Technical Bad Idle, Judder and Brake revving

[DTanner]

I have had some intermittent problems with my Panda for a while, but recently they have gotten worse so I have looked more closely today. It is a 1.2L Fiat Panda Dynamic Eco 2009 with 110k miles on the clock.

Main issues I see are:

1. Idle is not steady – RPM can drift around a lot and generally is higher than expected (e.g., idle on a warm engine at 1,500 RPM or more)
2. Pushing the brake causes the engine RPM to increase (or can cause it to begin to stall if I pump the brake fast enough)
3. Sometimes the car can periodically judder - like a little kick you feel while driving – this is indiscriminate hot/cold, underload or not, any gear, driving fast or slow
4. The EML has been on and I got this code: P0120 - Throttle Pedal Position Sensor/Switch (TPS) "A" Circuit Malfunction

So where has my investigation gotten me? I am building data but need help figuring it out. This is where I am:

For issues 1) Idle & 4) P0120 TPS:
I bought and installed a new throttle body as the throttle position sensor is integrated. Bought a MAGNETI MARELLI 802001291204 Throttle body (see image) so did not skimp on cost and replaced the main gasket between the throttle body and engine. At the same time I replaced the spark plugs (they needed replacing anyway) and the battery (it was 7 years old and, though ok, the voltage was at the lower end of the range when disconnected and on vehicle start).

Two points with this:

• The two middle spark plugs appeared fouled. I could not really work out if it was oily or carbony – see picture above. The sheen made me think oil but they did not really seem wet. They are shown in position order (e.g., left most plug sits in the left most position in the engine)
• With these changes the car was brilliant and sharp and ran like new. I had 3 great days (maybe five 20 mile trips). Then it drifted back to old behaviour as listed above, though the high idle was always off the driving performance seemed much better and no juddering. Following this I replaced the MAP sensor (Bosch 0261230030 Map Sensor) but no luck - same idle behaviour.

For issue 2) Brake Revving:
I suspected a vacuum leak to the brake system. I don’t know much about this but was expecting that the brakes use vacuum assist from the engine and that a fault was leading to air getting pulled back into the engine, thus affecting the idle. I looked for splits or wear and sprayed a good amount of “easy start” around but did not find a leak. See the picture of the engine bay above, I have added an orange line next to the black tube I was hunting around, is it the correct one? This tube has clear marking saying "Lancia Fiat Alfa ... Hutchinson P008...". Anywhere else I should look? Any other ideas for this issue?

For issue 3) Judder:
I have no new ideas for this, I am hoping it would go away if the other issues were addressed as I don’t have good ideas for this. In the back of my mind I am wondering if an engine gasket leak is leading to oily plugs and intermittent loss of spark (and hence judder/kicks), but I hope not. I don't have any major oil leaks (i.e., nothing on driveway), the exhaust is hard to see so not obvious that oil is being burned off, but there is oil about the top of the engine and throttle body (but not massive amounts). Might give it a bit and then pull the plugs again to check.


Any thoughts or suggestions? Sorry, I know there is a lot going on in this post!


===============================
Some more info for those interested in live data:
===============================
This is the car now with a warm engine (after running for 30 mins) and sitting idling in neutral - looks like a control loop. Every 40 seconds there is a click from the engine, the manifold air pressure drops and the RPM drops accordingly, then returns to an elevated level (around 5 seconds later). I assume this is normal behaviour but the idle should be 800ish. The graphs looked the same with the new MAP sensor. I did a quick check with the MAP sensor disconnected and the engine idled badly / lumpy between say 500-800 RPM.
I was also monitoring the throttle position sensor data but it appeared fine - this is not shown but I ran it up and down from 0-100% and it looked fine.

This is the mess when I am pumping the brake (say pushing and releasing the brake repeatedly at about one cycle every two seconds). I start pumping just after 20 seconds on the plot below and stop at around 50 seconds (outside this is pure idle). You can see the O2 data as well. Is sensor 1 supposed to swing between max and min (see between 60 and 120 seconds, this is just idling in neutral)?

 

[koalar]

You have an error code for the accelerator pedal in a flyby wire system and have several issues with the high revs and sudden revs dropping

I’d be starting here, reading the live data from the pedals two tracks.

