FBG47 Rev. 1.2 - AC-CDI Unit for Honda CX500

Okay, finally I got time to concentrate to my replacement ac-cdi unit aka FinnBoxG47 (shortly FBG47) which I made for my old Honda CX500. I thought could I make it better and started to re-designing the electric board. There are a few points which can be made better, I think. Last summer everything worked just fine but you need to do something on these long winter nights here in Finland. And sure, if I can improve even a little bit the present unit, why not?

It's not a secret that the analog cdi unit of CX500 is not in principle very complicated. There is just a few points in the schematic which need to be understood correctly. At the wiki-pages of cx500 forum is shown a few schematics which are almost correct.
All begins from the used thyristors. At 25C and 100C thyristors gates have a little bit different the lowest trigger voltage which is enough to open the current flow from anode to cathode. In BT151-800R case, for example, the normalized voltages are about 1V (25C) and 0.7V (100C). The difference is not much but in the present circuit enough to change the ignition advance curve at different temperatures. To design the back effect for that phenomenon requires usage of NTC thermistors at pull down resistances of the gates. Also the advance lines resistances need to be matched with the pull downs. In analog cdi unit the pick-ups and advance pulsers produce analog signals, ie. the signals heights are dependent on the speed of the engine. More revs, higher the signal heights. In CX500 case the idling is at 1100 rpm and the advance start to change at about 2000 rpm. The max. advance is at 5500...6000 rpm. The slope of advance curve can be easily changed by changing the resistor values of the advance lines. But the original Honda kind of slope is good to maintain. To summarize all this is as follows:

* Calculate the needed pull down resistances based on the used thyristors' trigger voltages at 25C/100C.
* Match the resistance by using suitable thermistors (B?) in parallel and in series (R9 and R10) with 1% resistors.
* Calculate/test the suitable resistors for the advance lines (R2, R3, R4, R7, R7 and R8, left and right lines are not the same!).
* Match the negative pulse resistors (R1 and R5)

Well, it's easy to say afterwards.. Anyway, lately I added a signal diode and a capacitor to both of the advance lines simply to smooth and balance even more. I also changed the thermistors to more accurate ones (AVX). Have to test ride my cx when possible to see is there any difference. Hardly much but now I think FBG47 is in the final form..