Electronic ignition for a small 2 stroke engine

A

Andres Rodriguez

New Member
I am helping a High School student on his senior project and I need help to help him.
The project was defined by the student, approved by the teacher, and he asked me to be his designated mentor (poor kid). He needs to run a comparison analysis of a small internal combustion engine running on three different fuels and gather data on power output, exhaust temperature, and fuel consumption. He found a circuit for electronic ignition and added a Picaxe 20M. Attachment Elec Ignition 5V.pdf is the original design, it works but the coil was getting very hot. The specs are not very clear and we think it is rated for 3V.
The question is whether or not the circuit in attachment Elec Ignition 3V.pdf will work
 

Attachments

L

lanternfish

Senior Member
I'll jump in here

Without the program it is difficult to get a full appreciation of what is happening.

For instance, is the trigger pulse for the coil too long, therefore allowing the coil to heat up?

Is this a standard 2-stroke coil? What is its part number?

These may help the more knowledgable PICAXErs to assist.

Cheers
 
M

mega

New Member
My preference would be for an IGBT designed for a coil-on-plug operation and a small motorcycle ignition coil, most would operate on 12volts.

I like the circuit though....
 
I

inglewoodpete

Senior Member
If the 5v version works, then I can't soo why the 3v one wouldn't.

You've nothing much to lose: give it a try. The spark will either be good enough, or it won't. If it does work, the coil should run a little cooler.
 
B

BeanieBots

Moderator
Be aware that the output stage has a double inversion.
It might be that you are pulsing the output pin with the wrong polarity.
(it should idle HIGH).
The spark generated will be pretty much the same with either polarity but one method will dissipate MUCH more heat in the coil (and drive transistor) than the other.
 
A

Andres Rodriguez

New Member
The reflecting surface for the IR sensor set at 15 degrees before TDC and the code running on the 20M is as follows:
Code: 
#picaxe 20m

Symbol Points = Output0
Symbol IR = Input1
High 2
Main:
setint %00000010,%00000010
goto Main
Interrupt:
Low Points
Low 1
Pause 5
High Points
High 1
setint %00000010,%00000010
return
The next step will be to try to adjust the dwell time and the timing with the two pots (R1 and R10). Obviously, the reflecting surface will need to be moved more than 15 degrees before TDC. I am letting the student suffer through the maths.
Code: 
#picaxe 20m

Symbol Timer = W1
Symbol Dwell = W0

Symbol Points = Output0
Symbol IR = Input1
High 2
Main:
setint %00000010,%00000010
readadc10 7,Dwell
Dwell=Dwell*10/1023+3
readadc10 2,Timer
Timer=Timer*10/1023+5
goto Main
Interrupt:
Pause Timer
Low Points
Low 1
Pause Dwell
High Points
High 1
setint %00000010,%00000010
return
 
B

BeanieBots

Moderator
What's on output 1?
What's on output 2?
Presumably "points" is the output used to trigger the coil???

Maybe use pulsout instead of high/low.
No need to keep re-initialising the interrupt in the main loop.
Just one setint outside the main loop. (and in the interrupt where you have it).

EDIT:
You don't have a High points before the main loop.
If it IS the trigger for coil, that could explain the overheating.
When the PICAXE starts, the coil will be energised and could never be switched off if no interrupts arrive.
 
Last edited:
A

Andres Rodriguez

New Member
BeanieBots said:
What's on output 1?
What's on output 2?
Presumably "points" is the output used to trigger the coil???

Maybe use pulsout instead of high/low.
No need to keep re-initialising the interrupt in the main loop.
Just one setint outside the main loop. (and in the interrupt where you have it).

EDIT:
You don't have a High points before the main loop.
If it IS the trigger for coil, that could explain the overheating.
When the PICAXE starts, the coil will be energised and could never be switched off if no interrupts arrive.
Click to expand...
BB thank you for all your input.
I will change the code as you suggest and use pulsout
Points is used to trigger the coil
Output 1 is for an LED that is on when the coil is triggered
 
Last edited:
L

LizzieB

Senior Member
Whoa! Everyone who is suggesting that he can just hit the coil with a low voltage pulse (IOW have the drive transistor off and pulse it on) and get a decent spark needs to step back and think how the Kettering ignition system works. The system uses energy stored in the coil, its normal sequence is for the voltage to the coil to be on and only turned off for short periods to trigger a spark. The coil is inductive, the current rises slowly when the voltage is applied so the on (dwell) time needs to be maximized to have a good spark. Normally on, off for a spark is the way to go.

There are electronic ignitions that do use the coil as a pulse transformer like the (US) MSD system, but those hit the coil with a series of pulses something like 500 volts to overcome the inductance and get a fast enough rise in current.

I doubt you will see any real difference in short term performance just replacing points switching with transistor switching, all you can really optimize is the dwell time. Capacitor discharge is probably the simplest electronic option that has the potential to show some quantifiable improvement.
 
P

papaof2

Senior Member
LizzieB said:
I doubt you will see any real difference in short term performance just replacing points switching with transistor switching, all you can really optimize is the dwell time. Capacitor discharge is probably the simplest electronic option that has the potential to show some quantifiable improvement.
Click to expand...
Capacitive Discharge (CD) ignition systems typically charge a fairly large capacitor (10's of uF) capacitor to a relatively high voltage (300-350) and dump the capacitor's charge to the coil primary. The best improvement I've seen was on a 4 cylinder overhaed cam engine. The idle speed increased more than 200rpm when the CD system was installed - with no other changes made.

John
 
L

LizzieB

Senior Member
I just realized that output transistor is a Darlington (not shown as such on the schematic), so you need to pay attention to the Collector–Emitter Saturation Voltage especially when using a 3 volt supply. The spec is between 2 and 3 volts maximum, most likely the one that you have will be better than that but it will most likely need a good heatsink.
 
C

chipset

Senior Member
first go to MSEFI.com its a DIY opensource programmable fuel injection system. Read up on the ignition stuff and also look at the scematics for the coil drivers. They use IGBT's to directly run the coils. I think your issue with the coil getting hot is an issue with the Dwell as was brought up earlier. If dwell is set too long the coil will overheat and probably fail after a while (probably when its important for it to function, like during his presentation). Depending on the focus of his project, Id steer him away from making his own electronic ignition. To build a good unit that doesnt induce any variables into data collection would be a senior project in itself. An MSD 6AL can be had for fairly cheap nowdays and is dead reliable and does improve performance.
 
You must log in or register to reply here.
Top