EGT for mixture setting

[orlin steven jared]

> Does the EGT sensor have to be close to the head?  The reason I ask is 
> that I have headers, and the only practical place to mount EGT's would be
> in the collectors.  I assume that the temperatures would be somewhat 
> reduced, but still proportional(?)

It would be almost useless to put the EGT sensor in the collector. First,
you wouldn't be able to tell squat about mixture distribution between
cylinders.  Second, in order to get *any* correlation between mixture and
EGT temp, you'd have to check with a Horiba lambda sensor.  And that is 
assuming that such a correlation would exist and give you consistant results.

You have to use the EGT sensors about an inch or two away from the exhaust
port, for best results.  Incidentally, I don't know much about T/C's, but
I know that certain ones are rated at certain temps, so where you put the
thing would depend on what temperature range your T/C's can handle.

> How does the mixture affect EGT?  Do lean mixtures produce lower or higher 
> temps than rich ones?  

>From what I've seen, at mid RPM ranges, rich mixtures (below stoich) produce
lower temps, while leaner mixtures produce higher temps.  But this is
not set in stone.  It will depend on the operation of the particular engine
being tested.  And each relationship between mixture/EGT temp will most
likely be different.   Which brings me back to my original point (maybe someone
elses too?).  EGT's are best used for checking distribution, not mixture.

> Can you determine if you are at stoich with this method (with any accuracy)?

No.

> Is there a linear relationship between EGT and AFR or is it a complex 
> function?

It is probably complex, but as I said, could vary for each engine tested.

> Is this a practical/effective/reliable way to adjust mixture?  Is it only 
> useful at WOT, or does it maintain some validity at all throttle settings?
> 
> What other factors affect EGT?


The last time I checked, 8 thermocouples ran about $240.   The readout
box, which would give all 8 temps at once would cost about 2000-2500
to pay someone to construct.  I'm not an EE guru, but maybe there are
some people reading who could build this circuitry themselves for MUCH
less.  If so, I would think that it would be very cost effective in
checking distribution.

For fine tuning mixture, you simply need the UEGO sensor, Or some other
gas analyzer method which can give you an absolute reading of AFR.

Seat of the pants, and trial and error still work well though.


Steve