More DIY EGOR

[Bernd Felsche]

RDwoo10 at aol.com tapped away at the keyboard with:
> I think that we are in agreement here.  as the temp goes up, the
> resistance goes up.  as resistance goes up we want the current to
> go down (E=I^2R, AKA power).  if the voltage is constant, the
> current will go down (E=IR), because the temp went up.  this will
> happen by itself.  i guess, now that i think about it.  a constant
> voltage source is sufficient, as long as it does not take forever
> to get back down to temp.  if that is the case then we want to
> move the drive voltage inversely proportional to the temp so that
> the sensor gets back down to temp (or up) fast enough that the
> sensor is still outputting the right thing.  All this w/o making
> it oscillate wildly.  

The environment is hardly steady-state.  There are many possible
perturbations. We need to know how much the signal output changes
for set of interesting, fixed Lambda values, over a range of
operating temperatures to get an idea of how "bad" the numbers can
get. The gain of the system can then be reduced to a level where any
oscillations fall within an acceptable error-band.

Heat is coming from two sources; the heater element and the exhaust
gases. You only know one of the variables (element resistance which
corresponds to "temperature") so the quality of control is reduced.

Consider also that the heating element will be hotter than the
actual sensor during warmup and cooler if the exhaust gases are
heating the sensor above the desired setpoint.

Two underlying reasons why I posed the question about adding a
thermocouple to the "outside" of the probe. It gives you the
magnitude of the temperature gradient and the direction once
calibrated. Calibration for the pair could be done by slowly heating
the sensor in an insulated "jar" so that the heater and the
thermocouple are at effectively the same temperature.

The other advantage is that the heat loss from the probe to the
exhaust gas and the metal of the pipe/manifold can be allowed for
when heating the sensor initially; avoiding an excessive temperature
gradient which might crack the sensor.

-- 
 /"\  Bernd Felsche - Innovative Reckoning, Perth, Western Australia
 \ /      ASCII ribbon campaign | I'm a .signature virus!       |
  X       against HTML mail     | Copy me into your ~/.signature|
 / \      and postings          | to help me spread!            |
----------------------------------------------------------------------------
To unsubscribe from diy_efi, send "unsubscribe diy_efi" (without the quotes)
in the body of a message (not the subject) to majordomo at lists.diy-efi.org