Sequencer First attempt in car

[ca434sbc4]

Quick update..

-I found some 1F caps for a reasonable price - should be here Wed.
-Wired the injectors up to the sequencer - without a external prime its too lean and won't fire. Even with an external prime its so lean the engine dies after a few seconds. Currently its set for 2 point cranking pulse widths, I'll change over to the table set up and crank the pulse widths up until it starts.
- I need to finish wiring in the TPS and IAT sensors, currently the IAT and TPS signals are tied to signal gnd. The IAT will try and lean the mixture out when connected to gnd. as it thinks the air is hot.

Another round tomorrow evening time permitting.

[grippo]

You will probably need to increase the cranking PW considerably because with the present mode you are not squirting all cylinders at once, just two at a time in wasted. So you probably need a factor of 3 increase.

[TheMonkey]

...
I've been out of town, but I'll try my tach signal to my tach before I pull the temp wiring for final wiring and splicing.

i am not getting a good reading on my tach. brand new autometer. i made sure toggles were set for V8. i'll take a closer peek at the signal on the scope to see what the signal looks like. my scope will get data points as low as 5 microseconds if only using one channel, so i can see this signal close, but can't see too much detail on a 15 microsecond pulse.

at first it was just reading a fractional RPM (600 when motor was 1000), then next try it just flinched when turned on and didn't show any RPM.

why is the pulse so short? it's not possible to have an adjustable pulsewidth?

[krisr]

My Autometer tacho has always worked flawlessly when being wired up to the various CDI units i've used and they nominally use a 20% duty cycle on the tacho wire if my memory serves me correctly.... Looking at the circuit for the tach it it should be a nice 0-12V swing.

[grippo]

Enough people have had tach problems that I need to lengthen the pw in the next version. It is presently about .15 ms. The tacho works fine, I have checked it many times, but it's evidently just too short. So don't waste time scoping it.

[TheMonkey]

My Autometer tacho has always worked flawlessly when being wired up to the various CDI units i've used...

yeah, this tach worked fine with MS2 using MSD box tach output.

you're right Al- no need to scope it. i've seen it before, i know it occurs every pulse, and i know it's very short.

[Bruce Bowling]

Actually, 150 microseconds is not all that short, I am curious why some tachs cannot handle it. Part of the issue is that many tachs will not trigger off of 12V but instead react to the flyback pulse from the ignition coil primary side, something like 30V triggers it. There are simple tach adapt circuits using a transistor and inductor to generate the flyback voltage. Not sure what Autometer wants in way of trigger.

Anyway, Al will lengthen it. Now.. it'll be too long and high RPMs will not register because the duty cycle will exceed 100%...

- Bruce

[RedRocketRally]

Is it possible that its the polarity of the signal, rather than the pulse width? In a Tach Output thread in the Ignition section, someone noticed that the signal in the "old" software was 12V when the engine is off with 2-3ms ground pulses. In the new software, its ground when off and the pulses are 12V for .15 ms.

The instructions for my autometer gauge say to connect to the negative side of the coil or to the tach terminal of an electronic ignition. maybe the sensing circuit is expecting short ground pulses (that's how a coil would look, right?) and needs 12V for longer to charge (something?) back up to wait for the next pulse? Maybe it rejects short 12V pulses for noise reduction?

Just guessing, but if 150usec should be long enough, could we get a test version with the polarity flipped? That should just involve swapping the &'s and |'s in the code and won't affect any timing, right?

[TheMonkey]

well, my curiosity got the best of me, and i went and measured it.

the length of my pulse is between 96 - 101 microseconds (scope grabs 5 microsecond datapoints). the voltage was kind of weird; between 8.5v & 10.8v.

BUT... there's a hanging chad.... my wiring is still messy and susceptible to RFI. i'm not sure my example is meaningful until my wiring is final. next on the list....

[grippo]

The tachO output for the sequencer, in addition to appearing on the Ampseal as a tach pulse, also goes from the main to the router processor and is a key signal in keeping the two processors synched. I can't change this signal lightly. It's period has to be smaller than the smallest expected tach cycle time, which would be about 1.5 ms for a very high revving 4 cylinder.

If you wanted to test an inverted signal, you could just put it through a transistor on a perf board. A 555 timer could also be used to lengthen the pulse to say 1 ms.