Fuel consumption issue/intake noise

[larson701]

Update

Received a 2nd MAF from the ebay outfit. Threw a CEL as soon as the key was turned. I have requested a refund.

I have purchased an OBDII scanner to monitor the operation of the MAF and assorted sensors. The hardware is an Elmscan Compact from Scantool. The basic software supplied with the device allows monitoring but not graphing of the data. I will look for a good software package to go with the hardware.

 

[niky]

Buying E-Bay MAFs is a crapshoot. It could have the same outer casing but have internals for a different engine, since MAFs for many of these Japanese cars sometimes come from a single third-party supplier.

 

[larson701]

The internals for the ebay MAF look entirely different, more like the classic thermistors, wire-wound resistors. The original MAF may be a thermal film-style. I rationalized the difference as the only important thing is the output signal. I guess that was different also.

 

[niky]

Have a friend that went out to buy a surplus MAF and he had the same issue. Different looking internals... different output.

 

[larson701]

2 things need to happen before I can move forward at this time.

1, I need to settle with the supplier of the bad MAFs. Beyond that point I will look for a used MAF. I really don't know if the MAF is the source of the poor fuel economy so I do not want to spend a lot of money on an ineffective fix.

2, I have ordered better software (a release code) for my new ODBII scanner. Hopefully with more data a better decision can be made.

 

[larson701]

No word on the return of the MAFs yet. There are times I am really uncomfortable with this Ebay thing.

I am finally making some progress with the OBDII scanner. The software I am using is ScanXL and the device is a ElmScan 5 Compact. It's taken a bit to get everything working but progress has been made.

O2 sensor 1 may be a problem. Based on research, a multipoint injection system should see approx 5-7 cycles/second, and not slower than 1 second/cycle. Mine is on the order of 1.3 seconds per cycle. A bit slow, possibly lazy.

The test of the TPS indicated what my seat was telling me. The throttle is very sensitive for the 1st 10% of travel. The rest of the range appears normal. I have a replacement TPS coming. Some research indicates a TPS can affect the MAF testing results. The TPS is a bit of a shot in the dark since it is hard to believe it can have the overall effect on the fuel economy.

 

[niky]

My TPS goes all wonky when it gets hot... reading anywhere between 0%-30% when I let off the throttle... making for a weird idle.

 

[slavrenz]

No word on the return of the MAFs yet. There are times I am really uncomfortable with this Ebay thing.

I am finally making some progress with the OBDII scanner. The software I am using is ScanXL and the device is a ElmScan 5 Compact. It's taken a bit to get everything working but progress has been made.

O2 sensor 1 may be a problem. Based on research, a multipoint injection system should see approx 5-7 cycles/second, and not slower than 1 second/cycle. Mine is on the order of 1.3 seconds per cycle. A bit slow, possibly lazy.

The test of the TPS indicated what my seat was telling me. The throttle is very sensitive for the 1st 10% of travel. The rest of the range appears normal. I have a replacement TPS coming. Some research indicates a TPS can affect the MAF testing results. The TPS is a bit of a shot in the dark since it is hard to believe it can have the overall effect on the fuel economy.

Yes, but your scan tool may not show all of those. I have the Innova 3130 scan tool, which was originally almost $300 new, and even its live data function now only shows an O2 sensor change of about once per second.

It all has to do with the number of frames per second that your scan tool can handle. A consumer version may not have the capability to see 5-7 switches per second, because it would have to support 5-7 frames per second, which is very rare unless you paid like $1000 for the scan tool.

 

[larson701]

Slavrenz,

You are correct. However, I was looking for the sine-like trace that indicated I was getting a steady trace. Even at the 0.2 second rate for the scan, I obtained a good wave form. With 5-7 switches/second, I would not have obtained the same form, probably would have seen hash.

This is the reference used for evaluating the O2 sensor:

http://www.autodiagnosticsandpublishing.com/feature/o2-sensor-testing.htm

Would you agree that the 1.3 seconds/cycle is too long?

 

[larson701]

The TPS is installed but not real performance difference was detected. I expected this but gave it a chance based on the MAF 'out of range' indication.

A new oxygen sensor should be here early next week. I have a long drive over the Easter weekend and want the Protege to have its best chance of turning good fuel mileage.

And yes, it's turned into a parts change fest.

