of reliability issue but the fact the U.S. required machines made after 1967 to have brake lights that worked when the engine was off. That effectively removed a major market so there was no choice but to switch to batteries.
For ~15 years my ET-ignited C15S has spent 51 weeks each year at the back of a friend's garage in Ireland. We fill it with gasoline, start it, and ride it for four days in a Rally, then drain the tank and put it back to sit for another year. Try that with a bike that requires a battery for the ignition. I have battery chargers and cables cluttering the floor of my garage because of modern motorcycles, but my ET-ignited Triumph 500 only needs gasoline and it's ready to go.
As a preview of posts to come, I'll present a step-by-step troubleshooting guide. If someone who is facing an unfamiliar bike follows these steps they'll end up with a nicely running motorcycle. It should be clear -- I hope -- which steps in that comprehensive guide can be skipped when someone's familiar bike starts acting up (e.g. since they know their bike used to run, they'll be able to skip the step about checking whether the rotor has the keyway slot in the correct position, which my C15S did not when I got the machine). That will be followed by wiring diagrams and then by specialized information for those who want to know more (e.g. size and no. of turns of wire on stator and HT coils, inductances, part numbers, etc.).
p.s. I'll try to switch between ET and E.T. and use the term Energy Transfer occasionally so future searches don't miss the thread.
Hi Magnetoman, and thanks for starting this topic. I have a 67 T100C that I will be restoring soon. It had ET ignition originally but a previous owner took it all out. I have collected all the parts needed to go back to ET so look forward to your posts, as I am a complete numpty with electrical stuff.
The Lucas ET ignition system was introduced in 1959 on certain BSA and Triumph sporting machines and used the 2ET high tension coil (45112A/B or 45112D). Presumably there was a 1ET but I've never run across one so perhaps it was a number used only during the development stage. Sometime in1962 Lucas switched to the 3ET (45149B for the right side of twins; 45150A for the left) that remained in use until the end of production. Although, as can be seen from the next photograph, the 2ET is very crude and heavy, electrically it is the same as the later 3ET.
(left) Lucas 2ET and (right) 3ET coils.
Although based on limited data, I believe the high tension side of both versions has a fundamental defect that causes some or all of them to die of old age even when not used. This happened to both 3ET coils on my Triumph 500, to one of the two 3ET coils on a friend's BSA Cyclone, and to a new-old-stock 2ET that I tested. My speculation is the factory used slightly acidic flux to solder the very fine wire of the secondary (0.003" dia.) and that this slowly corrodes the wire with time. This is consistent with the fact that although the HT lead of one of my Triumph's 3ET coils had infinite dc resistance, the coil worked fine when I tested it on my Merc-o-tronic coil tester. Immediately after that test the dc resistance had the proper 5.5 kiloOhms, consistent with having developed a temporary conducting carbon track during the test which carried enough current to spark the tester.
In any case, you cannot assume your coil is good so you will have to test it to know. An inexpensive ohmmeter of the kind sold by Radio Shack for ~$15 will at least determine if the HT resistance is the proper 5.5 kiloOhms value. If it is not, the coil definitely needs replacing. Suitable replacements will be discussed in the next installment.
I have a pair of NOS 3ET coils, so I will test them to see if they at least pass the test you have described. The Triumph parts book for '67 says 2 x 45149 so both the same side. I guess they just mounted one forwards and one facing backwards? Or maybe its just a misprint in the book? Both of the coils I have are 45149s, I could never find any of the other side. Is anybody able to confirm how they are mounted on the later bikes? I have some photos taken for me by a Britbike member some time ago of a bike that appeared original and it looks like the coils are mounted facing opposite ways.
The Triumph parts book for '67 says 2 x 45149 so both the same side.... Is anybody able to confirm how they are mounted on the later bikes?
