Fuel consumption issue/intake noise

[larson701]

I recently purchased a 2002 Mazda Protege5, 68,000 miles, automatic transmission, 2.0 litre. This isn't our first venture into Mazdaland, having had 3 Mazdas before this, an '86 323, a 2000 Protege sedan and a 2007 Mazda3. I wasn't thrilled about the 3 due to the sedan construction and visibilty from the driver's position. And I could not put my golf clubs in it easily. ;-) However, I have never driven a car with a better suspension and the 2.0/auto ran strongly and gave reasonable fuel economy.

The Protege5 is a style I have always liked but is very rarely available in our area (central North Dakota). When we found this one, we decided to reshuffle our fleet.

However, 2 items of concern soon became apparent. One, the fuel economy is abysmal compared to the 3, around 20 mpg or less around town. Two, there is a knocking note to engine area. I use and have used for some time a Scangauge to monitor the engine so the fuel economy issue became apparent quickly.

Looking over the posts, there appeared to be some concern with the EGR. I pulled the EGR and attempted to clean it. This effort appeared to be futile and the EGR valve will likely need a replacement soon. The engine runs a bit smoother but there was no effect on the fuel economy or engine noise. There was no CEL prior to working on the EGR valve.

Any suggestions to items to look at relative to fuel economy other than basic maintenance items will be appreciated.

The engine noise may be a rod but for now, let's assume it is not. The noise is very similar to the popping of a single throat carburetor without an air cleaner and is proportional to the engine speed. Is there anything in the intake that would create this sound or something related to the emissions control system. This does not sound like an exhaust leak, either.

Any responses welcome and this should get the discussion started.

Thanks!

Les

 

[larson701]

Update.

When the EGR valve was re-installed, the solenoid was not re-installed with the valve. The solenoid was left hanging but plugged into the wiring harness so that the CELs would be minimized. I am getting a CEL for a low EGR flow periodically but am able to reset the CEL with the Scanguage.

I installed new plugs and an air filter and checked the compression. Based on recommendations on this board, the new plugs are NGK V-Power ZFR5F-11. I checked the gap before installation and did not have to adjust the gap. The old plugs were the BKR5E-11 and were in basically good condition. Those plugs will be saved as spares. The old plugs had been installed about 8000 miles prior to our purchase of the Protege, according to CarFax. The air filter was replaced because I wasn't sure of the condition of the existing filter. The new filter is a Wix-brand, nothing special. The net change of the plugs and filter was a possibly smoother idle in gear but no noticeable change for fuel economy or engine noise.

The compression, checked with the engine warm, was between 170 and 175 for all 4 cylinders. No problem there! The even compression values give some confidence that the engine hasn't sucked in a VICS or VTCS screw. At some point I will check the condition of the butterflies, possibly with a borescope.

The next step will be checking the cam timing. Based on the posts here and an online manual, this can be done without removing the front cover. Once the valve cover is removed, the engine will be rotated to find TDC of the compression stroke on the number 1 cylinder. Then the alignment of the marks on the cam sprockets can be checked.

Later!

 

[larson701]

Update.

Well, the valve timing appears correct. This was checked by removing the valve cover and spark plugs. By placing a rod in the number 1 cylinder, TDC was approximated and the lines on the cam sprockets appeared to line up. TDC was matched by a notch on the crankshaft pulley lining up with the 'T' on the timing belt cover.

I played with the VICS and VTCS actuators, by removing connections on the solenoids and checking the actuator action and by driving with the connections removed and connected. No significant effect on the fuel economy was obvious nor was any change in the driveability noticeable.

The noise in the engine is notable only under load (accelerating) tho not if the transmission is in gear, the foot brake depressed and the throttle applied. The noise is more notable at lower speeds and probably fades into the background of wind and road noise at higher speeds. The noise was not affected by anything done with the VICS and VTCS solenoids/actuators.

I am running out of ideas!

 

[larson701]

update

The VICS and VTCS appear to be working correctly. The VTCS has been unplugged and immediately threw a P1569 code. Performance wasn't bad once the P5 was warmed but very sluggish when accelerating to 60 mph this morning at -10F. I am not certain, but believe the transmission would not shift into overdrive (actually a torque converter lockup, I think) this morning. This may have been due to the CEL.

The MAF was cleaned last night using a MAF cleaner. While the car felt a bit smoother, the idle drops lower than I prefer when coming off-throttle up to a light or stop. The idle recovers but I wish it did not do that.

With the removal and reinstallation of the MAF, the top of the filter box seated better and the intake noise was reduced somewhat.

Sorry folks, out of ideas to try to improve the fuel economy. It is now typically sub-20 coming into work.

 

[MaxPower]

Couple thoughts:

1. The idle dip can be caused by a dirty EGR valve or cracks in the air intake tube. If it's the EGR, replacing the EGR is the best solution, though some people prefer to clean it. If it's the intake tube, taping it up works temporarily, but it should be replaced. Both the EGR and air intake problems are common with the Pro.
2. Have you checked the oxygen sensors?
3. My Protege's fuel economy dips from a high of 32mpg in summer to about 24mpg in 0F weather. Not entirely sure if this is normal.

