Mk7 mid-rev light-throttle judder issue 1500-1700 revs

[Shankie]

Mine had a judder at mid revs, when accelerating. I replaced the n/s driveshaft and that judder has totally gone.

Hah! It's amazing how wide range of causes and solutions people are coming up with for the same symptoms! New injectors, new driveshaft etc. (Mostly) unrelated powertrain components – I guess this all points to Martyn's original statement that it's an argument in the map between the requested torque and observed torque.

[Airthies]

Judder only started after replacement of 4 injectors, swapped out two and it's been fine since.
There's also a difference in people's perception of the eronious word Judder... and that there's also a difference in engine size and drive train FWD v RWD

Sent from my anu5 using crapatalk

[DAZZ]

Funny I'd forgotten about this issue until I saw this post had been added too, but since the weather has warmed up my van hasn't done it.....

[Shankie]

Crazy isn't it. Perhaps the problem genuinely is caused by a variety of hardware issues (temperature sensing, driveshafts, injectors etc), but there's a particular part in the map that's overly sensitive to minor fluctuations in demanded torque and cause this "argument".

I'm a programmer by trade (though not ECUs!) and there's the concept of a "debounce" algorithm, that is, a function that prevents a condition which would keep fluctuating rapidly from doing so, effectively saying "I know this output isn't exactly right based on the inputs, but we only just changed 0.5 seconds ago, so let's not do it again in order to give the user a better experience". It's possible that the ECU isn't effectively debouncing between slightly different readings, causing demanded torque to 'wobble'.

This is all just guesswork as I don't really understand ECUs that well. But it would explain how different unrelated causes could manifest in the same symptoms

Sent from my ONEPLUS A5000 using Tapatalk

[Eoinom]

Today is the first day I didn't experience the gearbox Shudder (fingers crossed)

Mine is a 07 Mk7 110T300 5 speed 2.2 and all the same issues.
Hesitation at 1500-1800RPM depending on gear before engine gets up to temp almost like slight fuel starvation.
Gearbox Shudder at 1700RPM 50MPG 5th Gear if engine is not up to temp
Hard to start if sitting for a week.

Cold Weather:
Symptoms mainly when engine is cold
Symptoms worse in cold weather(5C) and do not go away under 0C.
Engine very Hard to Start if temp below 0C, eventually coughs, splutters and starts in a cloud of black smoke as if the fuel had waxed up, but engine does eventually start and keeps running but is noisy and clattery like an old vw bus until warm.

Other things I've noticed:
Fuel Guage not working - sometimes shows eratic readings if driving over a bumpy road.
"soot" smell recently started coming from exhaust when cold with a small amount of black smoke when cold - nothing worse than my landy td5 used to do.
Slight engine hunting (3 or 4 times) if clutch peddle pushed down at 1700 rpm.
Recent Fuel Starvation - 5l of diesel from a can in the back didn't help... limped it to a garage 10 miles away, filled up and eventually the van started and ran without a problem - this was either related, or my fuel consumption dropped massively and I ran out of fuel.

All of this only started since I've had my 4 injectors replaced.
Since then I've also had the EGR, EGR Cooler and Injector Rail seals replaced.

IDS showed problem with the fuel sender unit - 2 codes... sorry have lost the sheet - it was diagnosed in Germany with a bad translation as "the fuel pump in the tank has no signal" but okay to drive to Ireland where of course the symptoms go away with the heat.

After getting a stern talking to from a long-term transit driver over the risks to the fuel pump of contaminated fuel, especially with 195k milage, the fuel starvation incident and the heatwave finishing meaning the symptoms returning I eventually bring it to my mechanic back home last week. When they shake the van the voltage changes and the fuel gauge starts to move erratically, but not when they trace the sender wire. Straight away he suspects the BCM - checks it with another 2.2 and finds a voltage difference and thinks there's signs of water damage.

BCM replaced, reprogrammed and the fuel gauge now shows full and the miles to zero shows a number. Fuel tank dropped, drained and cleaned. A quick check of the diesel that came out of the tank and there was black bits of crud at the bottom and a cream "jellyfish" of slime about the size of a jam jar lid floating about about an inch up from the bottom. There was also some slime around the fuel sender unit base and the float arm was rusted in the upright position.