I can confirm how they were mounted on earlier bikes (at least on one of them). The following photograph is of the left and right coils from a 1966 Triumph T100C:
(the left coil is on the right in the photo and vice versa in order to place the coils themselves closer together; the photograph is from the perspective of looking up from the engine, i.e. the HT leads are on the bottom of the coils when mounted on the bike)
As can be most easily seen from the exit of the primary wires from the same end as the HT lead one coil and the opposite end on the other, the coils are not identical. Interestingly, though, the 1966 parts manual shows 2 ea. of part no. 45149 "Ignition coil, type 3ET, rather than two different part numbers.
Since I removed these coils from a 1966 T100C and are known by me to have been on that bike from within a year of being new, whether all '66 T100Cs came with two different coils, or some were shipped by the factory using the same one on both sides as indicated by the parts manual, is a mystery for concours judges to sort out. Of course, 1967 was a completely different year...
What is the difference, reliability wise, between a flat battery and a dead coil?
The answer, of course, is there is no difference. However, those aren't the only two choices someone has. To quote my earlier post: "Suitable replacements will be discussed in the next installment." With such a replacement the choice will be between a motorcycle with a flat battery and one that is fully functional (albeit, with a 35 W headlamp).
Ah...the primary wires coming from different ends of the coils might have put me wrong with the picture I have. I cant really see clearly but could see the primary wires and (probably) wrongly made the leap that they were mounted facing opposite directions. If anyone can provide a definative answer on this I would be grateful (and also a bit gutted since it was bloody hard finding the pair I have).
1) Step 1 of this troubleshooting guide is to test the resistance of your HT coil. Three electrical leads come from both versions of the ET coil; two wires (terminals, in the case of the 2ET) and one internal spike to make contact to the spark plug lead when it is pushed into the hole. As shown in the table below the resistance between the two wires should be ~1/2 Ohm and the resistance between either of those wires and the internal high tension connection should be approximately 5-6 kiloOhms. If your coil has approximately these values it probably is fine. However, if either value differs significantly (most likely, the high tension circuit having infinite resistance), you will need a new coil.
----- Sidebar on Measuring Low Resistances----- Although an inexpensive ohmmeter will have trouble distinguishing between ~0.5 Ohms and a dead short of 0 Ohms, if you touch the probes together to see what the meter reads when they are shorted this way (typically, slightly offset from 0), then touch the probes to the coil leads to determine its resistance, you should be able to see a slight difference in the reading for ~0.5 Ohms. You might have to do this several times to convince yourself you're seeing a real effect and that the resistance isn't actually 0. Alternatively, I recommend the Wavetek LCR55 meter since it also will be useful for setting the timing using its inductance scale.
As can be seen, on its lowest resistance scale the Wavetek allows zeroing the meter to compensate for the leads and then measuring to a resolution of 0.01 Ohms. That said, because the contact resistance becomes significant for these small resistances the exact value of such two-lead measurements varies so can't be taken too seriously, especially on corroded connections like on this ET coil. Only a four-lead measurement can provide accurate values. Despite this, the 0.01 Ohm resolution of the Wavetek is very useful. ---------- end sidebar -----------
Luckily, Honda, Yamaha, and others used an equivalent ignition system on some of their motorcycles during the 1970s so excellent replacement coils are available from motorcycle dealers. Also, recently an aftermarket copy of the Lucas ET has appeared as well. However, I haven't tested it so I can't comment on its quality or whether it has the correct electrical or mechanical specifications. This aftermarket copy is easily identifiable from its use of red epoxy to encapsulate the coil.
If your Lucas coil is bad, and even if you can find a genuine Lucas ET, I recommend swapping it for one of these Japanese coils even though you will have to fabricate an appropriate bracket to mount the coil to the motorcycle frame. The coils on my BSA C15S and Triumph 500 are from a 1974 Honda XL125 (part no. 30500-355-003, superseded by 30500-950-405), with a list price of ~$75 from the dealer. Used ones also can be found on eBay. Appropriate ET-type coils were used on 1972-76 XL70, XL100, XL125, XL250, XL350, TL125, and MT125 Hondas (later machines with the same model numbers used different, non-ET coils). An ET-type coil also was used on 1978 Yamaha IT175, and quite possibly on other makes and models.