 

[larson701]

Thanks!

1) The EGR is non-functional at this time but I will either replace the EGR or just re-attache the solenoid. One screw was wrecked during disassembly for cleaning. The solenoid is plugged into the harness but is not attached to the valve. The intake hose will be inspected closely for cracking.
2) I will have to see what can be diagnosed on HO2Ss. I have not seen much on any of the postings. The manual has very little about the sensors. I want to see if I have coding that will allow monitoring of the O2 or air/fuel ratio on the Scangauge. The coding input can take some time because there are a lot of options.
3) I did not consider my Mazda3 to be good for mileage and seeing that the P5 is worse is unpleasant.

 

[P-Funk!]

Could a cat be clogged/failing?

Bad coils and or wires?

I know you said 'assume it is not a rod' - but i gotta' say - it might have a rod getting ready to go...

Clogged injector? I would run seafoam[tm] through the gas tank and through the pwr brake vacuum line to clean her thoroughly.

Replace the EGR - it actually reduces cylinder over-temps. Doesn't make sense - but it does

 

[larson701]

Update

The EGR has been placed back into service since there may be a positive effect on the engine as long as the EGR is working correctly.

I now believe the 'intake noise' is related to the VTCS butterflies since the noise decreases tremendously once the engine is over 150F or if the solenoid for the VTCS is pulled. I may cap the vacuum lines so that the VTCS becomes non-functional but there is minor value in our very cold weather. The engine feels dead when accelerating to highway speeds and the engine is not fully warmed.

With all the work performed so far, including the checking of various engine-related components, the engine actually runs pretty good. It's smooth, accelerates well, no CELs...

The Seafoam is something I plan to do.

Looking at the way the engine runs, the fact that there are no CELs (monitored with a Scangauge) and advice that the O2 sensors will throw a CEL if there are any problems is starting to lead me to believe the problem may not be the engine. If a cat was plugged, I expect that the engine would not rev freely.

That said, I still need to check for dragging brakes but the car does coast well so I think that will be a dead end.

Transmission? It is a slushomatic with the tiptronic or whatever you want to call it. There is no switch to lock-out the 'overdrive'. If everything is not warmed up, say to 170-180F, it will not engage the lock-up torque converter. I am just fishing at this point.

Thanks for your interest.

 

[larson701]

Update

I purchased a O2 sensor removal socket, $9.99 at O'Reilly's. While the sensor looked somewhat sooty, there was no obvious damage. The sensor was cleaned with carburetor cleaner and looked a little better when done. In the past, I used a Bernzomatic torch to heat a sensor, burning off all of the contaminants. Using a torch then worked but was not done in this case.

Perhaps the engine ran a bit smoother. The idle speed did not drop as far as was the case before the cleaning so something positive came from the cleaning.

Based on what I have read in other posts, I wonder if I will notice a fuel economy change in less than one tank. There is some thought that the ECU may take a while to apply the most recent changes to the engine operation.

 

[larson701]

Update

I purchased a plug to block the VTCS solenoid outlet port, effectively jamming the VTCS butterflies in the 'open' position. Some folks here noted that this would create cold running issues. The temperature here this morning is -9F. The engine ran fine with the VTCS open all the time.

 

[larson701]

Update

Checked the engine vacuum. Steady state was about 18 inHg and the response to throttle open/close was textbook for good operation.

 

[larson701]

Update

Checked the brakes for drag. No problem noted.

Changed the ATF on the off-chance that the last change was done with the wrong fluid. This is not a 100% change, which would require the transmission to be flushed. The pan was drained and the fluid replaced to the correct level. This took 3 quarts of Dex-Merc. I will probably do this again just to dilute the old stuff, which looked good, even more.

I did the ATF change on the basis of the process of elimination. While the transmission is not likely to be the source of the poor fuel economy, there is nothing in the engine or the rest of the car that I have been able to determine is a problem. The engine runs good!

I use a Scangauge to monitor some of the engine functions but the Xgauges on the 2002 Protege5 are not nearly as uselful on the Mazda3. The number 1 thing missing is the ability to monitor the 1st bank air/fuel ratio. I really would like to see if this thing is running rich. That said, the plugs looked good tho the tailpipe is sootier than I like to see.

 

[larson701]

Update

The bellows on the intake have been taped to determine if there is any leakage between the MAF sensor and the throttle body. The bellows has some surface cracking but nothing that is thru-wall that would be the source of leakage. The car did not perform any differently after sealing that part of the intake.

I have added a can of Seafoam to the gas tank on the off-chance that the fuel injectors may be dirty. At some point, the fuel pressure will need to be checked as well.

 

[slavrenz]

Did you ever find your engine noise? I wonder if you have an exhaust leak?

 

[larson701]

I believe the noise that caught my attention is due to the VTCS butterflies. I disconnected the vacuum line and the sound is much reduced, pretty much a non-issue at this point. I had a Mazda3 and never noticed the intake noise. Makes me wonder what the differences in the engine designs may be.

