Digital programmable CDI ignition for 2-stroke engines

[bob weir]

june 27
this post is a follow up to the 'may 12' / june 12 post. and over the next several posts how the testing was done will be shown. that is , how the signals board circuit with its pulse generator input signal & opto-isolator was connected to the CDI ckt bd. and a spark plug.

the june 5 post referred to the p.8 7-24-16 post for its sketches. one will be repeated here. below.

this 7-24-16 circuit layout is the starting point . its a point along the evolution of a way to connect the 2 PCB's together. and since then some of its errors have been discovered
and changes made. today they are being listed. and later will be posted a re-done sketch
of the changes that represent how it looks now.

1. the date said 7-24-31 and should have been 7-24-16
2. the opto p/n is shown as FOD867 .... its really a FOD817
3. the pin no's for the opto were left off
4. the red ink pulse sizes suggest the last 3 are the same size, about 10V.
........... not so. the one at U1:2 is only about 2V. yet this is the logic '1'
this IC needs to drive Q1 to be 'ON' .
............ it is good to take a look at the U1 37322 data sheet table 7,5 to see
that only 2V is needed for an input pulse to be taken as a logic '1' .

5. that vertical line showing where the signals board ends and the CDI begins
............. has the opto on the CDI board. not so. its now on the signals PCB.

on looking at just the opto IC output side we see that 3 connections are
needed to be made to join this 'signals PCB' to the CDI PCB. one for +12V ,
the opto emitter output pulse and the opto's ground .

some thought will lead one to see these 3 connections can be made with
jumper wires to the U5 DIP socket with its 4047 removed.

we are fortunate here in that U5 is not necessary for the case of having an
external pulse signal make the spark fire.

by wiring the 2 PCB's together this way i was able to get a spark with
every push of the 'manual' button on the pulse generator. and doing this
simulates the flywheel signal causing the spark plug to fire.

so then what is left to post is :
1. a new sketch of how the 7-24-16 circuit layout looks now.
2. and a sketch of the overall testing method
3. and some scope signals to add what they can to the picture of what is happening

 

[bob weir]

july 14
the first EasyEDA PCB order came. late june. and none of the boards had traces of copper
on them to connect the parts. so i wrote to Nicole and she asked for the Gerber files to be re-submitted.
that was done. and hardly a week later a new set of 10 boards arrived that had the expected traces. they sure look good. and the price is unreal.

first thing was to 'ohm' point-to-point on the PCB for continuity according to the schematic drawing.
that checked out. so the voltage regulator section parts were soldered on along with the
power-on LED1. it lit up fine and the voltage was 4.95

next the rest of the parts will be added. and the 508 uC coded with a 1,2,4,8,16,32, 64,128 ...
to appear sequentially on the 8 leds.
i.e. just one led 'on' at a time.

this effort was a get-started with learning the E-E s/w with a simple circuit with which
i was already very familiar. a training exercise if you will. with as much known about
the circuit as could be. to not detract from the major s/w learning curve effort.

many beginner mistakes were made. chief among them was not changing any of the
s/w default sizes. thus the signal traces are narrow. and the pads small. too small .
and this makes for difficult soldering.
the solder tip had to be filed to shape it from a 1/8" chisel into more of a 1/16" point.

the narrow Vcc trace could be too small for when most or all 8 leds need to be 'on' .
if so then 2 wires will be soldered from the U3 5V reg to each of U1 and U2

also learned that more care is needed in label placement. or else the mounted part can
eclipse the label. another goof was wrong part size for the cap' lead holes. thus all of
the caps need their leads bent for a narrow hole spacing fit

and you can see that the corner mounting holes for standoffs were forgotten
with the 2'd PCB

the 2-pin female headers ( ex. P9 , P10 ) all fit well. so did the 14- and 8-pin DIP sockets.

overall though i am pleased to have them as - is. and to be on the learning curve.
and with many new ideas of how to do better next time

 

