RPM drops, occasional DSC light on (2014 Mazda5)

[hleboklop]

2014 Mazda 5 AT, 88K miles, mostly short city drives all the time.

Recently, I noticed that the RPM is dropping to around 750 almost always when the gas pedal is not pressed.
The issue is not happening when I am driving with cruise control engaged, - the car goes normally.
But as soon as I drive without cruise control and release the pedal, the RPM goes down - even when I'm rolling downhill and not pressing the gas, which I find weird.
Usually, when I take off, AT switches to the second gear and after that, I release the gas pedal and keep rolling - the car goes on. Now, however, it seems to drop RPM and after that, it's switching back to 1st (it seems).

Feels like a drag, like brakes are stuck - however, the brakes are not overheating, car is going straight and not pulling to either side, no vibration when braking, no burning smell coming from the brakes. The rear brakes are almost worn out, though. Brake fluid is less than 1 year old, and I'm am almost sure there is no issue with brake lines, since the fluid level is normal.

The air filter is fresh new clean, MAF sensor is clean - I just checked and cleaned it, just in case.

I have an OBD2 sensor constantly hooked up to my cheap Chinese Android stereo, which makes it possible to monitor various parameters while driving.
The only thing I can see that is slightly off is the Long Term Fuel Trim (LTFT) is 7.5%, which suggests that the engine is running lean, which can indicate a possible vacuum leak. However, it's still within the normal range, so, I am not sure about this being the reason.

I recently changed the engine oil (DIY 1st time), however, I don't see any leaks and even if the oil filter is completely clogged, this would result in such condition, I think. The oil level is just fine, I don't smell anything burning.

Also, used the snake oil fuel cleaner liquid (techron) recently and went for a long highway trip just fine.

Hours of research suggest that the torque converter or transmission is having an issue - i.e., slipping and going out, however, I don't hear any noises from that department.

On top of that - no CEL, no codes, nothing!
The only indicator that came up today is the DSC light, which blinked a couple of times and then got stuck until I did a stop engine start engine - which makes me believe that I should change brakes and rotors because these were changed 4 years / 34K miles ago.

So, I'm thinking about replacing the brakes and servicing the calipers now. I am still not convinced that there's a vacuum leak, however, I'm going to double-check this.

Has anyone ever experienced a similar issue? The only topic that is slightly related that I could find is this.
Any ideas?

 

[hleboklop]

update: this is nothing related to the brakes. If calipers ever would get stuck, the noise and pull would be noticeable.
Driving around while collecting data from intake airflow, throttle position, long term fuel trim (LTFT) indicates that this must be somehow related to either a vacuum leak or a problem with MAF or TPS sensors.
I'll swap the sensors, check and clean the TB and will see what changes.

Interestingly, after unplugging the MAF sensor, starting the car, failing to start with MAF unplugged and then re-plugging in the MAF sensor, I was able to briefly make the car run "normal" - without throttle getting closed asap. Although, this brought up the CEL light, which I cleared and returned to the starting point.

 

[hleboklop]

Update: swapped 2 different MAF sensors - no change, still low RPM at idle 650 or so, throttle closing shortly after gas pedal is depressed.
Next check - throttle body, leaks, see if changing the TPS sensor can help.

 

[negusm]

Does this car have an IAC? If so, clean it and the throttle body and if needed, replace the IAC. Also suspect the TPS like you mentioned.

 

[hleboklop]

Changing one parameter at a time:
cleaned the throttle body, as it was quite dirty, replaced the O-ring, cleaned the intake, checked for any cracks or holes - everything is more or less okay. 1 air filter clip is broken, though, and the TPS sensor screws are now not super tight.

The RPS doesn't drop as much as it used to. I would even say, this condition is gone.

The only problem I'm observing now is a very low, <600 RPM, at idle. For example, when stopped at the red light, with the shift in D, regardless of whether the AC on or off.

ECU is surely considering the current air-fuel mixture as lean: LTFT is never around 0 - the lowest I observed was around 5, mostly it's between 6 and 10; even 16 sometimes. As far as I remember, this car have always been running with a very low MPG - I always thought that this is due to driving pattern, which mostly consists of short infrequent trips.
And the spark plugs look like the engine was running rich all the time.