To make sure the shop doesn't accidentally sell you an incorrect coil from a mislabeled box, externally you will find it similar to a complete Lucas ET assembly. That is, it consists of an epoxy-encased coil riveted to the jaws of a C-shaped laminated iron yoke. Note that certain coils designed for CDI ignition systems look somewhat similar to the Honda coil pictured below although they will not work on an E.T. system. If there is any doubt, they can be distinguished from their HT resistance, which is only ~2.5 kiloOhms. If you use something other than the Honda coil I am recommending, you are on your own.
Coil______________Primary_______High Tension Lucas 2ET___________0.6 Ohm____________? Lucas 3ET___________0.6 Ohm________5.2-5.6 kiloOhms Honda XL125________1.0 Ohm________5.4 kiloOhms Yamaha XT500_______0.75 Ohm_______5.7 kiloOhms
Honda coils come with a 0.25 microFarad condenser permanently riveted to them so the ~0.20 microFarad Lucas condenser inside the distributor or points housing should be removed (it's already at least 30 years old so likely has failed by now anyway; if not, it will fail shortly). Eventually, like all such inexpensive condensers, the one hard-wired to the Honda coil also will fail so I recommend cutting its wire and reconnecting it using a male/female set of electrical connectors. This will enable easy testing of the condenser and, when it becomes necessary, removal from the circuit and bypassing with a new condenser.
Honda ET-type coil that is an appropriate replacement for the original Lucas, although requiring fabrication of a mounting bracket.
If you choose to use a Lucas 2ET or 3ET coil it is very important to use HT leads made with actual wire as the conductor, not carbon string. The coil makes contact to the HT lead using a metal spike and the arc from the spike to the lead that occurs every time the plug fires will slowly erode a nonmetallic conductor. This will slowly open up a gap inside the HT lead that will increase in size with time. Once the gap is long enough starting the bike will become increasingly difficult and eventually it will become impossible.
[to be continued]
Last edited by Magnetoman; 01/05/161:56 pm. Reason: added paragraph about using HT lead with wire
2) Next, make sure your points cam is of the correctly shaped "ET-type," with an egg-shaped cam (left on the composite photograph below) that leaves the points closed for most of the rotation (~320o). The more common battery ignition system uses a cam that is of a slightly larger diameter that holds the points open for most of the rotation (~285o) to keep the coil from overheating. Instead of a protrusion it has a pronounced flat region where the points close long enough to allow sufficient time for the current in the coil to "saturate" (~75o) after which the end of the flat region is reached and the points open (right);
(left) ET and (right) battery/coil points cams.
3) On an engine with a distributor make sure the points plate has a 10o advance range (equivalent to 5o of advance on the cam). It should be stamped "10" on the auto-advance bob weight just below the points. Since they are mechanically interchangeable it is quite possible someone replaced the points plate with a "standard" one having a 20o advance which is too large to function with an ET ignition. With too much advance the bike can be made to either start, but to cut out as the rpm rises, or run well, but be very difficult to start. There also is an intermediate 15o points plate that should work, but it's safer to reduce the range of that one to 10o. This can be done with a bit of solder in either case, as described in the next figure. Even if yours is stamped 10o you will be checking it later in these instructions to make sure;
This figure shows where to apply a small amount of silver solder on a non-ET points plate to reduce the advance range if necessary.
I too with a A65H awaiting reincarnation am interested. Are you aware if anyone has successfully developed a technique to 're-pot' the windings on these. Cracking seems to be an almost inevitable fate for 3ET coils.
BSA 1969 A65F BSA 1966 A65H Triumph 1968 T120 Kawasaki A1R & too many projects!