[bob weir]

july 16
good news to report. when the july 14 PCB was fully populated and powered it ran as expected.
the 508 uC code was some old test code . it counts 0,1,2,3,... 255. and sends the byte out
the DI and CLK pins seriallly. this code was tried out with the 508-2 uC chip you
see in the foto of this new PCB circuit. and it ran fine. the leds all lit evenly ...1 or all 8.

the foto below is with the leds being lit by the on-board uC 508. the purpose is just to
test that the 74164 is operating correctly. by taking 8 input bits and displaying them
in parallel with leds . here is seen 5F hex

the led along the top right edge of the PCB is a power-on indicator.

next the PCB will be tested with DI, CLK and GND from a separate uC ckt.
this will check that it runs as designed. and shows a byte from one
of the memory or special function registers in this separate/remote uC

note : when remote DI and CLK signals come onto this PCB the onboard
uC 508 socket will be empty. this will eliminate any pin/signal conflict
between the 2 uC's.

note : the external incoming DI and CLK signals can be from any uC,
not limited to a 12F508. for instance when work resumes on the
separate Sport Devices code project the DI and CLK will come
from a 16F84A

for those of you who are new or have forgot what good this ckt is
you can read about it on earlier posts. see list below ...



on looking back thru the pages of this project here are some of what was found
of the 74164 in use :
p. 7 4-29-16 ckt sketch w/ data table
p.8 7-24-16 ckt sketch
p. 10 3-7-17 ckt sketch
p. 10 3-14-17 foto 6"x3" bread board w/ 74164-led circuit
p. 10 3-15-17 ckt sketch

and i think code too was posted somewhere. that shows how a 508 uC sends a byte out
in serial fashion to the 74164.

 

[bob weir]

july 16
a mistake on the july 14 post was to call P9 and P10 ( P11 too ) a "... female header" ... if you go onto ebay and look for them there
the keywords are : " 2 pin 2 way 2.54mm 0.1" pitch screw terminal block connector "

expect to pay 20 cents each depending on quantity.

 

[JakeStew]

my attitude of trying to keep this post as 'clean' as possible comes from observing
a similar ignition effort over at RCGroups.com . its a mess. with little to no effort
to clarify what is or has been happening. its an example of what doesnt work

No doubt. You probably didn't even find the good stuff in that thread. Hope you checked out my code at least.

 

[bob weir]

july 24
fresh news of the 2'd PCB order with EasyEDA. new to arrive are 10 boards of the 'signals' type.
they look good. they were last mentioned on p. 12 june 5




ATTACH=full]35360[/ATTACH]

 

[bob weir]

july 24
this 'signals' board works with the main CDI ignition board. it accepts the flywheel timing
signals, crunches them and with the 508 uC delivers a signal to the CDI board
that causes it to make the plug spark

here in the foto they both are side by side.

the CDI board is the 2012 version via a purchase from dr alfrey.
it is being used as a full size reference PCB to try and copy with the EasyEDA s/w .

eventually the 2012 circuit will be edited to remove the parts not needed
for an ignition application . then the 'signals' circuit parts added to it
so that just one PCB does it all ... makes my motorcycle run again

 

[bob weir]

july 24
an error was the mounting holes being too small for the standoffs that will join the 2 PCBs
so with a metric mike the thread diameter was found , about 2.74 mm . then
a 3mm drill used to enlarge the holes

 

[bob weir]

july 24
here the 2 boards are joined as they will be for the immediate future.

already one problem is seen. in the fit. C5 sticks up high and the signals board will
need its corner trimmed. or possibly just add more standoff pieces so the 2 PCBs
are further apart.

more problems may turn up during the bare signals PCB testing. before
being populated and then later during the operational testing.

following that will be yet more testing to see how they play together.
there is some 508 coding yet to do to decrease the RPM range size

for those of you with more time and energy than i have can charge ahead
and have a finished ignition circuit & board for your motorcycle.
everything you need has been posted here already.
you will have to adapt what you've found here to your particular engine
and its timing diagram requirements. very do-able
give a hollar at this thread if you have my same yamaha 225 engine & get stuck.
or write me via [email protected] if your engine is something else