My next step is replacing the TPS.

Is it a hard job?
Can this be done without disconnecting the throttle body from the coolant lines?
Any piece of advice you can share?

 

[hleboklop]

Wrapping this up: after swapping various parts and checking everything I came back to where I should start: CLEANING THE THROTTLE BODY

Cleaning the throttle body, resetting the computer, and doing a drive cycle resolved the issue of RPM dropping after the gas pedal was released.
While at it, I also changed the spark plugs since they were pretty fouled. I think this is due to this particular car's driving pattern: lots of short trips, idling, and probably not warming up most of the time.

Things I learned:
- FORSCan on a PC with a proper OBD2 cable is super useful for diagnostics
- FORSCan's "Mode 6 scan" can show when the car has completed the full drive cycle - mode 6 scan results must show "o2 sensor ready" to pass SMOG, for example, while not being ready for evap is not a deal breaker - at least, in California

There's still a vibration at stops when the RPM drops to around 600, but, I think this is normal for a 10-year-old car. The rear engine mount must be pretty worn, probably.

 

[Lazy2.5]

The RPM's dropping, such as when rolling down a hill is NORMAL. This is the torque converter unlocking. Do not worry about it!

 

[Lazy2.5]

Your LTFT are for sure too high, though i'm sure its "normal" considering the fuel quality we have here in NA and the air quality in the city, short distance driving.. What do your STFT's look like? Do you know what your AFR ratio and ignition timing looks like?

 

[hleboklop]

The RPM's dropping, such as when rolling down a hill is NORMAL. This is the torque converter unlocking. Do not worry about it!

This one feels like "car's downshifting", I noticed this behavior before and read about it.
The problem described in the first post felt different - at first, it felt like somebody was stepping on the brakes right after I released the gas pedal, even when rolling on a straight-level road.

After replacing the spark plugs and driving around for a couple of days LTFT dropped to 0-4 and the vibrations at the stop with D are gone too!

I didn't check those yet, will drive around collecting more data, and also, will try to compare this to data I collected before all this nonsense started. (I was experimenting with torque reporting to the web server, I set up my own server as I was planning to help a student build a product prototype for a school project.)

 

[Lazy2.5]

This one feels like "car's downshifting", I noticed this behavior before and read about it.
The problem described in the first post felt different - at first, it felt like somebody was stepping on the brakes right after I released the gas pedal, even when rolling on a straight-level road.

After replacing the spark plugs and driving around for a couple of days LTFT dropped to 0-4 and the vibrations at the stop with D are gone too!

I didn't check those yet, will drive around collecting more data, and also, will try to compare this to data I collected before all this nonsense started. (I was experimenting with torque reporting to the web server, I set up my own server as I was planning to help a student build a product prototype for a school project.)

Have you changed your transmission fluid and filter?

 

[hleboklop]

Have you changed your transmission fluid and filter?

Yes, the transmission fluid was changed about 8 month ago, OEM at the dealership.

 

[Lazy2.5]

Yes, the transmission fluid was changed about 8 month ago, OEM at the dealership.

The filter?

 

[hleboklop]

The filter?

No, they didn't change it. And by that, I mean that they didn't remove the transmission pan - they just flush or replaced the fluid.

As far as I know, "the filter" in the transmission is not a filter, but rather a magnet?
Do you think this have something to do in this case?

 

[Lazy2.5]

No, they didn't change it. And by that, I mean that they didn't remove the transmission pan - they just flush or replaced the fluid.

As far as I know, "the filter" in the transmission is not a filter, but rather a magnet?
Do you think this have something to do in this case?

The transmission has a filter/strainer too, not just a magnet.
If your transmission is not behaving the way you want it to, have the pan dropped, cleaned out, filter replaced. I think it'll solve your problem.

 

[Silentnoise713]

When low idle occurs, shift to Neutral, does it go away?
Edit: Some additional thoughts from my experience.

Restore lost MPG

I've been getting unusually low MPG the last few fill ups, avg. 15MPG city. I don't recall it being this low but shrugged it off. Replaced the fuel injectors a short while back with reman. ones (also rough idle) to see if it might kill two birds with one stone - not really any different. Then...

mazdas247.com

'Suspect' the torque converter is worn (internal leak, not holding pressure). Not much you can do except replace it with a new torque converter or live with it (prevent it from getting worst). Periodically replace trans fluid to keep it fresh, this trans is known to run hot, which is the crux of the issue.

