Need your input on new ignition "brain" design

[Jcary85]

So, here's the current schematic for this. I'm struggling with two things currently:
1. The latching start/stop circuit is not working in practice. There's a bunch of noise causing inconsistent on/off latching. I tried adding a bunch of diode protection before the raw 12v gets to the BJTs but it didn't help. Also added a varistor. didn't help. This is going to be a tough one to solve. I'm debating doing it in software as I can "filter" much easier there.
2. The ECU won't provide any spark unless there is something (a USB-TTL or HC06 bluetooth module) connected to TX/BT_RX/GND (5v is not connected). This one really has me stumped. There's not a whole lot going on on those lines except for a simple voltage divider to drop the BT rx signal to 3.3v. Makes no sense at all

If anyone has ideas, I'm all ears.

Vent:
I'm at a point in this project where I'm closing to giving up. I have so many hours into this its insane. I'm not an electrical engineer but I really wish I was. I'm confident I'll get this solved, I just might need to walk away from the project for a bit.

 

[eric_f]

1. The latching start/stop circuit is not working in practice. There's a bunch of noise causing inconsistent on/off latching. I tried adding a bunch of diode protection before the raw 12v gets to the BJTs but it didn't help. Also added a varistor. didn't help. This is going to be a tough one to solve. I'm debating doing it in software as I can "filter" much easier there.

Am I correct that the "Start" and "Stop" are the buttons on the handlebars in the latching circuit?
If I recall, the start button provides +12v and the stop provides ground, is that correct?

In the schematic, it looks like both transistors are going to try and turn on at the same time. 12V > R26 > R28 > base AND 12V > R27 > R29 base. Which one will be enabled first? I don't think it will be consistent and they may fight. I think this is the culprit of your latching not working correctly.

This may be a better option. As shown, both switches must go to ground; S1 is Start and S2 is Stop. If I was correct and the start button provides 12V, then rewire S1 so instead of being at the junction of R2 and T2 & GND, have it between T1 and R1 and 12V. Then it can't be a dead-short to ground if both switches are pressed.

T1 is your power MOSFET and then you only need one transistor and two resistors, almost half the component count.

 

[eric_f]

2. The ECU won't provide any spark unless there is something (a USB-TTL or HC06 bluetooth module) connected to TX/BT_RX/GND (5v is not connected). This one really has me stumped. There's not a whole lot going on on those lines except for a simple voltage divider to drop the BT rx signal to 3.3v. Makes no sense at all

Does it boot up and function but not provide any spark, or does it not fully boot up?
I wonder if the bootup sequence is trying to initialize the serial connection and it's failing, so it's getting stuck.
Alternatively, does disconnecting the voltage divider help? What if you disconnect the voltage divider and try a pull-up resistor?

 

[Jcary85]

Does it boot up and function but not provide any spark, or does it not fully boot up?
I wonder if the bootup sequence is trying to initialize the serial connection and it's failing, so it's getting stuck.
Alternatively, does disconnecting the voltage divider help? What if you disconnect the voltage divider and try a pull-up resistor?

I haven’t debugged this at all yet. It’s not waiting for a connection though - simply plugging in a usb/TTL converter (not even hooked to a laptop) lets it run.

 

[Jcary85]

Am I correct that the "Start" and "Stop" are the buttons on the handlebars in the latching circuit?
If I recall, the start button provides +12v and the stop provides ground, is that correct?

In the schematic, it looks like both transistors are going to try and turn on at the same time. 12V > R26 > R28 > base AND 12V > R27 > R29 base. Which one will be enabled first? I don't think it will be consistent and they may fight. I think this is the culprit of your latching not working correctly.

This may be a better option. As shown, both switches must go to ground; S1 is Start and S2 is Stop. If I was correct and the start button provides 12V, then rewire S1 so instead of being at the junction of R2 and T2 & GND, have it between T1 and R1 and 12V. Then it can't be a dead-short to ground if both switches are pressed.

T1 is your power MOSFET and then you only need one transistor and two resistors, almost half the component count.
View attachment 436276

Nice I’ll simulate this later today. My circuit uses ground for both start and stop. I actually think there might be less noise if I used 12v like the factory did. Thanks for the help though. This might work better

 

[RiverRides]

So, here's the current schematic for this. I'm struggling with two things currently:
1. The latching start/stop circuit is not working in practice. There's a bunch of noise causing inconsistent on/off latching. I tried adding a bunch of diode protection before the raw 12v gets to the BJTs but it didn't help. Also added a varistor. didn't help. This is going to be a tough one to solve. I'm debating doing it in software as I can "filter" much easier there.
2. The ECU won't provide any spark unless there is something (a USB-TTL or HC06 bluetooth module) connected to TX/BT_RX/GND (5v is not connected). This one really has me stumped. There's not a whole lot going on on those lines except for a simple voltage divider to drop the BT rx signal to 3.3v. Makes no sense at all

If anyone has ideas, I'm all ears.