Update

Based on some web cruising, I am taking another look at the O2 sensor. If I can believe what I have read, the sensor can become fouled and not trip a CEL, significantly affecting fuel economy. I cannot monitor the A/F ratio to determine if this is true.

As a result, I removed the 1st O2 sensor again (amazing how quick this is with practice). This time, I used a propane torch to burn off the residues. Also, the engine computer was re-set once again. More driving is needed but there is promise that the O2 sensor is involved at least somewhat in the fuel mileage loss. Even with the cleaning, it is possible that the sensor is still partially fouled because the shield protects the actual sensor from direct cleaning.

 

[BlkZoomZoom]

How about this one, the rattling is affected by your playing with the VTCS because your VTCS butterfly screws are loose, or missing. There is a RECALL on this.
Bad mileage? How heavy is your right foot? What are your fuel trims? What type of gas do you use? Have you tried different brand gases?

 

[larson701]

This is an early 2002. The car's history does not indicate that the intake was worked on to check for the VICS or VTCS screws. The noise I am talking about is not the rattling noted in posts regarding the screw failures. I have not seen any damage to the plugs indicating a screw was bouncing around in a cylinder. That said, removal of the intake is on my list of things to do. I may remove the VTCS permanently at that time rather than just disabling it as it is now.

My normal fuel is E10. My concern with the fuel economy on the P5 is relative to the Mazda3 we had just before the purchase of the P5. In other words, the type of fuel I use is constant. The fuel trim varies from slightly negative to slightly positive. There is no bias in the fuel trim that leans heavily one way or another.

 

[larson701]

Update

Unable to isolate anything but suspecting an O2 sensor, the car was taken to a local fixit that I have a decent level of trust in, for a scan. We discussed what they would look for. The owner had every confidence that if there is a problem with an O2 sensor, they would be able to detect it.

The result was quite different. No problem with an O2 sensor was found. The mechanic, not the owner, explained that there is a leak in the intake tube (stock 2002 P5) and was quite confident that this was creating some system problems. The other thing he noted was that because of the intake problem, the MAF sensor was reading out of range. They recommended that I replace the intake tube ($60, on order from local Mazda dealership) and return with the car for a quick scan to check if the fix was effective.

I examined the tube very closely and could not find a crack or damage. Also, I was unable to get a response from the engine when I washed various parts of the tube, including the end connections with carb cleaner. At this point, I am trusting the mechanic.

Note that there have been no CELs.

 

[larson701]

Update

In an effort to determine if an intake tube may be a source of leakage, I removed the tube, capped the ends with Saran wrap and pressurized the tube while holding it underwater. There were no visible leaks. As noted above, a new tube has been ordered. While probably not the source of the 'out of range' MAF indication, the old tube does have cracking starting in the corrugations so replacement is probably a good idea.

Taking this a step further, I studied how the MAF sensor could be tested. Short story long, the on-car method was the desired method but the Mazda's MAF connector is well sealed. The alternative is the off-car method, which is pretty simple. To test the MAF sensor while removed from the car, a voltage source (12 volt) and a digital multimeter are needed. The power is applied to the sensor and the output is measured. According to directions on the Web, a reading of about 2.5v with no air flow is normal and blowing thru the MAF should cause an increase in the signal. In my case, the no-air-flow signal was 8.08v and did not vary with air flow, indicating a malfunctioning MAF sensor (correct me if I'm wrong). This makes sense, I believe. The 8.08v signal could be seen as legitimate, tho it would be very high for almost any throttle setting, especially if the signal verification programming isn't very sophisticated. I don't know how this works with the O2 sensor in any detail. Anyhow, my impression is that this thing is probably running very rich, especially just off idle. As the engine speed rises, the mixture called for by the false signal approaches what is really needed by the engine. Or something like that.

Adding to the above, I figured out how to back-probe the connector (sharp ice pick) and check the MAF on the car. With the key in the 'on' position, the sensor voltage was about 1.49v. With the car idling, the voltage was about 2.2v. Revving the engine, can't tell you how fast exactly but in the 2-3k range, the sensor voltage would increase to about 2.7v and no greater revving would get the sensor voltage to increase further.

By pulling the plug while the engine is running, a CEL can be generated. No CEL is generated with the MAF sensor plug connected.

Anyhow, my interpretation of today's testing is that the MAF sensor is bad. Conflicting opinions?

Thanks in advance!

 

[larson701]

Update

Received a new MAF (FS1E-13-215) from an Ebay vendor. Threw a CEL right out of the box. The internal design of the old MAF looks more like a vortex-type MAF. The new MAF looks like a resistance-type.

Requested a refund or replacement MAF.

Re-installed the old MAF and everything works well. Note that previous testing by a garage indicated the old MAF was 'out of range', and recommended the replacment of the air tube between the MAF and throttle body. The tube was replaced and did not change anything.

Testing data of the new MAF showed that the MAF is somewhat different from the old MAF, right up to the point that the new MAF did not work.