EGR experiment (Increase EGR flow for fuel economy)

[t vago]

Reference: Increasing EGR flow for better mileage

Today, I modified my EGR system - hopefully for better fuel economy.

First, I took some measurements of my EGR valve. I had a spare EGR valve assembly from a very similar Chrysler engine, so I partially dissected it. The valve controls consist of a solenoid, and a plunger-style potentiometer to measure the solenoid position. The engine computer reads the position of the EGR solenoid (thus, the position of the valve itself), then sends a variable width 128 Hz pulse to the solenoid to open or close the valve to a desired percentage. It goes without saying the the engine computer completely shuts the valve when the engine is idling, at full deceleration, or at moderate (or above) acceleration.

The potentiometer varies along a 2.7k ohm value, and there appears to be a 490 ohm resistance in line with the wiper. Not too concerned about that, though.

Once I got the values, I spliced in a 1 kohm resistor, along with about 3 feet of extra wire, into the EGR sensor supply wire. This willl have the effect of reducing the voltage that the potentiometer sees, thus also reducing the apparent position of the EGR solenoid. The engine computer should then increase the EGR position to compensate.

Looking at the FSM, there does not appear to be any rationality testing for the EGR system. Basically, the EGR valve can pop a code for "No voltage to potentimeter," "Potentiometer shorted to ground," "Potentiometer shorted to supply voltage," or the generic "EGR system not responding." In other words, splicing in the 1 kohm resistor should not pop any codes.

I tested the EGR system with my trusty Snap-on MT2500, and it still responds normally. It's amusing to cause the car engine to start idling horribly when the EGR system is commanded to kick in... Heh. I will take a test drive shortly, and will report my initial feedback here, at some point in the near future.

[pete c]

I am a far cry from any sort of expert on modern engine control systems, but, I wonder if this will cause issues with things as the TPS position will be different from what it thinks it should be.

My guess is that you may have some drivability issues around town but, not so much at steady state cruise on the highway. Maybe wiring things so you could switch between the standard setup and a modified one would be possible once you have attained cruising speed. Perhaps even use a variable resistor so you could fine tune it as you drive.

Good luck with it.

[t vago]

Initial feedback: #@$%^%$&$%^#$

Of course, the car kicked out a P0404 code, meaning that the EGR position sensor performance is faulty. Also, after about 5 minutes of operation, the EGR system got commanded to a continuous duty cycle of 0%. I think the engine computer is not liking that resistor in series with the sensor power wire.

I must tackle this from a different direction, it would seem.

Originally Posted by pete c

I am a far cry from any sort of expert on modern engine control systems, but, I wonder if this will cause issues with things as the TPS position will be different from what it thinks it should be.

I am not sure this will be the case, at least with the TPS. MAP sensor readings might, however, cause the engine computer to throw another code.

Originally Posted by pete c

My guess is that you may have some drivability issues around town but, not so much at steady state cruise on the highway. Maybe wiring things so you could switch between the standard setup and a modified one would be possible once you have attained cruising speed. Perhaps even use a variable resistor so you could fine tune it as you drive.

Yah, that's a good question. I'm not sure this mod is even allowable for my car.

Originally Posted by pete c

Good luck with it.

Thanks! heh...

[pete c]

Sounds like that damn engine computer is a little too smart for its own good. It looks at the various inputs, does the math and says something is screwy. This job sounds like it will require the ability to actually modify the computer's code. Maybe you can get away with smaller changes. Rather than the fixed resistor you added, you could try a variable. Start it at 0 Ohms and slowly increase it, a bit at a time.

Good luck with that! You'll need it.

[t vago]

Yah, I will likely next go with the variable resistor route. I just have to find a small variable resistor (<= 1k ohm or so).

And if that doesn't work, I got a similar EGR valve sitting right beside me. The EGR valve in the car is driven by a pulse-width modulated 128 Hz signal. I'll hook the spare EGR valve up to the engine computer, piggyback the solenoid driver signal into a small microcontroller (likely a Microcore-11, since I have a few of them here that are gathering dust), and use that to drive the actual EGR valve. The microcontroller would then read the variable resistor, and multiply the EGR PWM signal by some percentage determined by the variable resistor.

[t vago]

I could not find any potentiometers smaller than 5k ohms at Radio Shack, and was not willing to wait around for Digi-Key/Mouser to ship me one.

So, I put a 1-pin weatherpack connector on the wire extension I installed yesterday, then installed a 1-pin weatherpack jumper with a 220 ohm resistor on it. Now, I can very quickly switch between stock and added resistance. Hopefully, the 220 ohm resistor will be small enough to be acceptable by the engine computer. If not, I might as well go to plan B above, because I just don't see that it's possible to get any sort of meaningful increase with anything smaller than 220 ohms.

I've added the below table in an attempt to illustrate the # increase in EGR valve opening with the added resistance. "Stock" is what the engine computer would get if it commanded the EGR valve open at a certain percentage, without the added resistance in-line with the position sensor. With the feedback from the valve position sensor with the added in-line resistance, the EGR valve would open to the "modified" percentage value.

% EGR Opening, Stock and with 220 ohm resistor added

Stock	Modified
0.0	0.0
5.0	5.4
10.0	10.8
15.0	16.2
20.0	21.6
25.0	27.0
30.0	32.5
35.0	37.9
40.0	43.3
45.0	48.7
50.0	54.1
55.0	59.5
60.0	64.9
65.0	70.3
70.0	75.7
75.0	81.1
80.0	86.6
85.0	92.0
90.0	97.4
95.0	100.0
100.0	100.0

[Daox]

Very cool. I've wanted to do this for a while with my Paseo, but never had the time. Looking forward to seeing how it works out!

[AndrzejM]

Interesting thread. Subscribed

[t vago]

Feedback after a 120+ mile commute to and from work: Eh... Maybe?

The engine computer seems to be satisfied with the 220 ohm resistor. There were a few instances where the engine computer appeared to double-check the EGR valve position, both to and from work. However, it did not throw any code, nor does it have a pending code.

If I am to believe the trip fuel consumed figure of 2.24 gallons consumed, on my return leg commute, then it would appear to be that I somehow got 27.1 MPG on that 60.6 mile leg. This would be better than the 24 MPG I would otherwise get on this leg. However, many things occurred which call this figure into question - namely, it rained (so the road was wet for most of the commute), heavier-than-normal 1 AM traffic proved to be somewhat of a challenge, and the turnpike section was completely under construction (forcing me to slow down to below 60 MPH). (side note - You know, Rt 8 well and truly needs to be resurfaced, and has needed it for several years now, but they decided instead to resurface the far left lane of I-80 - which had just gotten resurfaced last year! @$#%#^$)

[AndrzejM]

So maybe that's a good moment for A-B-A test? ;-)