Cylinder Identification (CID) Sensor

[johnc32779]

Looking forward towards sequential injection, one must synchronize injection with the intake phase of the cylinder. As far as I am aware, all MS work has assumed this synchronization is accomplished by sensing cam position. The purpose of this posting is to encourage discussion and consideration for using an ignition event for synchronization.

Background:

My framework to this point has been limited to my engine, a BMW M30B35 with a 60-2 timing wheel, a conventional distributor for spark sequencing, and originally came with a Bosch Motronic 1.3 ECU. Of particular interest is both Bosch Motronic 1.1 and 1.3 systems employs a cylinder identification (CID) scheme which detects when a spark occurs on cylinder 6. The firing order is 135246, and injection uses two alternating banks (i.e. 135, and 246). According the only Internet reference I can find, the Motronic ECU uses this signal to synchronize which injection bank to fire so as to provide a smoother idle.

The moulded CID sensor, which slips over the high tension ignition wire, appears to be coil which directly connects to the ECU. Referring to the attached is a Motronic 1.3 schematic, the CID input and it's associated input circuitry can be seen on the left side of page one about half way down the page. It appears that the high tension pulse induce a small AC signal which is returned to the ECU and transformed into a digital pulse recognized by the CPU.

Proposition:

The proposition, or better yet the questions is, can a scheme be developed where the Bosch CID sensor can be use to (1) synchronize sequential injection with distributor ignition, and (2) synchronize both sequential injection and COP ignition? In the latter case I am thinking that by firing the ignition in waste-spark pairs, it may be able to distinguish between the real spark and the wasted spark – don't know, just a thought.

If the CID scheme is workable it should be much cleaner install than having to add a cam sensor.

Johnc

[UnaClocker]

It would also work nicely for engines that don't have overhead cams where getting a sensor onto the camshaft could be a really difficult thing.. Sounds like a nice trick to me..

[Andy_Stprbeck]

This reminds me of a claim that i heard once about the Aprilia 450 V-twin dirt bike/supermoto bike. Supposedly this bike has sequential injection but no cam sync sensor of any kind. The claim that I heard was that when you first start the bike, and it doesn't know if it is correctly synced or one revolution off. The ECU basically just guesses, waits to see if the bike starts, if it goes through some number of ignition events without starting, it switches to the "other" revolution and so on.

I suppose the Megasquirt could be programed to do something similar. Guess, wait a couple of ignition events and if it hasn't started switch to the other, wait a few and so on.

Not sure if it's practical but if it was it would be pretty sweet for the people switching from MSII to MSIII/sequential if they didn't have to add another sensor. Unfortunately even with my very limited knowledge about it I could see how this could be a huge headache or just not possible. Might be easier to just add the sensor.

[pascal]

The proposition, or better yet the questions is, can a scheme be developed where the Bosch CID sensor can be use to (1) synchronize sequential injection with distributor ignition, and (2) synchronize both sequential injection and COP ignition? In the latter case I am thinking that by firing the ignition in waste-spark pairs, it may be able to distinguish between the real spark and the wasted spark – don't know, just a thought.

Sounds very appealing, especially as switching between semi- and full-sequential for even fire engines seems possible. And the upcoming Intellitach would certainly help in "reading" the signal...

Pascal

[jsmcortina]

I can only see this working with a distributor so I'm not really sure how value it would have.

James

[pascal]

Misunderstanding on my side. I expected to see different current on the HT leads with a COP/CNP setup fired in "fake" wasted spark mode. And did probably mix in all the things I read about ion sensing (not necessarily PPP though...).

Sorry for any confusion.

Pascal