[Diy_efi] Re: [Gmecm] 305/350 Swap 10-month Progress Report...

John Gwynne jsg001
Wed Mar 16 02:11:35 UTC 2005


Mike Frels wrote:

>I thought I'd drop my favourite lists a quick report on well overdue progress on my TPI 350 of significance. The last time I spoke up it was with the results of my failed 15mph treadmill smog test and how my NOx levels were too high. Suggestions were made regarding timing reduction to pull the NOx down. Well I reduced my initial from 6BTC to zero and the only thing that came down was my gas mileage. I put the 6 back in and finally broke down and installed a new cat. Car passed 15mph with NOx 10% below max allowable. Strange though that the 25mph test remained virtally unchanged from when I had it tested with the gutted cat. The good thing about all this testing was that the tests confirmed the idle/low speed 160 BLM's I was getting
>  
>
I would bet your high NO levels are a result of being too lean.  Were 
your CO levels 0%?
If so, I had the same problem two years ago.  My numbers were

    test   units limit
HC  13.4   ppm   55.0  Pass
CO  0.00   %     0.30  Pass
NO  1429.2 ppm   382.0 FAIL
CO2 14.52  %           N/A

A three way converter needs excess CO for the reaction in the front catalyst bed
NOx + CO -> N2 + CO2.  Assuming everything else is running right and you are not producing excessive NO levels, this is where NO are reduced.

In my case, I learned that not all EGO sensors are alike. I had not installed a true gm sensor.  When you look at the feedback PID controller in the code, it is biased slightly rich (O2 hiding in other forms post engine/pre cat).  Changing the code to the maximum richness while still keeping the loop stable and using the non gm EGO resulted in NO levels of 1136 ppm for my Jeep (5L TPI/ vortech SC).  Still not rich enough... but moved in the right direction....

    test   units
HC  13.4   ppm 
CO  0.00   % 
NO  1136.8 ppm 
CO2 13.39  % 

The EGO sensor I was using had a steeper slope than the gm part as you 
moved away from stoichiometric.  Enough so to change the relatively 
precise balance needed for emissions.

If you have trouble on your next test, you might want to take another 
look at you EGO type and placement... FWIW.

john gwynne






More information about the Diy_efi mailing list