Are you aware if anyone has successfully developed a technique to 're-pot' the windings on these.
It would be straightforward to use an existing coil to create a mold from which subsequent ones could be cast using an appropriate epoxy matched to the amber of the originals. However, even if the windings in an existing coil were fine finding a chemical that de-potted the epoxy without removing insulation from the wires would be problematic.
If I were going to make a coil that needed to pass concours inspections and also be highly reliable I would de-pot an existing ET coil, remove the wires, wind a new primary and secondary, and recast the epoxy using a mold I made from an existing coil. As for time, I think within a few weekends' of work something meeting expectations could be made, after which additional ones could be made more quickly.
A second approach, if I trusted the internal connection wasn't in steady decline as I speculated earlier, would be to create the necessary mold, "etch" the epoxy on an existing coil by ~1/8" or so, and then use the mold to create a fresh skin.
You can think of a standard ignition system as the battery supplying the necessary current to the coil and the rotor/stator simply in the background keeping the battery charged. Hence the relative position of the rotor with respect to the stator when the points open is irrelevant and, in any case, it is determined by a keyway. However, it isn't the same with an E.T. ignition system since it is a magneto.
Both battery and ET systems have rotors and stators with six segments each. As a rotor segment passes by a coil the current induced in it passes through 0, reaches a maximum, then decreases and passes through 0 again on its way to a maximum of opposite sign. The AC currents from the coils are appropriately combined and then rectified before being sent to the battery as DC. The relative position of the rotor and stator are irrelevant in a battery system because the output power is in the form of rectified DC.
In contrast, an E.T. system is a magneto so it is essential that the current from the stator coils be interrupted by the points at the moment the current is at a maximum. For this to happen the position of the rotor with respect to the stator when the points open is critical. In a conventional magneto, like a K2F, this is taken care of by a notch in the end of the armature and a corresponding 'pip' on the points plate ensuring the plate can be attached in only one position. That position being the one that ensures the points open at the moment when the induced current in the primary of the armature is maximum.
If the current in an ET system (or a conventional magneto, for that matter) is too far off maximum when the points open there won't be sufficient voltage generated to fire the plug. This can happen for two reasons even if all the components individually are in good condition: 1) the range of advance is too large, or 2) the rotor is incorrectly keyed to the crank. Odds are that no. 2 won't be the case, but those odds are not zero so it has to be checked.
To check of the advance range of the points cam as well as the orientation of the rotor and stator:
4) Set the timing by locking the advance mechanism in the fully advanced position and adjust the points plate so the points are just opening when the engine is at the proper value for running at high rpm (34o BTDC for a BSA C15S). Unlike with a conventional ignition system an ohmmeter cannot be used as an aid for this because the difference in resistance between the points closed and open is only ~0.5 Ohms. However, the difference in inductance of the primary is a factor of ~1000x so a meter like the Wavetek LCR55 mentioned in an earlier post makes determining the precise position where the points open easy.
Compare the relative position of the rotor and stator when the points are just opening when fully advanced with the image at the right in the next figure (also read the information in the caption). This is especially important when working with an E.T.-equipped motorcycle that you do not know for certain has run properly before (e.g. a basket case you assembled).
(Left) Approximate position of the rotor with respect to the stator when the points open with the auto-advance mechanism retarded. If there is a larger gap on the left of the magnet than on the right (i.e., if the rotor is positioned even a few degrees counterclockwise of the position shown here), it will not provide enough voltage at kickover to fire the plug. However, it is OK if the gap on the left of the magnet is smaller (or non-existent) than on the right when retarded. (Right) Approximate position of the rotor when the points open with the auto-advance mechanism fully advanced and the engine at the fully-advanced firing position (34o BTDC, in the case of a BSA C15S). If the edge of the magnet is further than ~1/16" to the right of the stator core -- the laminated iron protruding from the coils of wire --when the points are fully advanced, the voltage generated at higher speeds will be too small to fire the spark plug and the engine will cut out as the rpm increases. However, it is OK if the gap between the magnet and the stator core is smaller than the 1/16" maximum.
