Have some more info that may be helpful.
Because my last attempt seriously overheated the modules, and I was jumping off a car battery, I remembered Lou's description of the positive DC feedback, and wondered why the voltages were off so much that they were loading the modules with the full 5.8 amps @/or 6+ volts as they climbed.
I switched gears and just started looking up the OKI MPS-200 module on Google. I also had a question on what was the GM coil matched to the original 4 prong, at least on one application.
I looked up a 1980 Corvette with a 5.0L, and it has the 4 prong module with a slightly different coil due to the fact it is sitting below a 8 cyl distributor cap. It has 3 wires out and a common coil terminal.
http://www.rockauto.com/catalog/moreinfo.php?pk=935138You can see that the primary and secondary measurements are "OK" up to a .1 and 1k Ohm difference than the C849 coils.
I tried to look up how the Corvette was wired, but I could not find it. I did find wiring diagram for a 5.0L Bronco.
In this diagram, you can see that the black middle wire (probably) goes right to ground, as I would not expect anything else for a black wire with an eylet on it to be put except under a metal screw. This kind of goes along with what we know about the DIS type ignition, that when the double coils we are using fire, one spark goes the "proper" way from electrode to ground, but the charge coming out of the opposing Secondary terminal jumps the other way, from ground to electrode, which is what causes the wear. The other idea I had is that maybe it goes to an isolated resistor block that filters the signal for the tachometer and engine control feedback (like the 7 prong modules have for EST). The other thing I found in the schematic is that this particular setup uses a start ballast resistor. When the ignition button is switched, full power is given to the coils, but there is a ballast resistor that is in between two closely spaced wires to the module for power. This would mean to me that the ignition module is getting either half or double the voltage as per here:
http://moodle.student.cnwl.ac.uk/moodle ... istor.htmlSince I was looking up the OKI MPS-200, I did come across two schematics of an 82 and 83 Honda Goldwing 1100, BOTH which have ballast resistors right on top of the coils.
http://goldwingchrome.com/Manuals/GL110 ... ematic.jpghttp://goldwingchrome.com/Manuals/GL110 ... ematic.jpgWhich I verified on Ebay as well as searching for the OEM measurement, which turns out to be 3 Ohms @10%.
The Goldwings have a different firing order, 1,2 and 3,4. As we know, no one has ever complained about a Goldwing 1100 being a "cold hearted beast" to start like the 650 is. Nevertheless the ballast resistor is there. My suggestion here is that just like my 81 Suzuki, it would seem easier to get a bike started if at least one of each side fired correctly, meaning it would be theoretically possible to get started with 1 or 3 WITH 2 or 4 than our 650s are, this is because the two spark events happen next to each other which has more inertia. But yet the Goldwing and at least one Bronco out there use a low Ohm ballast resistor to "equalize" the voltage across the system, but is bypassed on start for the Bronco, and is "on all the time" for the Goldwing.
I also found another site that would confirm this:
http://www.network54.com/Forum/47850/me ... n-+HOW+TOOBut because of the high-end "Touring" characteristics of the Goldwing, it is confirmed that once the GM 4 prong is used instead of the MPS-200, this ballast resistor can be removed.
So I am wondering now...Maybe the low resistance GM coils NEED this ballast during the start phase. Instead of trying to calculate across a 530 +/- 50 Ohms on your reluctor side, why not put a 3 Ohm Wirewound resistor (or an original Goldwing Ballast) on either the coil side or the module side?
I am still using a second relay trying to "break ground" to the reluctor negatives as two of the early attempts only got the bike to start by breaking ground reference with the white wires.
Instead, I could switch this relay to feed off the coil relay for the through voltage, which would completely skip any corroded ignition/starter switch resistance (which is sending the modules into feedback/infinite dwell) and apply battery voltage minus 3 ohms of ballast to the modules (as appears to be in the Bronco) then when the motor catches and the starter is released, the 3 Ohms of Ballast feed the coils.
So in fact, we isolate both circuits from the white/black ignition wire altogether, and only power the relay with the white/black wire, and use the starter button to interject 3 Ohms of resistance either to the coils AT START or RUNNING (which would begin to use the current limiter in the modules if primary resistance then became 3.2-3.7Ohms) or ONLY for the modules, which if wired similarly to the Bronco, for example, the MODULE would also receive half/double/nominal voltage at start, but then would revert to full voltage (albeit switching the ballast to the coils)
I hope this helps! It makes sense to me in my mind. Please notice that in the diagram of the GM Module on the 1100, he also only uses W being fed by the positive side of the reluctor, but simply grounds both whites. He also does not address the green wire harness ground, and ONLY uses the heatsink bolted to the frame as ground, leaving G unconnected on the module and no harness ground (which is kinda sorta like 'breaking ground')
So then on the four prongs, we can take the white wires of the harness and just jump them to the green wires, but keep them off of the heatsink ground, add a proper ballast of 3 Ohms to seriously increase either the starting voltage, and/or regulate the running voltage (as on the stock Goldwing with the MPS-200 and identical coils).
It would seem easier in theory to me. The Goldwing is going to have a higher output Amperage for all the lights, radio, cb, and accessories, let alone turning an 1100cc motor, as well as a higher CCA battery to start with compared to the CB650...