The IC inside the CDI

[Louis]

I'm not sure if we could easily read the IC. The first step is to try and get a datasheet for the IC (or similar IC) and create a logical schematic of the printed circuit board to identify which pins are receiving which input/output. If it is an eeprom then you might be able to get a dump from it but I'm a bit hazy on this.

I'm looking following lines now for a couple off hours. Tying to find some logical in it


This will probably not be the way, but this kept my busy for a wile.
Most logical I thought can't measure the signals (jet) but the ic will need a mass and a working votage.
So there I'm looking fore, this IC has it's pin nr correct (pin 21 and 42)

Only the ex and x are not connected. (don't know if this is logical)
http://datasheets.chipdb.org/Fujitsu/MB8840.pdf#search='fujitsu+42pin'

Now I'm thinking of connecting the cdi and measure on the pin's where voltage is found.
Will this be a logical step?

What I think you are trying to get to is an understanding of the 3 dimensional ignition advance map. This might be easier to get by treating the CDI as a "black box" and feeding it a range of inputs and correlating the outputs. I think this is what Borat  (?) does.

That I'm already busy with
Some forum members will already know this as I have send them a question about which electro motor I could use best.

Now I have found a electro motor which hopefully will do the job.
And there quit cheap to, have ordered a washing machine engine (second hands of course)
The engines are capable off doing 13000rpm and will have the power to turn the flywheel to.
Sadly the engine did not arrived jet (otherwise I could post a picture)

Will try to make it like a real engine so stator and rotor pickup will be mounted.

[Paul]

I'm watching this thread with interest. We have stripped engines, carbs and suspension, is this the last outpost? Stripping and understanding the ignition box? I wish you clever lads luck. Louis, washing machine, oh my Lordy! Interesting beyond belief.

[Louis]

Stripping and understanding the ignition box

Hopefully in the end we do

Louis, washing machine, oh my Lordy! Interesting beyond belief.

Yes I know it sounds funny , but they are quit strong engines witch are controllable
http://www.youtube.com/watch?v=BZolEeqbWz0

[Louis]

Just did a quick measurement.


This is what I found..(Measured ignition switched on)
Like the results off pin21 and pin42

[yanw]

The datasheet says that unused pins have to be tied up/down with a resistor. If you can see on the PCB (assuming it is not multi-layered) which pins go to just a resistor and then ground or VSS then you will be able to eliminate them from the equation.

[Louis]

Ehhhhh.

The datasheet says that unused pins have to be tied up/down with a resistor

This means that every pin not in use should have a resistor somewhere mounted on the circuitboard?
The problem is, there also going some lines under the IC that I can't follow.

(assuming it is not multi-layered)

Don't know what you mean with multi layered?
The lines on the circuit board are different between top and bottom.
(the pattern on top does not mach the pattern bottom side)

If you can see on the PCB which pins go to just a resistor and then ground or VSS then you will be able to eliminate them from the equation.

Meaning the ones that read 0V should have a resistor?
That will be difficult, what I can see for now is:

-ex not connected
-x not connected
-reset not connected
-irq (doubt if its connected)
-sc/to (not connected)
-tc (not connected)
-p0 (seems not connected)

Does this make any sense

[Toop]

+0v may be a channel signal 
(this is a common thing on CAN-bus system for example )

[Louis]

Thanks for cheering me up

Looks like thinks will be much complicated then I thought.
But that will be the challenge
Already know more from the cdi then I did a week ago, so we ar going to right way

[Louis]

Hello Guys,

After a good night sleep, came up with the next thought

If Yamaha has placed a order to make the (in this case) 3MA CDI's
The developers worked with ic's that existed at that time.

So looking fore a Fujitsu IC's build in 1989 will not be the case, as I doubt they will developed a ic specially for the 3ma cdi's

Also a bike that is coming on the marked at 1989 will be made/tested in begin 1988

So in the seach for a IC that could be used in the 3MA cdi, i have looked to IC's that where on the marked before 1989

And then it becomes more logical to me.

IC 88400
http://pdf.datasheetarchive.com/indexerfiles/Scans-059/DSA2IH0049322.pdf

IC 88500
http://pdf.datasheetarchive.com/indexerfiles/Scans-007/Scans-00141501.pdf

IC 88530
http://pdf.datasheetarchive.com/indexerfiles/Scans-059/DSA2IH0049323.pdf

In short terms, the 88000 onwards series could be the one used.
Specially the 400 and 500 series are close to each other
The 530 is a bid different, how can we emulate this type


Now how does this idea sounds to the ones who are more into the IC's them me.
Personally think I'm close.

[tzr-v4]

Hi Louis,

You should list what are the input/ouput of the CDI unit from Yamaha :
- TPS = input / analog input should be 0-5v signal = on pin ?
- RPM signal = input / digital imput could be 0-5v rising edge or falling edge or pulse
- Spark signal = output / digital could be signal going from 0->5V or 5->0V one
- airjet solenoid = output / digital could be signal going from 0->5V or 5->0V one
...

We can imagine replace the ship by a PIC µC board like this one :
http://www.tiertex.com/index.php?route=product/product&product_id=51

[Louis]

Hi Olivier,

That is exactly what I'm planning, only I don't have a device that can generate a rotating engine.
So I need to build that.
(currently doing as it turns out the washing machine engine was delivered at the neighbors)

More info:
http://www.tzr3ma.com/start-build.html

What I'm trying to solve for now is getting some base line.
If the IC has some fixed measuring points witch I can already measure.
(without the need of a running engine)

like

-VSS
-VCC

What ells do have a steady Voltage measuring point?
-SO?
-SI?
-IRQ?
-EX?
-X?

If for example IRQ needs to be 4,98V to work (Or delivers 4,98V when ic is working), and I don't measured it on that pin number off the IC (sheet). Then I can remove that IC from the possibilities.

[Louis]

Just looking more into this, have to say the idea is super

If I understand it right, this could replace the whole IC With a programmable?


There are more types

[Louis]

Need to let it rest a bid, as the more I read. The closer I think I am

@yanw and Olivier (and the rest who is following this)

Have put the measured values to the (what I think) the possible IC
(used copy/past so its not completely clear)


Now with the description what the values should be


It looks more logical to me
(My head hurts)

[yanw]

By multi layered I mean that there is not only tracks on the top and bottom of the board but also on other layers sandwiched in-between. I suspect that any pin at 4.96V is "tied up" and unused (don't quote me etc).

From the quick scan of the datasheet I think these ICs have a writeable memory that allows you to "put a program on it" (that's the 2K program memory on page 2). If they are like the PIC then you should also be able to pull this program off. It will be in HEX but you can use the datasheet to reverse that into the program code to map what it is doing. Youy will probably need a development board and the SDK to do this.

I suspect the programming was either done by Yamaha or contractors in-house to Yamaha.

I will try to have a better look at the datasheet over the next few days but I suspect they feed the flywheel timing signal into the timer/overflow on the chip and use the throttle etc settings to calculate the advance and then send the ignition pulse to the coils/turn the powervalves.

[Louis]

You make it sound that easy. wow

This is the sheet I used to drop my voltage readings in.
Hopefully its the on sitting in the CDI.
http://pdf.datasheetarchive.com/indexerfiles/Scans-059/DSA2IH0049322.pdf

Thanks for the afford you already put into this

Louis