5) Unlock the advance mechanism so the springs pull the cam to the fully retarded position, thus causing the points to close. Now rotate the engine toward TDC and find where the points just start to open again (after removing the backlash, i.e. rotate closer to TDC then backwards until the points just open). Check that the rotor is aligned with the stator as in the left image of the figure. If it is, the advance range of your points plate is acceptable. However, if necessary, solder the slot in the points plate to reduce the range of the ignition advance.
If you find it impossible to adjust the points to achieve the relative positions of the rotor and stator shown in this figure you will have to deal with this situation using instructions that will be given later in a section on the rotor.
6) Check that the contact breaker gap is 0.014"-0.016";
7) Gap the plugs at 0.018", the same as used for conventional magnetos.
8) If you have made any adjustments to the points plate again lock the points cam in the fully advanced position, rotate the engine to the correct number of degrees BTDC for running at high rpm, and confirm that the points are just starting to open. Tighten the setscrew to fix the distributor or points plate in position, and thus set the timing of the engine at full advance;
It might seem that ET-equipped bikes would need to be kicked over quite briskly because of the limited range of the auto-advance mechanism since the 10o points plate results in fully retarded being ~15o BTDC (35o - 2x10o = 15o) rather than the ~5o BTDC (35o - 30o = 5o) on a standard battery/coil system. Still, in spite of this, my Mikuni-equipped C15S normally starts with just one fairly gentle prod when it is warm, and seldom takes more than three kicks even first thing on a cold morning. The same is true of my AMAL-equipped Triumph 500.
9) First, if any of the wires from your stator contains the color green, it is not an E.T. stator. Over the years the E.T. stator evolved from having 3 wires (late 1950s), to 4 wires (early 1960s), to 5 wires (mid- to late 1960s). The next photograph shows an older, open winding, 4-wire stator on top and a newer, encapsulated winding, 5-wire stator on the bottom.
The older unencapsulated stators have the advantage that if they break they sometimes can be repaired by soldering the break. However, the disadvantage is the wires making up their coils are much more susceptible to abrasion due to the vibration and thus much more likely to break or short. BSA issued a service bulletin at the time advising racers to encapsulate their stators using glyptal because of this problem. Based on one bad experience with an unencapsulated stator that failed I advise buying the more reliable later encapsulated type if you shopping for one on eBay.
Wiring diagrams with each of these stators will be shown later in this thread. As already mentioned, the 6 individual coils making up the stator are exposed in early versions but later were encapsulated in epoxy (there is also an uncommon, cut-down racing stator with only ignition windings). There also are internal wiring differences. For example, I am aware of at least three variations of 5-wire stators; one of these has the ignition windings completely separate from the lighting windings, while other two have the stop lamp draw its power from the ignition windings. However, the latest 5-wire stator can be used in even the oldest 3-wire application by making suitable connections as will be described later.
10) Check the electrical resistance of the stator. The ignition windings have ~3x more turns than do the lighting windings and since they also use finer wire in order to fit the additional turns in the same available volume this results in the ignition circuit having significantly higher resistance than the lighting circuit. With the stator leads disconnected an ohmmeter should read approximately 5 Ohms for the ignition circuit, approximately 0.5-0.8 Ohms for the lighting circuit, and infinite between each of the wires and earth.
The wiring diagrams in a later post will have specific values for several variations of the coils. Even if the lighting windings turn out to be bad, unless they are shorted to the housing they will have no effect on the ignition portion of the stator, so there might be no need to worry about a problem with this if you never intend to ride at night, or need a brake light.
