Since I fried that ignitor module with my tig welder, I have been increasingly concerned about the scarcity of suitable modules on the used market.
I have one coming from EBay, no worry, but still.....
One good thing about the 87-88 version of the module is that it is not potted, so that it is possible to get a rough idea about what it contains, and I see a lot of passive components, resistors, capacitors, diodes, a darlington transistor, a comparator and a microprocessor.
This about sums it up, and the microprocessor already means that the unit is unrepairable, there is no way to retreive the firmware from it. And even if not, the surface of the parts is covered with ugly goo that makes it a mess to work with.
It would probably not be all that hard to build one from scratch, though.
See how messy the ignitor looks from inside:

This is what I had sort of figured out on my own, then I engaged chatGBT for a sanity check, and things got really interesting pretty quick.
I asked how a CDI or TAI system calculates the timing, and here is the first part  of the discussion that followed:

ChatGPT sagde:
A CDI (Capacitor Discharge Ignition) system calculates timing advance based on engine operating parameters, primarily engine speed (RPM) and sometimes load (e.g., throttle position or manifold pressure), depending on whether it's a simple or more advanced system.
Here’s how timing advance is calculated in a CDI system:

? 1. Trigger Source Detection
CDI systems use a sensor to detect the position of the crankshaft or camshaft:
Pickup coil (often near a magnet on the flywheel)
Hall effect sensor
Optical sensor
This sensor sends a signal each time the engine reaches a specific position, often before top dead center (BTDC).

⏱️ 2. Measure RPM (Timing Between Pulses)
The CDI unit calculates the engine RPM by measuring the time between trigger pulses:
Shorter time = higher RPM
Longer time = lower RPM

⚡ 3. Use a Timing Map or Logic
CDI systems use one of the following to determine how much to advance the spark:
Fixed advance curve (analog CDI): Uses resistors, capacitors, or basic circuitry to delay the discharge based on RPM.
Digital advance map (digital CDI): A microcontroller reads a preprogrammed timing map. Input: RPM (and maybe throttle/load). Output: spark advance angle.
The idea is:
At low RPM, spark is near TDC (Top Dead Center) to prevent knocking and ensure smooth combustion.
At high RPM, spark is advanced (occurs earlier BTDC) to allow more time for the air-fuel mixture to fully burn.
Example:
RPM
Spark Advance
1000
5° BTDC
3000
15° BTDC
6000
30° BTDC

? 4. Trigger Capacitor Discharge at the Right Moment
Once the correct timing is calculated, the CDI circuit:
Charges the capacitor
Waits for the right delay based on the timing advance
Fires the SCR (silicon-controlled rectifier) or transistor
Discharges the capacitor into the ignition coil → produces high-voltage spark

✅ Summary
A CDI system calculates timing advance by:
1. Detecting crankshaft position.
2. Measuring RPM (time between signals).
3. Applying a programmed or analog timing advance curve.
4. Firing the spark earlier at higher RPMs to optimize combustion.
Let me know if you want a visual diagram or how it differs from TCI or ECU-based systems.
Du sagde:
I want to know how the signal from the pickup coil relates to tdc.
ChatGPT sagde:
Great question — understanding how the pickup coil signal relates to TDC (Top Dead Center) is key to grasping how timing works in a CDI ignition system.

? What Is the Pickup Coil?
The pickup coil (also called the pulse generator) is an inductive sensor that generates a voltage signal when a magnet or metallic protrusion (on the flywheel or crankshaft) passes by it. This signal tells the CDI when the engine is at a specific rotational position.

