i was working on a project and i discovered something odd about the stock honda distributor that might affect the autronic, or any sequential efi system.

from my understanding, the sync sensor on most aftermarket computer systems that don't have a customizable trigger setups are supposed to trigger somewhere near tdc of cylinder 1.
http://www.autronic.com/software/trigger4cyl.pdf


this is a picture of the sync sensor when aligned.


this is the position of the rotor, when the sync sensor is aligned. its has just passed cylinder 3 and is approaching 4.

What this makes me conclude is this.... if you configured your injectors in the standard manner meaning:

injector output 1 = cylinder 1
injector output 2 = cylinder 3
injector output 3 = cylinder 4
injector output 4 = cylinder 2

the Autronic, or whatever ECU now thinks cylinder 4 is cylinder 1, cylinder 2 is cylinder 3, cylinder 1 is cylinder 4 and cylinder 3 is cylinder 2.

this might seem like a huge enough problem where the car shouldn't even run, but because the actual spark is directed physically by the distributor rotor, the spark ends up in the right hole.

if the above is true, then the problem here is that your injectors are squirting fuel at the wrong time, totally negating the benefits of sequential fuel injection. also, you're individual cylinder timing/fueling are indexed to the wrong cylinders.

the correct configuration would be this.

injector output 1 = cylinder 4
injector output 2 = cylinder 2
injector output 3 = cylinder 1
injector output 4 = cylinder 3

Anyone have any thoughts? Are my assumptions correct?

 

Your correct...with this setup, number 4 actually fires as number one. You just have to stager which wires go to what injector...that way they will run seq., and not just fireing 1,2,3,4....you run them in the fireing order.

 

to get the sync signal u need to take the 2nd magnet down in it... find which tooth out of the 16? teeth on it best lines up with cylinder 1 and cut off all the rest of the teeth

 

bump for a good thread.

 

Would Coil Packs and a CAS be a good work around? I've always wanted to run DIS anyways

Suprdave

 

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by ninesecrx &raquo;</TD></TR><TR><TD CLASS="quote">Your correct...with this setup, number 4 actually fires as number one. You just have to stager which wires go to what injector...that way they will run seq., and not just fireing 1,2,3,4....you run them in the fireing order.</TD></TR></TABLE>

hmmm, thats weird, I would just pull the magnet off and align it how it should be, maybe thats a SOHC or LS distributor shaft on a dohc vtec distrib, evne the h22a shaft is different than the b-series vtec or maybe cam receiver is off 180 degrees ? or is that how the autronic setup comes ?

greg

 

Andy,
Good eye on catching that...........but here's the story..........

The reason the sync happens before the #1 cylinder is so the ECU knows that the NEXT EVENT will be #1..........if the sync happend actually on the #1, it would already be over and the syncronization would then be out........

that's why the Autronic always fires the ignition outputs starting with the last one first like this

4 1 2 3 {4 coils}
3 1 2 {3 coils}
2 1 {2 coils}

On a distributor car, the sync signal could be off, and the car will run fine, but when you have individual coils it must be sequenced correctly in order to run.

-hope this helps.

-Ben

 

Interesting observation but I don't think the ECUs are actually injecting or igniting the cylinders in the wrong sequence if you setup the firing order correctly. And as far as which cylinder goes first, it is of no consequence. For example a typical Honda Civic has 1-3-4-2. From the stock ECU's stand point, it's the same as 3-4-2-1, 2-1-3-4 as long as the ECU know when cylinder 1 gets to TDC.

Now due to the sensor setup, your sensor for determining TDC of cyl 1 may be triggered X degrees before TDC of cyl 1 (or equivalently 720-X after TDC of 1). And as everyone probably noticed, Honda's distributor sensors aren't all lined up nicely. And determining the timing of the sensors in the distributor is also challenging as it is really reference by the key to the camshaft.

What would be helpful is to put a particular distributor's signals in question on an oscillascope and get a picture (this isn't an actual distributor sensor signal capture, but an injector output signal reference to the TDC sensor) like this to tell exactly how many degrees BTDC (can be easily derived from measuring the period of the signals) is the CYP sensor (this is the sensor in a typical Civic distributor that's triggered once per cycle or 720 degees) in reference to the actual TDC of cyl 1.