11) Although I am not certain if there were variations, it appears that throughout the entire time of production of BSA and Triumph twins that, rather than the E.T. rotor being keyed to the crankshaft, it was oriented by means of a dowel protruding from the engine sprocket which itself has to be properly oriented on the crankshaft splines. This is shown in the next drawing from a 1961 Triumph manual:
As this drawing shows if the sprocket is removed it should be replaced on the splines such that the peg protruding from it is at the approx. 9:00 position with the engine at TDC in order for the phase of the voltage output of the ET stator to be correct. Although there are 6 splines on the crankshaft and 6 magnets on the rotor, if placement is off by 1 (or 3 or 5) splines the polarity of the output voltage waveform will be reversed. However, later BSA and Triumph manuals don't mention this detail so the polarity of the tip of the spark plug could be reversed with less than optimum results (such is always the case with one of the two plugs on a twin with a magneto). Again, although a slot is broached in the rotor supplied with these twins a Woodruff key isn't used to locate it directly on the crankshaft. The peg in the sprocket turns the rotor.
Rotors for these E.T. machines have two or three holes on their inside face. For a BSA these are marked "S" (34o) and "R" (37o), and for a Triumph they are "S" (37o), "M" (39o) and "R" (41o). This is shown by the Triumph rotor at the left at the following photograph.
These holes allow the timing of the rotor to be closely matched to the opening of the points to ensure maximum current since racing machines typically having their points open a few degrees closer to TDC than do less sporting machines. It also makes it possible to mix up rotors intended for the other marque machine resulting in reduced output because the timing of the rotor/stator doesn't match that of the points.
In the case of BSA singles it appears the rotor was always held directly on the crankshaft by a Woodruff key in the slot broached in the rotor. However, the sprocket on the twin you are working on might have been replaced with a "standard" one without the peg, and the rotor replaced as well, so it might be necessary to alter the rotor as discussed below.
12) Rotate the engine to the factory-recommended value for timing when fully advanced (in the case of a BSA C15S this is 34o, or 9/32" / 7.1 mm, BTDC);
13) Check that the magnets in the rotor are aligned with the stator windings as shown at the top right of the figure in step 4) in an earlier post. If so, it means the hole in the rotor was drilled in the right location, or the keyway was cut correctly, and you can proceed to the next step. If not, it means the factory screwed up and you will have to alter the rotor before the timing will work. If you are using a non-ET rotor, i.e. one without the holes in the back face, in conjunction with an ET sprocket you will have to use the figure in step 4) to find the correct orientation and then drill the hole yourself to fit the peg. Similarly, for a rotor keyed directly to the crankshaft you may have to broach a new keyway to get the correct timing.
The orientation of the rotor might seem an unlikely thing to have to check, but enough ET rotors must have come from the factory with their keyways incorrectly cut because the BSA importer issued a service bulletin describing how to broach a new one, and even had a program to loan the necessary 5/32" broach and 3/4" collared guide to dealers to correct the problem. Also, I suspect at least some rotors for non-ET bikes have their keyways in arbitrary positions and one of those rotors could have made it into the box of parts from which you have built up your basket case.
I drafted the original version of this troubleshooting guide as a result of having to diagnose the ignition on a rolling basket case BSA C15S I bought. As described in the BSA service bulletin, I had to broach a new keyway in the rotor when I couldn't get it to run after rebuilding it. In any case, the necessary broach and guide like shown at the right in the above photograph can be obtained from a machinery supply company like McMaster-Carr.
Finally, when your machine passes all of the above tests, it is ready to fire up:
14) Just to be sure, replace the plug with a new one, gapped to 0.018";
15) Start it. If your carburetor is correctly jetted, it should now start easily and run fine.
My '67 ET Victor uses a woodruff key and not the pins to locate the rotor, as its a single and not the twins you referenced (although the stator, 54215824, does have the 3 locating holes on the back) as it's It's timed at 28o btdc. The shop manual lists a spark plug gap between 0.020" - 0.025", and I usually set it near the middle. Will using the slightly smaller gap of 0.018" that you suggest help my hot-starting issue, and/or in some other way?