The oil on the two centre plugs looks like it’s coming from the top

Cam cover leaking, spillage, breather pipe off or similar.

 

[koalar]

Ignore above. I should have looked at the photos closely which clearly shows a cable throttle

I don’t understand why

I have a 1.2 dynamic eco

The coil pack is a single unit under the airbox

The throttle is flyby wire as was my 1.2 05 dynamic


Anyhow no matter there two types of cable throttle and most still have a potentiometer built in them. Which is what the error code is complying about.

 

[koalar]

 

[koalar]

For issue 2) Brake Revving:
I suspected a vacuum leak to the brake system. I don’t know much about this but was expecting that the brakes use vacuum assist from the engine and that a fault was leading to air getting pulled back into the engine, thus affecting the idle. I looked for splits or wear and sprayed a good amount of “easy start” around but did not find a leak. See the picture of the engine bay above, I have added an orange line next to the black tube I was hunting around, is it the correct one? This tube has clear marking saying "Lancia Fiat Alfa ... Hutchinson P008...". Anywhere else I should look? Any other ideas for this issue?

eliminate by unplugging and blocking the end with your thumb

 

[koalar]

The correlation between revs and pressure is normal

As the revs drop the pressure should rise towards atmospheric

 

[koalar]

Every 40 seconds there is a click from the engine, the manifold air pressure drops and the RPM drops accordingly, then returns to an elevated level (around 5 seconds later). I assume this is normal behaviour but the idle should be 800ish.

Correct

Can you feel a vibration from the idle control valve as the click happens.

 

[A3jeroen]

How is the checkvalve in the line to the brakebooster?
Any other vacuumleaks? Inletmanifold (especially around cil 2 and 3) Ok?

gr J

 

[DTanner]

Thanks @koalar and @A3jeroen,

Yep, my car has a throttle cable so I assume there is nothing to look at on the pedal end (e.g., sensors etc.) just the TPS on the Throttle Body end. I tried a throttle relearn procedure, even though I am not sure if it is needed (I used the procedure mentioned in this: https://www.fiatforum.com/threads/throttle-acts-up.269804/).

@koalar, thanks for looking at the telemetry and marking it up – your comments are in line with my expectations.

Sounds like I have some actions that I will undertake and report back:

• Blocking the vacuum port to the brake booster makes a lot of sense, to isolate it and see if the issue persists
• Will look for a check valve in the brake booster line. Is this integrated into the hose or part of the brake booster itself? I assume it is a passive mechanical part (i.e., a check valve like an umbrella or duckbill valve).
• Check the idle control valve click on the throttle body, just for interests sake
• Check the inlet manifold to engine interface for an air leak (I saw a video on this today that showed such a leak) and generally look again for any other leaks

Thanks, will let you know what I find but might not get to this until the weekend.

 

[Goudrons]

The oil on the middle two plugs is on the threads, so I would guess oil is leaking down the back of the cylinder head and running down the threads from the outside.

As you appear to also have a air/fuel imbalance, what is the condition of the oil breather pipe?
The black pipe in the centre of the image, from the cam cover to the airbox (with a small pipe off it to the throttle body).
Check this carefully as it can split.

When checking the brake servo line, carefully check the ends of the line.
They are hard plastic to stop them deforming under vacuum but this can make them brittle as they get older.
I haven't had a Panda one crack but have on another car. The end of the pipe, where it's flaired to fit the barb broke up.

It's worth starting at one end and working back, if you can reach, block it off at the manifold and test, if it still does it then it's not leaking further back on the pipe or servo and you can move on. Other wise, start working back.

 

[DaveMcT]

It might not be coincidental that plugs 2 and 3 are oily. The coils fire the plugs in pairs 1 + 4 and 2 + 3. If you have not already done it, swap the coils so they each connect to the opposite pair of plugs and see if the misfire moves.

 

[DTanner]

Thanks @Goudrons, I have had the breather pipe off last weekend and a good look at it. It was in really good shape. There was some crusted oil in the the connection to the cam cover and there was oil around the small barbed connection into the inlet manifold (think that is where it goes) but nothing that looked concerning. The pipe itself was pretty solid. As for the harder brake booster line I will try and take this out for a good inspection. I have tried to find a leak but nothing obvious (i.e., spraying easy start on it with the engine running). However, I could do with getting a better look. First step will be to remove the connection at the inlet manifold and block it as suggested above.