Top signal is injecotr n, middle is conditioned TDC signal and bottom is unconditioned TDC. Notice the injector timing and duration in relation to the rising edge signal of the TDC for cyl n. This signal was captured from a Civic OBD2 ECU without any other sensors connected (thus in safe mode).

Once you have determined the timing, then you need to program the ECU to tell it what angle the crankshaft is at when the CYP sensor is triggered or in another word, tell it how many degrees into the future to expect the TDC of cyl 1 when the CYP sensor is triggered.

There is something that I am not 100% about most aftermarket ECU is the timing resolution they are capable of for these sensors. In a Honda Civic as an example, does the CYP have to occur many degrees before the actual TDC of cyl 1 for it to accuratly calculate where the TDC is? Also, during cranking, the crankshaft speed fluctuates tremendously and to what degree can these ECUs cope with the fluctuation to obtain synchronization?

Eddy

 

hey eddy,

i'm thinking that the injectors fire in the wrong sequence when in standard configuration because the autronic doesn't have the capability to user define at what angle the sync (CYP) sensor triggers. it is set to expect the trigger near tdc of cylinder 1 during the compression stroke. in most aftermarket ecu's, the triggers are setup to accommodate the particular specifications of the ecu. I guess what's great about the motec is that they can adjust for situations such as these.

about syncronization, when experimenting with the motec, i noticed that it took approximately 2 secs of cranking before a spark plug would actually fire. that's considerably slower than the stock ecu, as it fires almost instantaneously. I wonder if the stock ecu during cranking switches to some sort of non sequential fueling method, and switches back to sequential once its syncronized?

also, did Dan tell you we got his car fired up with the individual coils? it was a lot harder to do than i expected.

Andy

 

Yup, been through that when I set my E6K up. Had to tear apart the dist because i couldnt find anyone who knew how to set one up. Anyway, had to wire my injectors in order of the firing since the Haltech has no way of programming in the firing order of the engine. It doesn't know which cylinder it is firing.

 

Eddy,

Do you know if the signal from the ECM to the distributor ignitor is a 0 to 5 volt signal or a 0 to 12 volt signal?

 

Yeah, if the firing order is not programmable then the wires would have to arranged in firing order. And if the sensor angle can't be programmed then the sensor would have to be moved.. What a pain..

And yeah, I noticed that stock ECU fires up fast. I think during development of road going ECUs that starting up on first crank within 1-2 cycles is key criteria. If I recall correctly, the stock ECU generates injector and ignition signal on the first CYP sensor pulse if not the second pulse. If you think about it, any ECU programmed correctly should be able to synchronize within 2 cycles. But that can easier said than done. On Civics, the ECU does have something to cheat. For example, the starter signal may be used by the ECU and also the CKP sensor can give a signal even 1/4 of a cycle and therefore be able to synchronize the ECU to the engine a lot quicker.

I can't say for sure but it would be interesting to see if there are any "interference" problem on the sensor signal from the distributor to Dan's motec on a scope. He said that you get a sync error above 7000rpm. Sounds like noise to me.. Maybe you can play with his Motec to tune the sensor filter and also the trigger voltage level.

The coils are magical.. That's why..

Eddy

 

The signal is helf at 12V and drops to 0V by the ECM when it signals the ignitor to discharge the coil. You can look at the graph. Each grid vertically is 5V so you can tell that the voltage swing is a little more than 2 blocks.

 

Actually,
The Autronic ecu with the 1.92 chip version {latest} has a user selectable trigger angle for the crank signal that goes from 45-112 degrees

Then there is a huge window of opportuninty for the ref signal.......on a four cylinder honda that window would be something like from 30 BTDC compression all the way to 130 ABDC combustion #1

There is a neat program that you can get for free at http://www.autronic.com that will allow you to figure out your trigger set-up for whatever chip version you have and how many cylinders your engine has very easily.

-Ben