This is just some informal "thinking out loud" about the GL1000's ignition system. It is repeated for the most part on the Naked Gold Wing forum at this link because that is for whom I first wrote it.
GL1000 Kettering ignition owes much the the CB450, and it in turn to Honda's original two-cylinder cars (which by the way whose engines were built much like the CB350 twins'). Two ignition coils, two points sets. The difference of course is the Wing's four cylinders, which means its two ignition coils have two plug wires and have to fire twice as many spark plugs, so let's deal with that first. Honda was using half as many ignition coils to run twice as many cylinders as early as its 160 twin days, and probably before that but I am not as familiar with those machines going back into the dark halls of history. (A motorcycle older than I am is just too old to bother with in my view...) Most of the multicylinder four-stroke engines of the 70s and into the 80s used this so-called "waste spark" system, wherein one ignition coil connected to two spark plugs. Even car makers got into the act, later than the bike makers in an unusual reversal of their usual roles, in the early 1980s. They called it DIS, or "distributorless ignition systems.". Means just what it says, no distributor. Just like on bikes, pulsers to coils, coils to plugs, transistors for switching. This distributorless or waste spark system relies on the fact that sparking a dual lead ignition coil twice as often as previously necessary is a convenient way to fire two plugs in two distinct cylinders. The double firing means each cylinder at alternate times receives a spark when it doesn't need it, so that the other cylinder can get one at the same time that it *does* need, and this trades back and forth. The previously "dud" cylinder becomes the active one and the previously active one becomes the dud. This system doesn't hurt anything because the exact opposite point in each combustion event in a four-stroke engine is an exhaust event, meaning the unused ("waste") spark occurs harmlessly (if uselessly) within a combustion chamber filled with exhaust gasses. Works fine. There is a lot more that could be said about this, but interestingly, the wasted spark system requires that one spark plug fire from center electrode to ground electrode, and the other plug on the same coil fire "backward" from ground electrode to center electrode, after the electricity has traveled though the cylinder head. Yup, the two plugs fire in series. Banish that image in your mind of electricity traveling down the two plug wires to the plugs and after firing just disappearing. That ain't what happens. Electricity has to travel in a circle, and the circle includes the two plugs and their wires.
You have noticed there are two marks on the GL1000's flywheel viewable through the ignition timing inspection window. One is "T" for top dead center (TDC, used while adjusting the valves) and the other "F" (fire) which is used for ignition timing, Those relatively new to engine tech will wonder why ignition can't commence at TDC. Well, it could. The reason most engines don't is it takes, at idle rpm, a few crankshaft degrees' movement to get combustion enough underway to have the resulting cylinder pressure near its peak as the piston positions itself for its rapid downward journey. In fact every engine has its own ideal amount of "advance" i.e. head start, it likes to have to get this result. Some 10 degrees, some only 5, and there have indeed been engines that timed at idle right at 0, or TDC. Car engines, mostly, and for emissions reasons primarily. You may also have heard of ignition advance spoken of in another way, and that is what is often called "full advance," referring to as the engine is revved, the timing needs to advance still further than that initial 10 degrees to as much as 35 or 40 degrees before top dead center. On all the four cylinder Wings, the spring-loaded mechanism found hiding behind the ignition points plate takes care of this chore, with flyweights that react to rpm. Six cylinders advance electronically. Full advance is necessary because a faster turning engine's ignition needs more of a head start than it needed at idle. Engineers have for many years spoken of full advance figures, such as 24, 27, 38 or 42 degrees, and related them directly to combustion efficiency. That is, how evenly and promptly combustion is happening in the cylinder. Engines requiring large amounts of full advance are those whose combustion is not as smooth as it could be. The early Wing has virtually hemispherical combustion chambers, which while not the worst they could be, are not the best burning either. (Another topic for another time.). Engines requiring lower numbers, i.e. smaller amounts of full advance, are engines having superior combustion characteristics, and for a given engine size, tend to make more power. These are engines with almost flat chambers, whose flame movement is unimpeded by odd combustion chamber contours and combustion is therefore tidily prompt. Today's sprocket rockets in fact run timings in the 20s. Their combustion is so good, they not only get by with half the ignition advance older engine designs such as our GL1000 need, they do it with cylinders running 12.5:1 compression and some of them at least using 86 octane unleaded! It is no coincidence either that these new bikes have valve angles that are only 12-15 degrees from completely vertical. Older engines that needed airflow over their combustion chamber roofs had to have their valves splayed outward far, which in turn dictated hugely voluminous combustion chambers, which needed lots of ignition advance. Later engines are made with very steeply positioned valves, are liquid cooled to compensate for the lost access to the top of the chamber by airflow, and make tons more power, with half the total ignition advance. So the full ignition advance requirement of an engine ultimately goes back to its valve angle, but in any case speaks volumes about its engine's combustion efficiency.