Vent:
I'm at a point in this project where I'm closing to giving up. I have so many hours into this its insane. I'm not an electrical engineer but I really wish I was. I'm confident I'll get this solved, I just might need to walk away from the project for a bit.

Electrolytic capacitors can be used to filter between your DC positive to ground where it's needed to remove ac noise and clean it up. Place them wherever u think you may need them and test with your oscilloscope to see if you've eliminated the noise. Id still use your diodes. On your schematic I drew them in as a red line for an example but put them wherever you want and what farad size u need.

 

[Jcary85]

Electrolytic capacitors can be used to filter between your DC positive to ground where it's needed to remove ac noise and clean it up. Place them wherever u think you may need them and test with your oscilloscope to see if you've eliminated the noise. Id still use your diodes. On your schematic I drew them in as a red line for an example but put them wherever you want and what farad size u need.View attachment 436286

awesome! thats an easy thing to try. I'll give it a shot and keep you posted. Thanks!

 

[RiverRides]

awesome! thats an easy thing to try. I'll give it a shot and keep you posted. Thanks!

Please do!

 

[eric_f]

Nice I’ll simulate this later today. My circuit uses ground for both start and stop. I actually think there might be less noise if I used 12v like the factory did. Thanks for the help though. This might work better

The example I provided uses ground for both buttons but changing the start to 12V is easy. My thought is that using the factory set up as much as possible means fewer mods and therefore less likely to be installed incorrectly.

I haven’t debugged this at all yet. It’s not waiting for a connection though - simply plugging in a usb/TTL converter (not even hooked to a laptop) lets it run.

Sounds like it's pulling the pin up. Maybe try adding a pull-up resistor to the TX line?

 

[Quinc]

@witrebel

 

[witrebel]

Nice I’ll simulate this later today. My circuit uses ground for both start and stop. I actually think there might be less noise if I used 12v like the factory did. Thanks for the help though. This might work better

I am going to be testing bistable latching relays this weekend to adapt kawasaki trim buttons to a bilge pump. Press up, latch on, press down, latch off. I think it might work for you on the 12v supply you are trying to latch as well? Not sure on vibration resistance, but should be a lot more immune to noise.

https://www.digikey.com/en/products/detail/panasonic-electric-works/DE1A1B-L2-12V/647198

 

[Jcary85]

I am going to be testing bistable latching relays this weekend to adapt kawasaki trim buttons to a bilge pump. Press up, latch on, press down, latch off. I think it might work for you on the 12v supply you are trying to latch as well? Not sure on vibration resistance, but should be a lot more immune to noise.

https://www.digikey.com/en/products/detail/panasonic-electric-works/DE1A1B-L2-12V/647198

I was originally using TE bistable latching relays. I switched because I think it’s a mistake to use anything mechanical in this application. Even the current shifty module has issues with hard flat landers sometimes.

 

[witrebel]

I was originally using TE bistable latching relays. I switched because I think it’s a mistake to use anything mechanical in this application. Even the current shifty module has issues with hard flat landers sometimes.

Yeah this makes sense, in my application I can always just hold down the bilge button. It would be a much bigger deal if the vibration killed your ignition. If the electrolytic caps don't do the trick let us know, I'd be happy to work on a solid state latch circuit.

 

[Jcary85]

Yeah this makes sense, in my application I can always just hold down the bilge button. It would be a much bigger deal if the vibration killed your ignition. If the electrolytic caps don't do the trick let us know, I'd be happy to work on a solid state latch circuit.

I played around with the caps quite a bit. Seemed like it maybe helped but didn't quite solve it. I've walked away from it for the past week or so. Was getting frustrated with the project. I was debating moving the latching circuit after the 5v regulator. This would mean you need to disconnect your battery at the end of the day but would remove all the noise for sure.

 

[Jcary85]

made some great progress yesterday. Hooking up a decent oscilloscope really helped. One simple well placed zener diode on the start input dropped most of the noise and the latching circuit seems to work reliably now. Hoping I'll have time to hook it up to a ski tonight and test run it.

 

[witrebel]

Hows this going? Hopefully you're enjoying the summer and getting some tray time on your prototype!

 

[Jcary85]

Hows this going? Hopefully you're enjoying the summer and getting some tray time on your prototype!

Going well. Running well in my DVX. Just got next rev in from china. Having an issue with crank signal but it’s probably something simple. I’m hoping I solved the slow starting issue by getting rid of the boot loader.

 

[Jcary85]

Got the new revision running. No hardware changes needed! I was so focused on the crank signal circuit that I forgot I had to populate a capacitor on the 5v rail which is through hole so china factory doesn’t install it. Noisy 5v signal was the problem. Without the boot loader it starts super fast now as well! Water test coming soon!