The IC inside the CDI

[Louis]

OK so the curve in ICP12 is OK with 0- 1/8 throttle

Yes its perfect, but only from onwards 92Hz till (example) 420HZ

Not clear to me right now 

That would probably have something to do with my poor explanation.
Here we have the signal coming from pin 23/24 (in yellow)


It is there from 20Hz till 170Hz, then gone.
It would probably be used somehow in the Oem configuration, but will not be used on our setup.

Can change the value every 500rpm

Ok, than it will only be curve a that i need

The only problem we need to solve is the part from 92Hz and downwards

I can only make steps off 500RPM with my pulse generator, but if use my engine pulse generator can see clearly that everything under 92Hz is not stable and is moving around.
(you can see it clearly with the use off a timing light)

First it thought with the use off a running engine (my own made pulse generator) more signals where given on the Oem CDI that perhaps could cause this jumping around.

But if I use the (digital pulse generator) no other pulses are given on the CDI, it is still the same.
You see the shifting/dancing on the oscilloscope.

Could it be that your firmware generator, has different values/calculations from that 92Hz and downwards?

Or,

Perhaps the long pulse is messing thinks up as the icP12 can't hold that pulse for that time?

[tzr-v4]

Louis,

You have an email...

- ignition signal is 200µS at low state
- curve a
     - 0rpm -> 2000rpm => 19°
     - 2500rpm  => (19+28)/2 = 23°
     - 3000rpm -> 14000rpm => 28°

You should compare for a given rpm
- pickup + OEM pin 22 => measure the time between signal => time equivalent to advance
- pickup + ICP12 pin C0 => measure the time between signal
- OEM pin 22 + ICP12 pin C0 => should be synchronous 

[Louis]

Olivier,

Thanks for the fast firmware postage.
Try to test it this week, and calculate the angular at serval points where it goes wrong with the iCP12 firmware.

Will report back

[Louis]

Got some results,

Think we are getting closer, but some small obstacles need to be taken.
all in the lower RPM

For some reason the firmware starts at about 29 till 30 degrees?

Have made a few setups to get a better idea about this

First setup)
Starting with the firm ware and with the use off the digital pulse generator

Calculated back it is 30,26 degrees (can be seen to with the timing light)


Calculated back it is 27,40 degrees


Second setup
The use off a running engine (oem IC)
Clearly you can see the holding off the 19 degrees generated by the T2333 as a big fat pulse
It is there all the way till 60Hz (as we know there is the point where the oem IC takes over as the ignition curve starts at 1800RPM)


all 19 degrees till 60Hz, here you have a shot off 59,07Hz


And here on a touch over 60Hz, as you can see the oem IC is moving the pulse


Third Setup.
Know this, I combined the Oem Ic pulse and iCP12 pulse to getter
Good to know is the iCP 12 gives a 5V pulse and the oem IC a 4,8V (to know witch is witch)

Here you have Oem en iCP12 (ignition pulse at 29,4 degrees)


as RPM rises it goes slowly towards the OEM IC pulse
(Good to see here are the (fat) pulse generated by the T2333 the small pulse in frond off that is the oem IC pulse and in frond off that pulse is the iCP 12 firmware)


At about 90Hz they start to meld together (Seen in the two pulse heights)


After 90Hz they stay melted and are exactly the same in the complete RPM range (great work Olivier)


As we are getting close, First question is way is the firmware not at 19 degrees?
I belief you start the curve at 19 degrees from 0RPM till 2000RPM?

Can see clearly you have shorten the pulse, think that works great as the shifting around ios almost gone.
We need only to solve the first part

[tzr-v4]

I've just collected from Japan a 3XC-30 CDI to look at and then make it a donor for programmable CDI.

[tzr-v4]

for detail check the 2XT CDI thread.

[Louis]

That looking very nice

Will connect everything back up, so I can start measuring where left it.
(From memory), the problem what's there, is only on the lower rpm and only there when using a running engine (washing machine setup)
When using a digital setup, all what good.

