[Diy_efi] re: Problem with cam mounted optical trigger sensor

[Mike]

At 05:43 AM 4/11/06, you wrote:
>There is a 5v pull up inside the ecu. This pulls any rising edge to the full 5v. When you rotate the sensor and it is not connected to the pull up, the rising edge of the signal gets smaller the faster you rotate. It gets down as low as 0.4v. 

If you are measuring this with a meter, then thats what you'd expect as the
frequency response of the meter will effectively see the average, unless its
an RMS meter - either way the reading is meaningless...

>  I was originally feeding the sensor only 5v and the signal was even lower. I now feed the sensor a 12v supply and the signal is higher at 0.4v. Possibly the noise is creating a tiny rising edge and the pull up inside the ecu is pulling this high to 5v thus creating a false reading?

You really need a good CRO/probe to see the edges, if the sensor is meant for 5v
then do you think its safe to feed it 12v - especially when its feeding into a ECU
with only a 5v pullup ?
   
>  The pull up resistor is 680ohm. This seemed a very low value to me for a pull up but thats what Walbro told me to fit. Do you think changing this value would help me?

Thats par for the course, you dont change the value, you are not redesigning
the ECU, hopefully, you are only fixing a sensor or power or cable fault...
 
>  I am going to try to get a scope that can log output. Then I can confirm that there are extra rising edges. I can't tell with my cheap nasty scope because its just a very jumbled square wave.

Ah ha, did you use this to get the 0.4v reading above when it was connected to
the ECU (ie When the pullup was operating), if so then it sounds like either
your sensor output stage has been damaged or its connected incorectly or if
you are measuring this when your cable/loom is connected then you have a 
cable/loom fault, does the body of the sensor need a ground connection ?

Just out of sheer curiousity, what sort of CRO and probe do you have, I ask
this because when we had a student in the mid 80's work part time with
our nucleonic ore flow gauges he used an old cro with an abysmally low
input impedance and probes that had a broken compensation cap, so the
net effect was the pulses he was seeing from the scintillation counters looked
like tree silhouette's from a nuclear test, ie flopping all over the place and
no chance he could see a stable trigger - as you seem to have trigger
problems I wonder whats up with the cro or sensor ?



Regards from


Mike Massen
Network Power Systems
Lab 08 9444 8961
Mb 0438 048961
Perth, Western Australia
* VL/VK Commodore Fuse Rail that wont warp or melt !
* Skyline/NIssan/VL Milspec ignition driver in development
* Twin tyres for most sedans, trikes and motorcycle sidecars
* Industrial grade PolyVinyliDeneChloride (PVDC Copolymer) in bulk, the best
  oxygen and water protective barrier you can find for printed circuit boards.
Web site under construction   http://niche.iinet.net.au