This matter of the GL1000 having wobbly ignition timing is an interesting one. Dealer techs of course glommed onto the fact immediately (I was discussing the split timing method with Fred Germain, Honda corporate's L.A. area tech rep in the Spring of 1977), but it was nonetheless interesting for me to read about it in print for the first time just a few weeks ago, having acquired a copy of the Wingworld magazine (thanks, Jeff!, aka California Wing Nut) in which Mark Overby outlines the technique. There is no better way to do the job if you're going the static timing route, and I do prefer it myself, though I have tried both static and dynamic method and in fact started with the dynamic method back in 1975 and continued with it for two years, using a custom-made dwell meter created for me by one of the shop's customers. Before timing your GL1000's ignition you need to make very sure of the cam belt condition and more importantly tension. I have a particular method of tensioning which involves finding the loosest spot by hand and not relying on crankshaft positioning. But it is probably mostly being anal on my part.... You will still have the camshaft clearance in its bearings to contend with, unfortunately, even with properly adjusted belts. Putting a dial indicator on a cam still in the engine will show between 0.001"-0.003" total runnout, not surprising given Honda's designed-in loose fit of the cam to the head (all their overhead cam bikes are like this, and are worsened of course by wear) and the effect valve spring pressure has on the rotating camshaft, causing it to be pushed back and forth slightly in its bearings. Nothing to be done about it, I am afraid, though I have experimented with specially modified advancer assemblies, ones hand filed (stoned, actually) to compensate for the camshaft's wiggle. I have had appreciable success in getting the two different timings on one set of points (the crux of the issue) to within 1 crankshaft degree, almost eliminating the "ghost" effect (what is seen on a strobe type dynamic timing light). But it's painstaking work and is very hard to do exactly. It does however validate what is happening at the camshaft. For my own or special customers' bikes, it is worth it, but most folks aren't going to want to do it. So do the split timing thing, and while you're at it, don't worry about the gap overly much. Now finally we are to the controversy, and I wish I could avoid it, but I guess I can't. Overby makes much of point dwell in his article, and while I understand his point I think it unfortunate as it complicates things and thus discourages owners from doing the timing job themselves. My opinion is, if both points are within the factory 0.012"-0.016" range, you're good. Don't worry about it. Doing the timing is hard enough by the static method. You don't need added worries thrown in.
Now let's get practical. The real difficulty in timing, dynamically or statically, the GL1000 (and any of Honda's points powered multis), is two-fold. First is the fact that Honda designed the points assembly to not have any concentric pieces. Nothing is on center! That is, when rotating the base (backing) plate to time the left side, the left side gap changes at the same time. The same is true of the sub plate (where the right side points are timed). Pivot it to dial in right side timing and its points gap changes too. So after painstakingly getting both point assemblies perfectly gapped, then you rotate the base plate a smidge to get the timing right, and it ruins all your hard work, throwing the gaps (both left and right) out of adjustment. This is the paramount problem, and it is designed-in and cannot therefore be overcome. Making things even worse is that the fit of the points plate (backing, base) in the cylinder head is somewhat loose, with the result that each time one loosens the big Phillips screws to rotate the base plate to adjust timing for the left side points, the slop in the machined area that grips the plate makes the gaps in the two points assemblies change once again! So the gaps change with every timing adjustment, due to two system shortcomings -- non concentric mounting of the points to the backing plate, and loose fitment of this same plate to the head. I gently peen the cylinder head area around the backing plate whenever I do one of these (on SOHC inline fours too), but a) you can't do it too much or the backing (base) plate becomes impossible to remove, and b) you still have the non-concentricity of the pivoting parts to contend with, so little is gained. Take heart though, the exact same problems exist on the twins and inline fours. So the drill is, gap, then timing, then gap again, then timing again, in ever-smaller increments, until both are where you want them. It takes a pro about three go-rounds, so I imagine for a weekend tech it might take several. If this process isn't your experience and you are a master at static timing, I would appreciate hearing from you! I know of no other way. The only difference doing it dynamically is it takes slightly less time as you can see the gaps change by viewing the dwell, you don't need to physically keep regapping as you do with the static method. You just tweak the dwell each go-round as necessary. Of course the main reason folks like myself don't prefer the dynamic method is the difficulty of viewing the marks though the timing inspection window, even using the factory "bombsight," which I have, and even with the bombsight modified by the knocking out of one of its two windows, which I did 37 years ago. Still too hard to see in my view. Of course, now we have Randakk's new degree wheel that can be mounted on the left cam pulley, replicating the method Triumph innaugurated on their twins back in the 60s and Honda themselves resurrected on the 1979 six cylinder CBX. It promises to be useful.
Dyna and their ilk I have one problem with. Though they eliminate all this fussing with gaps and such, and also eliminate the need to maintain the points condition, being replacements for the points, they are not as reliable as they could be. I have removed as many or more than I have installed, if you can appreciate that. In addition, when fitting the Dyna timing rotor onto the stock advancer assembly, it is all too easy to end up with an advancer assembly that does not fully advance. It is imperative that you check full advance when installing an aftermarket electronic ignition. The problem was (it may have been fixed by now for all I know) the Dyna rotor was thicker than the stock points rotor, so the advancer flyweights came to rest in a partially deployed position, meaning that some of the advance was already used up. Watch out for this, as I say, by checking the full advance.
Powersports Ignition Evolution
Resistor Plugs, Wires, and Related Issues
U-Gap and Splitfire Spark Plugs
Optimizing the Honda SOHC Ignition
Honda DOHC, CBX Ignition System Troubleshooting
Reading Spark Plugs
High Performance Ignition Coils