? Relationship Between Pickup Signal and TDC
The pickup coil doesn’t detect TDC directly. Instead, it detects a fixed point before TDC — typically 20–40 degrees BTDC (Before Top Dead Center), depending on how the flywheel or trigger rotor is designed.
Here's how it works:
1. Trigger Magnet Position:
A magnet or metal tab is mounted on the flywheel or a trigger rotor.
This magnet is placed at a known angle before TDC (e.g., 30° BTDC).
The exact angle is fixed mechanically.
2. Pickup Coil Sensing:
When the magnet passes the pickup coil, the coil outputs a voltage spike (AC signal).
This voltage rises and falls as the magnet approaches and leaves — usually the zero-crossing or peak of the signal is used as the trigger point.
3. CDI Receives the Pulse:
The CDI now knows the crankshaft is exactly at 30° BTDC (or whatever the magnet angle is).
From this point, the CDI calculates a delay (in microseconds) to fire the spark at the correct time — maybe immediately (if it wants to fire at 30° BTDC) or later (say, 10° BTDC).

? Example:
Let's say your pickup coil is triggered at 30° BTDC:
At idle (1000 RPM), the CDI might delay the spark to fire at 10° BTDC → waits ~3 milliseconds after the pickup signal.
At high RPM (6000 RPM), the CDI might fire at 25° BTDC → waits ~0.28 milliseconds after the pickup signal.
So the pickup coil provides a fixed reference point, and the CDI determines spark timing by delaying from that known point.

? Visualization
pgsql
KopiérRediger
Crank Angle (°) --->

TDC ←---10°---20°---30° ← Trigger magnet here
                         ↑
                     Pickup coil signal sent to CDI
                         ↓
      CDI calculates delay → fires spark at desired BTDC

✅ Summary
The pickup coil signal marks a known crank angle (usually BTDC, not TDC).
The signal is generated when a magnet on the flywheel passes the coil.
The CDI uses this signal as a reference to compute when to fire the spark, based on engine RPM and timing advance logic.
TDC is calculated by knowing the fixed angle offset between the pickup trigger point and TDC.
Let me know if you want help calculating spark timing delay in microseconds for a given RPM and advance angle — it's a fun bit of math.

Here is a link to the chat for anyone interested, It is a good half hour read:
https://chatgpt.com/share/684b273e-5dc8-800e-99df-41a7802e9d39

Cheers, Finn Hammer

FinnHammer wrote on 06/12/25 at 13:08:42:

Thanks for sharing the pulser signal trace, very interesting!

I would hypothesize that the single reluctor bar that we see on the LS650 Stator  is placed such that it will trigger the spark with the max. possible advance.  (You could check this mechanically by noting when the pickup  coil is aligned with the beginning and end of the reluctor lump, and seeing how many degrees of rotation from each of these points till TDC)
https://www.thebonnevilleshop.com/ignition-timing/

The specs say timing for a LS650 should be 5 deg BTDC below 2000 rpm and 30 deg BTDC above 4000 rpm.   ( page 194 for a 1987)  https://www.scribd.com/document/362846253/suzuki-ls650-savage-boulevard-s40-1...

Then, I would guess that the engine will add delay, reducing advance, based on rpm.  I would guess the ICM doesn't trigger on the first pulse, but waits until it has a good idea of timing.

It might be that the ignition controller is timing between when it sees the  beginning of the reluctor to the end of its passing, (which I would suppose are the negative / positive spikes in the signal you showed us), and getting rpms from that, or even looking at the ripple, which is a function of the poles on the 3 phase stator which are generating power pulses at regular spaced intervals.  But, I would imagine that the rpm calculation come from the full cycle, which I would imagine is from negative going pulse to following negative pulse.

We can see that in 1996, Suzuki changed things a little, the reluctor node  for '96 until the eng of production grew quite a bit longer, which suggests to my mind that Suzuki was using the start and end of its passing for some purpose, and wanted a longer interval to sample.  https://www.ebay.com/itm/146200094003

Can you show us the signal for the full cycle, from negative pulse to negative pulse?  Also tell us the vertical scale on the scope trace, and about what rpm the engine was spinning at?

Thanks again for sharing this investigation.

