Digital ignition timing

[Anders Dinsen]

Back in the old days, tuning ignition timing was a matter of replacing weights and springs inside the distributor and by trial and error on the road assisted by a strobe light, determinng the optimum advacnce curve. When our dear Murena's came out, the there was specialised equipment available for mechanics to test the advance mechanism of distributors to ensure they were within specs. Today, of course, digital is the way forward and you can buy aftermarket distributors with bluetooth interfaces having replaced the weights. Tuning with these devices is best done on a rolling road. The potential is great!

But even in 1981, when the first 2.2 engines were fitted to the Murena with mechanical Bosch and Ducellier distributors, there were some electronic tools available, and these were in fact used on the Murena. We know, since the 2.2 was fitted with a crank sensor connected to the diagnostic connector. It was there for diagnostic purposes as the picture from my "Nachtrag der bedienungsanleitung" for the 2.2 shows. Note the "Geber für Zündzeitpunkteinstellung" and the "Diagnoseanschluss"!

I started wondering, a few years ago, about this sensor. How was it used? Noone I have talked to has seen a device that uses is. So was this just a meaningless item on a feature list? Perhaps! Yet, I kept thinking that it should be quite easy to get a signal from the crank sensor and compare the timing of it with the timing of the ignition signal, measuring RPM and advance.

Well, "Quite easy" turned out to be quite a project.

I have a fully working prototype now and will use this thread to document the way it works and how you can build one yourself. The picture is from when I was debugging my design in my lab. I will post pictures later of it working on my car and details about the design.

[Anders Dinsen]

So here's a bit more info about my device. The picture attached shows my #2 prototype board, the one I'm currently using on my car.

As can be seen, all components are leaded so it does not require any special tools to build. The LED 7-segment displays are fitted on the rear of the PCB. Essentially all components are modern deriviatives of 1980's era components.

The big black device in the middle is an 8052 microcontroller. This can be programmed through the connector seen on the left side. The lower part of the board contains the analogue circuits resposnible for detecting the signals from ingition and crank sensor. To the right of the microcontroller is a circuit that does some basic filtering of the signals before supplying them to the microcontroller for actual measurements and calculations. There is a voltage regulator on the board responsible for the 5V supply to the system and a separate one supplying the 9V supply which is used to generate a permanent magnetic field in the crank sensor coil. Since the crank sensor is unmagnetic, it needs a constant DC current flow to generate a small amount of magnetic flux. When a notch in the flywheel passes by, the flux is distorted and an additional current is generated, and this additonal current is detected in the analogue circuit. A fuse and protection diodes protect the device from voltage spikes and reverse polarity. The top part of the board (unpopulated) contains an isolated RS232 interface which I'm still working on.

As mentioned, the basic idea is to measure the time between the ignition and the crank sensor triggering, divide that by the time between crank or ignition signals (1/RPM effectively), and multiply by 180 to the the advance in degrees. There's also a factor of 28 degrees to subtract from that number due to the fact that the crank trigger is not at TDC on the flywheel.

There are two triggers per revolution of the flywheel since there are two notches in it. This corresponds to the four trigger points inside the distributor, which runs at cam speed, i.e. half the crank.

I'll give an update later explaining how I am driving the crank sensor and getting useful signals out of it.

[Oetker]

Before I sold my 2.2 I set up a Speeduino.
Its around 150 Euro to bulld one.
Th idea was to made the 2.2 injection because the car was drivenig like sh*t.
I made the kit and puted the Arduino on it and installed firmware.
This allows you to make the injection and all bells and whistles for a modern system.
I stopped the project because I got the CIC carburator good after working 150-200 hour on it.
The original carb is best for the Murena, and after years of driving with the CX carb I experienced the full potential of the car.

Now the car is sold I was thinking to use it for my 1.6, but I am selling that one to ending the hobby because of age.
However I think that system can adress al problems concerning fuel and ignition.
Its highly experimental and universal.
It usues freesoftware and evrything can be adhjusted by a laptop.

see....https://speeduino.com/home

[Anders Dinsen]

Was the CX carb a Weber or a differently jetted Solex CIC? Carburettors can be really rewarding to work on and it sounds like the hours you invested paid off in the end! :-)

I also considered EFI at some point. The arduino project looks really interesting! The starting point for my project was however that I wanted to keep the original Bosch distributor (and my Weber DCOEs) but wanted to be able to verify the actual curve - and eventually tune it by exchanging springs and weights inside. For that I needed a device that would allow me to measure advance while the car is running on road or rolling road.

[Oetker]

When I bought the car in 2007  they just fitted the Citroen version at the local dealer.
This carb was a bad choice because out of specs.
The car drove with dips in several rpm and I always had to drive it with choke on to get some drive-ability.
The original was gone else  so I bought a second hand CIC Talbot but was badly warned and CO and acceleration jet where damaged.
I drilled the CO on a another point (Like the Citroen one) carefully not to danage other tunneling.
I replaced the axels because of leaking using material from Lada Carbs.
The carb was also bend so the sec valve did no fit good to produce the vacuum, and it was out of line for the progression holes.
I made the foot of the carb flat and produced a valve for the second hole from a bigger valve (Lada) and made it fit in the sec so it closed in the right angle to make it work again.
In experiments I made the 127.5 jet also 135.
The car runned a bit rich but no more holes in acceleration and top speed was possible without problems.
It took a lot of time but it was also a learning curve.












I know for sure that the original carb is the best way to go for the car but getting the setup right took me lots of time.

[Anders Dinsen]

Just wanted to share some data I got out of my ignition advance timer this weekend as I had it connected via a serial port to a laptop.  The data is processed in Excel.

The distributor is an unmodified Bosch distributor for the standard 2.2 engine. It looks like the distributor has two separate curves, but this is due the the inertia in the internal mechanics. The my engine has a well lightened flywheel so it spins up (and down) real fast.

I have identified a problem with the advance data stopping being calculated at rev's above about 3500, so I'm currently looking into that issue. I need to do some more signal measurements on the car.

Also, my device seems to miscalculate the advance on higher revs, so I have had to filter out some extreme values which are clearly incorrect in excel. More or less as I expected when I started designing this circuit, which needs to be sensitive to signals and robust to noise at the same time. There's a balance I still need to strike.

Also, because of my lightened flywheel, there is really too little metal to detect, but I'm going to try to see if I can get it to work all the way up to the rev limit of the engine.

EDIT: At least this gives a clear indication of the advance curve of the distributor!

[Anders Dinsen]

I realized, I haven't shared a block diagram of what I'm doing, so I drew one up this morning. A friend suggested that this is a distributor advance curve tester, and yes, that's exactly what it is, and since it's measuring actual data on the engine, it can be used also for a few other things.

You're also getting a picture of my set up last weekend when I finally got data out of it.

It's still work-in-progress to ensure it works perfect, but that's part of the fun now

/Anders

[Anders Dinsen]

I've agreed with Herman to split the topic to separate out his maintenance thread on the 34CIC, but there's a bug in the forum software which means I can't do it, so that will be later.

When I bought the car in 2007  they just fitted the Citroen version at the local dealer.
...