The OEM trans filter is not going to do much. It is mean to filter out 'larger' particles. Search BITOG for cutout visuals. If you must, you can easily add an inline magnetic filter by intercepting the trans fluid before it goes to the trans cooler. Note this will add a service item but it is easily to replaceable if you mount it in accessible spot. OEM pan has a small donut size magnet on the bottom to collect ferrous particles. It would be good to clean it but not absolutely necessary.


Does FORScan allow monitoring/control of torque convert clutch status? If yes, during the low idle scenario, try setting it lock vs unlock. This will help to confirm it.

 

[Lazy2.5]

At that mileage, you'd want to see what the inside of the pan looks like. I'm sure it will be covered in dark sludge. Once all of that is cleaned out, along with the magnet, and a clean filter is installed, the torque converter will perform significantly better.

 

[hleboklop]

I agree, the situation when the vibration happens on stop in D but is gone when switching to N does look like a worn torque converter. Last weekend I took a car for a spin on a highway and in the mountains and everything was super smooth and perfectly matched the expectations.
I had, however, one episode of transmission stutter when rolling up the hill, where it looked like the downshift didn't happen at the right time.
I'm still thinking about the worn engine mounts, as I can hear the slight "rubber" noise when switching from D to R to D, for example.

What is your opinion about installing an inline canister-type filter? Does it reduce the ATF flow significantly?
I've seen those modifications, I was even thinking about installing filter + transmission cooler since we live in a relatively hot place with lots of sitting in traffic.

 

[Silentnoise713]

I agree, the situation when the vibration happens on stop in D but is gone when switching to N does look like a worn torque converter. Last weekend I took a car for a spin on a highway and in the mountains and everything was super smooth and perfectly matched the expectations.
I had, however, one episode of transmission stutter when rolling up the hill, where it looked like the downshift didn't happen at the right time.
I'm still thinking about the worn engine mounts, as I can hear the slight "rubber" noise when switching from D to R to D, for example.

What is your opinion about installing an inline canister-type filter? Does it reduce the ATF flow significantly?
I've seen those modifications, I was even thinking about installing filter + transmission cooler since we live in a relatively hot place with lots of sitting in traffic.

From what I learn, worn torque converter leads to worm clutch, which creates micro ferrous partials. If left unchecked (unchanged), they end up in the trans and clog up check balls/valve bodies, and build up and affect EPC/shift solenoids (electro 'magnetic' switch) function.

I think the inline filter would be a great add on. You can always run one for a period of time, then remove it. I suggest you to contact the mfg. for flow specs or post your question in BITOG for unbias opinion.

Don't think an external cooler is necessary but I don't know what is your definition of 'hot'. Do flush your coolant!

 

[hleboklop]

<..>

Does FORScan allow monitoring/control of torque convert clutch status? If yes, during the low idle scenario, try setting it lock vs unlock. This will help to confirm it.

From what I observed over a brief drive monitoring session, it does engage/disengage as supposed to - see the OSS graph (orange). The big flat space between 100 and 128 is idling at a stop.

Is there anything wildly abnormal, that should further be investigated in your opinion?
I would greatly appreciate any advice on which parameters to group on this graph that would help to understand what is going on.

Also, I tried feeding this data into chatgpt4o and, to my surprise, it outputs something that now actually makes sense, although, the recommendations are very generic, i.e., "note whether a change in LPSA LPSB correlates with gear shift, look for random spikes on OSS, etc".

 

[hleboklop]

From what I learn, worn torque converter leads to worm clutch, which creates micro ferrous partials. If left unchecked (unchanged), they end up in the trans and clog up check balls/valve bodies, and build up and affect EPC/shift solenoids (electro 'magnetic' switch) function.

I think the inline filter would be a great add on. You can always run one for a period of time, then remove it. I suggest you to contact the mfg. for flow specs or post your question in BITOG for unbias opinion.

Don't think an external cooler is necessary but I don't know what is your definition of 'hot'. Do flush your coolant!

I will also look into installing the inline filter. Doesn't look too hard so far.