[Diy_efi] Info on how to make crank position sensor circuit?

Perry Harrington pedward at apsoft.com
Thu Jun 6 12:54:55 GMT 2002


I believe that the car makers already made it trivially easy for us.

I was thinking about this and realized something.  The 60-2 wheel
is to the developer's advantage.  You know that it's physically
impossible for an engine to double or halve it's rpm in 6 degrees
rotation, therefore the missing 2 teeth solves the problem.  If
the pulsewidth between falling edges of the teeth more than doubles
from the last sample, you know you hit the -2 teeth.

The pulsewidth between falling edge should triple in the -2 teeth
gap.  That's a very large sentinel value and does not require any
special PLLs or trick algorithms.

If you're looking for a good circuit that will detect an ignition
signal, refer to the DIS ignition article on the diy-efi.org website.
The article uses an LF357 opamp as a comparator to detect the zero
crossing point of the VRS sensor.  I used an LM324.

There are 2 issues with the design that you should be aware of however:

The 10M hysteresis resistor is prone to making the opamp oscillate when
any noise may capacitively on the inputs.  I built this
circuit last night and played around with it.  With the input open it
made no noise until I held the input wire in hand, then a ~60hz signal
was imposed on it.  I replaced the 10M resistor with a 1M and the noise
went away.  Using guard traces will improve this as well.

You can observe the sensitivity of the circuit easily.  With the 10M every
crossing of the VRS sensor would cause the opamp to flip-flop.  This
is handy for detecting very slow crossings, however it also makes it prone
to noise.

The other issue is the 10k trim pot to set the comparator voltage.  This
is useful in a DIY system for tailoring to your setup.  I would recommend
a 100k or 500k pot for this, as by the time you adjust the trimpot down
to the value you want, the current would be too high.  I found that a
good value for the voltage divider was 100k on the high side and 2.2k on
the low side.  The 2.2k keeps the current at sane levels (instead of 220ohms)
and the 100k doesn't impose any real issues.

Also, you can substitute out the 1M input resistor, I'd choose 10k.

You might also try adding a 100k input pullup.

--Perry

On Wed, Jun 05, 2002 at 01:02:24AM -0500, bcroe at juno.com wrote:
> If I wanted to resolve between teeth, I would keep track of the 
> time between the last 2 teeth passing.  I would divide the time 
> since the last tooth by the time between 2 teeth to find out how 
> far past the last tooth I am.  
> 
> Phase locked loops are a tricky business, esp for widely 
> varing frequencies.  Bruce Roe
> 
> On Tue, 04 Jun 2002 14:49:53 -0700 David Gravereaux 
> <davygrvy at pobox.com> writes:
> > bcroe at juno.com wrote:
> > 
> > >Just how much more accuracy than individual flywheel teeth 
> > >is really useful?  Bruce Roe
> > 
> > Whether I infer in hardware (integration in the loop filter 
> > of the PLL), or infer by time in between teeth, it's the 
> > same result (hopefully).  As I have no idea how to do a 
> > DPPL, I'll start with a 4046 and a timer used a 
> > programmable counter and see where it gets me.  A 
> > quarter of a degree
> > resolution seems like a good starting point.
> > 
> > If I can get it to pass the strobe test with a gear on a 
> > drill jacking the rpm all around, i'll be pleased.  From 
> > the archives, someone advised a DPPL to be more 
> > adjustable with regards to noise rejection because 
> > all the math is right in front of you.
> 
> > David Gravereaux 
> 
> _______________________________________________
> Diy_efi mailing list
> Diy_efi at diy-efi.org
> http://www.diy-efi.org/mailman/listinfo/diy_efi

-- 
Perry Harrington             Linux rules all OSes.               APSoft      ()
perry at apsoft dot com 			                 Think Blue. /\

Those who would give up essential liberty to purchase a little temporary safety
deserve neither liberty or safety. Nor, are they likely to end up with either.
                             -- Benjamin Franklin


_______________________________________________
Diy_efi mailing list
Diy_efi at diy-efi.org
http://www.diy-efi.org/mailman/listinfo/diy_efi


More information about the Diy_efi mailing list