@DaveMcT thanks. This is a nice link to be aware of. I will keep an eye on this and see what happens. It is not easy for me to detect a misfire - I only feel it when driving (car jolt) but not when it is stationary on the driveway and it is not all that frequent. In the car I cannot see or hear it (e.g., there is no bang or other event that I can see or hear). Given this I do not think I would be: 1) able to detecting a misfire generally and 2) would not be able to assign it to a cylinder pair. I will think about this though and knowing that a single coil fires the two middle plugs is useful!

 

[koalar]

Thought there might have been some progress or other suggestions

I see the throttle cable is disconnected. Was the original throttle body acting the same

Is there live data for the throttle position. I find it strange you have a P0120 error but no potentiometer

What’s the coolant live data like ?

You can try swapping the pre and post cat O2 sensors over

If you disconnect the MAP while it’s running what happens

Small air leaks normally don’t effect a MAP only system much as it’s still seen by the sensor

The only place of concern is post MAP sensor which is the joint between the inlet manifold and cylinder head. 2 minutes spray with water should eliminate this. Very rare and up to now only associated with an overheating engine.

 

[DTanner]

Hi @koalar, sorry I have not been able to look at this as my wife is away so managing work and the kids is enough until the weekend. Then I will work on this and provide some updates. However, it is really helpful for me to cultivate a list of areas to focus on when I hit the weekend.

Changing the throttle body had no noticeable effect, though (as mentioned) the car did feel great for a couple of days after this change (though the battery and spark plugs were also changed at the same time).

I have no idea what the error code relates to, I assume it is the circuit for the TPS on the throttle body – so could be the connector or wire harness. I will have a look at these. I did look at Absolute Throttle Position % live data and it seemed fine, i.e., I slowly varied the throttle with my foot and there was good movement and no spikes or jumps (though the acquisition rate is very poor, i.e., 1 Hz ish). You can see it on the graphs above for idle where the line for Absolute Throttle Position (%) is flat as expected. I have had a look through my data from the last check I reported on and found this – the car is idling so there should be no signal variation on the throttle position but it does fluctuate (see image below). The fluctuation is < 1% but I will try and collect data for longer to catch any larger spikes as the judder issue is intermittent.

I have not looked at coolant data so will see what I can do for this. I have never seen the dashboard temperature sensor go above the mid-range mark. However, it is a while since I changed the water pump. I will also swap the O2 sensors as you suggest to see if symptoms change. I have replaced these in the past, but again many years ago now.

I did electrically disconnect the original MAP sensor and record some data. The car idled really lumpy and felt like it could cut out (see image below). I assumed that it was not intended to run without the MAP so move on to other checks. Is this significant? How would you expect the idle to be with an electrically disconnected MAP?

I will definitely be checking the joint between the inlet manifold and cylinder heads. I will spray it plenty - I have not checked this previously.

Thanks.

 

[koalar]

Great

Interesting. With the MAP disconnect the high revs have gone

Lumpy idle is correct. It’s running like a fully functioning car with the map disconnected.

It can’t read the data because we unplugged it so is running a default set of parameters just trying to keep the car running the best it can

Confirms one of the sensors is giving or getting the incorrect data, and it’s running on the wrong map. Unfortunately as of yet we don’t know why.


I asked for a trace of the coolant sensor because I have seen these blip/spike occasionally to some crazy temperature something like 6522C for a fraction of a second but it’s enough to confuse the ECU

P0120 is normally a throttle pedal sensor. It’s a common fault on cars that kangaroo and flyby wire throttle. No idea why you have it if you don’t have a throttle pedal sensor.
Are you sure you pedal doesn’t have an electrical connection Looked at the wiring diagram for the cable throttle 1.2 with idle control valve and it shows one on the pedal. But I couldn’t find a picture of one online

 

[Goudrons]

P0120 is the Throttle Position Sensor.
There is a pair of them on the throttle body. Circuit A and Circuit B.
The system tests it's self against both signals and if it isn't right it will flag a code.