Axman88 wrote on 06/12/25 at 20:51:40:

That was my first thought, but i diode tested them, and all diodes show normal 0.6ish volt diode drop in reverse, so I figured they are ok.

Needles:

the igniter bit the grass along with the also electronic decompression relay after I had done some tig welding on the side stand. I was careless there, I see that now, but nothing bad happened when I welded the header last year, so I did not give it any thought before too late.
I actually don't see any swelling on the 'lytics, but they are heavily smeared in with conformity coating.

My guess is the starter arc, which is several kV

Cheers, Finn Hammer

Needles,

Thanks for your input. I found that the M59201P lisyed as different things, microprocessor one place, probably also as linear amplifier, but I picked the micro as the most probable, but that just relays how I would have done it if I was in charge.
Here is a simplified schematic, from the Haynes manual.

Axman,

When trying to determine whether a circuit can output, or source, power, you have to look at the output stage.
In the case of the schematic you posted before, the output stage is a single transistor configured as an open collector stage.
A transistor can be made to block or conduct current, and this is controlled by the voltage applied to Base input. If the voltage is low, the transistor is shut off, and does not conduct any current, but if the voltage is raised, some small current flows from Base to ground.
More important, the transistor opens up, and starts to conduct from kollector to emitter, connected to ground.

In this way, the transistor acts as a switch, just like the contact breakers, but with the added convenience that it can be controlled by a low voltage/low current signal, like the typical output from a logic chip or a microprocessor.
Another convenience is, that the collector can be connected to any voltage that the transistor can sustain, so that the driving circuit can be for example 5 Volts, but the ignition coil can be any voltage you have available.
An open collector can only "sink" current, It cannot "Source" current.
The schematic you show specifically states that the spark is initiated by a CPU, Central Processing Unit, and this is the typical description of a Microprocessor.

Whether these ignitors are actually using microprocessors or analogue circuits is still puzzeling me, and I have little insight into how this task can be solved with linear amplifiers, whereas the microprocessor solution looks simple to me.

Anyway, to your question: what is the 200+V?
It is the inductive "kick" from a coil that happens when the current flowing through it is interrupted, which means that the voltage across the base of the transistor is shifted from high to low, shutting off the current flow.

One more thing about the ignition system.

When the first spark has to be made, the timing circuitry has not yet received the necessary information about RPM to calculate the right spark timing position based on the advanced trigger.
I now think that this is the reason for the double humped trigger coil signal.
The first peak, the negative one is probably the one that is positioned 30-40 deg. before TDC.
The later one, the positive one: It is my qualified guess that it is positioned exactly 4 degreed before TDC, so that it delivers a "Hard" trigger for start up spark position.
Is this probable?
Let's say the engine cranks at 400RPM, not an unreasonable estimated RPM cranked by the starter.
There are 360 degrees per revolution, and 400 RPM, that means that the engine covers 360x400 degrees in a minute = 144000degrees/min
In one second it moves 144000/60 = 2400deg
The peaks appear 15 mS (0.015 S) apart (you can see that there are 5mS between divisions on the scope (circled in red)).
in 0.015 sec. it moves 2400 x 0.015 = 36 degrees.
This does not look impossible, and has to be investigated. (But I have not yet figured out how to do that).
One way to prove it would be to have a timing disc on the engine, rev the engine to see the timing advance accordingly. Then apply 12V to the blue wire input, rev the engine and see if the timing advance fails to appear.

Mike?
https://suzukisavage.com/cgi-bin/YaBB.pl?num=1521968829

The black wire from starter relay instructs the CPU to chose between the 2 peaks according to the position of the starter relay.
As long as the starter is cranking the engine, the timing is determined by the last positive pulse, and when the engine is running, the starter relay is released, the CPU switches to calculating the timing based on the RPM information that it gathered during the first startup revolutions, and the negative pulse which enables it to advance the timing as the revolutions start to climb.


This concludes my present understanding of the ignition timing.

Cheers, Finn Hammer