Seems like the 3MA circuit board, is getting a pulse somewhere with changes the signal a bid.
Perhaps the discharging of the capacitor (lower RPM takes more time) messes thinks up.

Will report back on this

[tzr-v4]

Yes I'll try to make an "In the box" loom with these 2 boards.
This setup will be useful for racing with stock looking box 

[Louis]

Just bought a 8CH Oscilloscope, with all probes
Will not be a high tech device, but good enough for measuring the thinks I like

The idea is to measure on the CDI

1) IC ignition signal (oem CDI)
2) IC ignition signal (Icp-12)
3) IC pickup signal (oem CDI)
4) IC pickup signal (Icp-12)
5) IC signal low rpm (oem CDI)
6) Pickup signal (at sensor)
7) Ignition signal (coil)
Free

But then all signals at the same time, it would give me some good data.

Will let you guys know when I have it

[Louis]

Hello Olivier,

When you have time can your read this part on my site?
http://www.tzr3ma.com/8-channel-oscilloscope-measurements.html

Got the 8 channel oscilloscope and have my first measurements.
Now you can see what I mend with under the 3000rpm the icp12 is dancing not holding his place.
All above the 3000rpm is great and it is spot on with the oem curve.

Personally I think something on the CDI it self is cousins this, and my guess for now is the T2333
At low revs the CDI is firing itself at the end off the pickup signal till 1850rpm (19 degrees line)

Think the T2333 is programmed to do so and the rising edge off the pickup is not good till 3000rpm
Will it be possible to program the iCP12 to fire at the back off the pickupsignal till +/-1850rpm and than to the frond?

There is a difference between measuring the data with a running engine or giving a digital pulse.

Here a small movie, you can see clearly that something is enlarging the oem ic pickup pulse and its firing directly after it after it.
https://www.youtube.com/watch?v=U1JWVjhB1rU

To explain
Channel 1) yellow: pickup pulse oem IC
Channel 2) light blue: pickup pulse iCP12
Channel 3) deep purple: iCP12 ignition signal
Channel 4) green: oem ignition signal
Channel 5) light blue: thiristor gate signal

Cheers

[tzr-v4]

Will look when I can take some time.

No bike related some in recent time.

[Louis]

Well guys, specially for this I created a new work place.


Ordered a frequency generator witch should generate the needed pulse for this project.
(as it is possible to draw your own pulse, and update it in the generator)
When I have this generator it will be possible to create a signal like the pickup or a signal needed for the IC

Sadly do to quick/wrong connection I destroyed my test 3MA CDI
So had to replace it with a new one, here we have the new 3MA test CDI. Uncut and connected to the oscilloscope


Measuring for now

1) Thyristor gate (switch for unloading the big capacitor)
2) Pickup signal (pickup from flywheel)
3) Pickup signal on the IC
4) Outgoing ignition signal on the IC


Target is to get it working this year with the Help off Olivier and forum enthusiasts

Will update this soon as I'm focused on this one.

[Louis]

As Olivier is picking up his CDI project I can't stay behind
The ordered pulse generator, if all goes well it will be delivered this week.
Hope this gives me more measuring possibilities.

But I couldn't wait, wanted to tell the problem more precise so other forum members can see/think with me.
So I used my testing machine, here a small explanation off the new setup.


More details can be seen/read here:
http://www.tzr3ma.com/solving-the-problem.html

[tzr-v4]

Louis,

Can you check if the light blue signal built using the pickup sinusoidal wave had always the same duration ?

This square pulse is made by the electronics for the IC.

Olivier.

[Louis]

Can you check if the light blue signal built using the pickup sinusoidal wave had always the same duration ?

Will check this, I suspect it is the same.
But perhaps with the changing at 60Hz it might be different.

This square pulse is made by the electronics for the IC.

Yes is the signal from the pickup lowered by components (mostly resistors) just before entering the T2333

You see it in my ''new setup'' picture
Just before the T2333 you see a soldered red wire that's the point where I measure this.