P0120 is Circuit A below or above expected OR isn't matching Circuit B.

Modern fly by wire cars would have a pair of potentiometers on the pedal to electrically work out it's position and send that to the throttle body to drive it's action.

With a non FBW setup, the pedal is physically connected to the throttle body so it doesn't need to electrically work out the position of the pedal to drive the throttle so there's no position sensor on it. The ECU just needs to know the position of the throttle.

Earlier Pandas had cables, later Pandas were FBW.

Since the you changed the throttle body has the code been reset and not returned?

To me, in your first image on your last post (post 14) the ATP has risen and fallen at around 260 but the rpms aren't really following suit. Which kind of suggests the signal the ECU is getting and what the throttle is actually doing aren't really related.

As the old TB gave you a position sensor code, I'm wondering if you have a wiring issue between the TB and the ECU.
Is something screwing with the TPS readings as you have throttle position problems on two different TB's??

You thought the first TB was faulty, but is might be it was ok, same as the new one but the problem could be in the wiring or ECU it's self.

What about logging that last test again yet wiggle the loom and tap and wiggle the connectors to the ECU?

 

[Goudrons]

To me, in your first image on your last post (post 14) the ATP has risen and fallen at around 260 but the rpms aren't really following suit. Which kind of suggests the signal the ECU is getting and what the throttle is actually doing aren't really related.

The Inlet Manifold Pressure hasn't changed either.
If the throttle opened at 260, you would expect the manifold pressure to change and the rpms to change but they haven't.

I guess you've not yet seen the P0120 return because you haven't really driven it enough for it to flag.
A lot of codes only set when they are repeatedly detected in the same drive cycle, often three or more times before they set a DTC.

 

[koalar]

P0120 is the Throttle Position Sensor.
There is a pair of them on the throttle body. Circuit A and Circuit B.
The system tests it's self against both signals and if it isn't right it will flag a code.

P0120 is Circuit A below or above expected OR isn't matching Circuit B.

Here what what’s displayed on the Puntos 2015 dash display

P0120-14 – Accelerator pedal potentiometer 1 – short to ground or open circuit

The internet often confuses the two as they can be both shortened to TPS

There several threads on here where blowing compressed air or reseat the connector on the pedal has fixed this error

Elearn and a list of fiat specific codes both list as pedal

 

[koalar]

If the throttle opened at 260, you would expect the manifold pressure to change and the rpms to change but they haven't.

Isn’t it a flat line just zoomed in. The graph full scale is only showing 1 percent throttle opening not 100 percent. I have never zoomed in that much so don’t know if it’s normal.

 

[Goudrons]

It's the throttle postion sensor that is on the throttle body as there is no pedal position sensor on this model.

If you think, both A and B circuits on a FBW car would be linked. Input from A at the pedal should match A at the throttle.
Same goes for B.

If there is a problem on either circuit it could be at either end, pedal or throttle end it would still think there is an issue with the throttle position.
With the right software you should be able to drag the extended DTC out of the ECU which would perhaps point to which isn't right.
There would be 4 values, 2 at the pedal and 2 at the throttle so in most cases it should be able to spot the one that's incorrect.

But there's no pedel sensors on this model only the throttle position sensors so using a car with FBW is like comparing apples with oranges.
The link between pedal and throttle is linked via cable, so there is only 2 position sensors on the throttle body linked to the ECU.
Now what has happened is the ECU has seen there's a problem with the throttle position, A as under or over reading or doesn't match B.
There's no other sensors involved, just the two differing readings from the two sensors on the throttle body.

I understand the value of the absolute throttle position is small, but these sensors alter a signal voltage of 5v and return the results back the the ECU as low as 0.1v, so I would expect some change in manifold pressure even if the throttle opening is so small, it would change the pressure in there and show some small change, but there isn't any.
Plus if there's no input on the throttle, where is this throttle position signal coming from?

I think the OP still changed the throttle body to solve a throttle position sensor fault which caused some running issues.
As this has failed to solve it, it's not impossible but it's likely it's not the throttle body, throttle position sensors or the pedal position sensors (which is doesn't have) which is why I said look at the wiring.