 

[larson701]

Update

The new O2 sensor has been installed. This is a Bosch aftermarket unit. I did not use the splice kit that came with the sensor. The instructions left me a bit cold regarding obtaining a positive connection with the wires. Instead, the wires were spliced with crimped butt connectors and shrink-tubed. I used the red connectors (18g?) and removed the plastic insulation. There were no CELs on starting the engine.

Once up to temperature, the scanner was attached to determine if any changes were noticeable compared to the original Denso sensor. The crossing time (period) was essentially unchanged, about a second. Perhaps the crossing time is determined by the ECU as much as by the condition of the sensor. The engine does appear to run a bit smoother at idle.

The effects on fuel economy remain to be seen. I leave tomorrow morning and the forecast tonight includes snow. Total driving time this weekend will be over 14 hours, all on interstate.

I received a notice from Paypal that the purchase price plus shipping for the MAF has been refunded.

 

[larson701]

The snow turned into snow, sleet, rain and lightning simultaneously. There was a good coating of slush in the driving lane. Driving too fast, I was going 60 mph during the first 50 miles. People who left town 1/2 hour later had to turn back. We received 5-6 inches of snow in Bismarck.

The trip to the Cities, was into a stiff headwind, on the order of 20 mph. The Protege5 averaged 25 mpg. Other than the stormy stretch, the cruise was set at 70 mph.

The trip home was into a headwind as well, carrying about 175 lbs of furniture from Ikea. The furniture was not really a significant load but every bit counts. The mileage was 27.5 mpg and the speed was set at 70 mph. The headwind was lighter but noticeable.

No mechanical or electrical bobbles were noticed. No CELs were illuminated.

I do not consider this to be good mileage tho it may have to be acceptable. The fuel economy is about the same as our 2009 Honda CRV.

Not sure what I am going to do next.

 

[larson701]

This will be my last post on this thread, a kind of conclusion.

The gas mileage is somewhat acceptable. Around town, I am seeing mid-20s and on the interstate, upper 20s.

Around town gas mileage appears to be influenced heavily by the length of the drive. My normal commute is 2.5 miles (small town by many standards). While the engine warms quickly, the transmission does not want to shift in to overdrive until the temperature is above 170F, and the mileage remains in the low-20s. Once the car has warmed and on a longer drive around town, the mileage creeps into the mid-20s.

Highway mileage appears to be affected by the temperature as well. In the earlier posts, the car was barely breaking 20 mpg. Now, the car is seeing upper-20s. The only major difference is the change in the weather.

I never found a significant problem with the engine. The only changes made were new plugs, throttle position sensor, front O2 sensor and a new intake tube. The EGR was cleaned.

I normally run E10 and have done so for many years on many different vehicles. I expect there was an exagerrated effect of the ethanol in the winter mix. Ethanol vaporizes poorly at low temperatures and the refinery may have added more volatiles to the winter mix to improve starting. With the warmer temperatures this spring, the mix is moving towards the summer mix, which is better for gas mileage.

I have plans to improve the fuel economy but those come under the heading of 'mods' and will be reported separately from this thread. In the interest of more immediate gas mileage gains, I will increase the tire pressure from 32 psig to something higher.

Pretty pathetic that the P5 struggles to get better fuel economy than our CRV. But I still prefer my P5 for everyday driving!

 

[dd123]

From your posts, it seems like you still made good progress in increasing your mpg...
I get somewhat same but in southern california, I get mid 20ies in all stop n go traffic.

 

[breiland]

I lived in Williston, ND when I first got my 03 Protege. From new, it never did any better than 25. When I moved to Kansas, I actually only got about 23 on the highway, but I've been averaging 30-32 since I moved to Colorado in October. Not sure if the higher elevation does anything to improve mileage, but I thought it was interesting, and I won't argue with it. Anyway...Mid to upper 20's with the weather the way it is in NoDak seems to be normal.

 

[larson701]

I am starting to see just above 30 mpg on the highway, around 25 mpg around town. There is probably no single reason for the increase in fuel economy. As mentioned in other posts, several tanks of gas may be needed before the ECU adapts to changes, such as a tune up or replacing the O2 sensor.

Also, I have taken to increasing the tire pressure to 38 psig for long highway trips, but reduce it to 32 psig for around town. We have a lot of frost cracks in our roads so the lower pressure feels better around town.

And then there's the warmer weather, which basically loosens up the car and makes 'summer' gas available.

And this is in spite of our winds!

Like I say, a lot of little things, probably no one thing.