Maraakate
Veteran Member
Now that the CDI project is more or less complete I think we should move the previous Change Relay discussion here.
Where I'm currently at with things is that the schematic I built partially works. The BJT transistor biasing is incorrect so it "thinks" it's in gear which prevents starting the bike with the starter. However, you can start it with the kickstarter. This also has the side effect that the 15 degree pulse is triggered at idle, instead of 7.5 degrees.
I did some basic testing over lunch today and have discovered that you can disconnect the light blue wire to the CDI entirely and this will disable the 7.5 degree pulse at idle. This opens up a possibility that if you have a bad change relay and cannot source another one then you should be to make a small harness adapter that completes the circuit for the lights and the starter button. Obviously, you will lose the 7.5 degree at idle timing, so it might not be a good idea to leave it idling excessively in Neutral (i.e. starting the bike walking away for 10 whole minutes). You also need to be incredibly careful to not trip the starter button while in gear. But, if you're in a pinch this should help. It also means if you have a faulty change switch (or it's been scavenged for some reason!) then its possible to find the right sized bolt, close off the hole, and still be able to use some part of the bike.
I have gone over the circuit with an EE friend, and he suggested just getting rid of the BJT entirely and going with a MOSFET as these are more reliable and it's what they would have used had they been more readily available back then. The MOSFET I intend to use is overpowered for the application, but cheap. It's the most ubiquitious one available from Digikey, a IRF540NPBF. His suggestion was to use that, and change R1 to a 1K and R2 to 10K and this should work. He said the values of R1 and R2 with the MOSFET aren't completely critical. I.e. could use 10K for R1 and 100K for R2. I'm ordering those parts tomorrow and will wire it up next week and see how it goes.
I'm reattaching the files from the previous thread. The PCB size is not optimized to fit in the side cover. It's large on purpose to make it easy for someone to wire one up on perfboard with point-to-point or scrap wiring. Once the circuit is confirmed working I will shrink it down and it may even be possible to get a SMD version so small that it could be a simple harness plug in that could float in the side cover. I would have to check pricing on pre-assy from China for this and/or finally learn how to SMD solder (it's been on my TODO list for 20 years).
Where I'm currently at with things is that the schematic I built partially works. The BJT transistor biasing is incorrect so it "thinks" it's in gear which prevents starting the bike with the starter. However, you can start it with the kickstarter. This also has the side effect that the 15 degree pulse is triggered at idle, instead of 7.5 degrees.
I did some basic testing over lunch today and have discovered that you can disconnect the light blue wire to the CDI entirely and this will disable the 7.5 degree pulse at idle. This opens up a possibility that if you have a bad change relay and cannot source another one then you should be to make a small harness adapter that completes the circuit for the lights and the starter button. Obviously, you will lose the 7.5 degree at idle timing, so it might not be a good idea to leave it idling excessively in Neutral (i.e. starting the bike walking away for 10 whole minutes). You also need to be incredibly careful to not trip the starter button while in gear. But, if you're in a pinch this should help. It also means if you have a faulty change switch (or it's been scavenged for some reason!) then its possible to find the right sized bolt, close off the hole, and still be able to use some part of the bike.
I have gone over the circuit with an EE friend, and he suggested just getting rid of the BJT entirely and going with a MOSFET as these are more reliable and it's what they would have used had they been more readily available back then. The MOSFET I intend to use is overpowered for the application, but cheap. It's the most ubiquitious one available from Digikey, a IRF540NPBF. His suggestion was to use that, and change R1 to a 1K and R2 to 10K and this should work. He said the values of R1 and R2 with the MOSFET aren't completely critical. I.e. could use 10K for R1 and 100K for R2. I'm ordering those parts tomorrow and will wire it up next week and see how it goes.
I'm reattaching the files from the previous thread. The PCB size is not optimized to fit in the side cover. It's large on purpose to make it easy for someone to wire one up on perfboard with point-to-point or scrap wiring. Once the circuit is confirmed working I will shrink it down and it may even be possible to get a SMD version so small that it could be a simple harness plug in that could float in the side cover. I would have to check pricing on pre-assy from China for this and/or finally learn how to SMD solder (it's been on my TODO list for 20 years).