All cleaned out, 20l clean diesel put in and the engine started first time, no smoke and sounded so much smoother than before. No errors in IDS, but they did notice a fluctuation in the fuel pressure - which they think might be the pressure sensor.

Anyhow they've given it back to me to drive - I have to drop it back as they had to do a bodge job replacing the arm on my existing fuel sender unit as the replacement was faulty. They also want to see if the pressure readings bed in with a bit of normal driving, and they didn't have one in stock to test.

I'd asked them to do a fuel pump learn, but I'm not sure if they have - I think they were just glad to get the space back in the garage after sitting there for a week. I just got my modified ELM today and downloaded forscan but I'm a little nervous about going in and doing any of the tests as a first time user. But guess what.... it drove great today. no hesitation, no shutter, and only the normal blip of the engine when you push in the clutch - plus the engine seems to be running much quieter and smoother.... quiet enough for me to now start hearing other noises so now a new hunt begins!

My mechanic reckons that a mix of the fuel sender not priming properly could've caused the hesitation and the sludge in the tank could explain some of the starting problems and why the problems are worse in the cold weather. Only time will tell!

[martynx]

Crazy isn't it. Perhaps the problem genuinely is caused by a variety of hardware issues (temperature sensing, driveshafts, injectors etc), but there's a particular part in the map that's overly sensitive to minor fluctuations in demanded torque and cause this "argument".

I'm a programmer by trade (though not ECUs!) and there's the concept of a "debounce" algorithm, that is, a function that prevents a condition which would keep fluctuating rapidly from doing so, effectively saying "I know this output isn't exactly right based on the inputs, but we only just changed 0.5 seconds ago, so let's not do it again in order to give the user a better experience". It's possible that the ECU isn't effectively debouncing between slightly different readings, causing demanded torque to 'wobble'.

This is all just guesswork as I don't really understand ECUs that well. But it would explain how different unrelated causes could manifest in the same symptoms

Sent from my ONEPLUS A5000 using Tapatalk

"Hysteresis" control its called Shankie. You're spot on

[DAZZ]

And the first cold snap.... had frost on the windscreen this morning and lo and behold the judders are back pmsl......

[Shankie]

And the first cold snap.... had frost on the windscreen this morning and lo and behold the judders are back pmsl......

Haha oh no!

Mine have been there all summer, but it definitely seemed worse this morning as well.

[Bluevanman!]

My Mk7 110 dose this Judder thing from time to time.

If its been running a while it doesn't do it.

Never did it until I updated the PCM. But then reading all the other comments it could be anything.

Something that hasn't been mentioned is If it starts to judder try switching on the Cruise Control, if you have it.

Tried this when it was in Judder mode one morning Activated Cruise control judder immediately stopped.....

Could be a Throttle pedal issue??

Another suggestion to add to the many haha

[Eoinom]

Its 3 weeks since I replaced the BCMii; dropped and drained the Fuel Tank; and did Pump and Injector Pilot Relearns and none of the symptoms have come back... The vans driving smoother and quieter than ever; starts on first turn of the key with no smoke; and no sign of the shudder, or hesitation around 1700RPM.

[twin--turbo]

I'm a programmer by trade (though not ECUs!) and there's the concept of a "debounce" algorithm, that is, a function that prevents a condition which would keep fluctuating rapidly from doing so, effectively saying "I know this output isn't exactly right based on the inputs, but we only just changed 0.5 seconds ago, so let's not do it again in order to give the user a better experience". It's possible that the ECU isn't effectively debouncing between slightly different readings, causing demanded torque to 'wobble'.

Debounce/Hysteria/Flapping/Governing...

Been around since the steam age ( literally ) and mechanical engineer should take this into account .

TT

[Ronald53]

.... and did Pump and Injector Pilot Relearns and none of the symptoms have come back...

I am interested to know if the injector pilot setting needs a special procedure to do the learning. I meanwhile reset the learned parameters for the injector pilot setting twice. For me, this helped to reduce judder. But what is the best method for learning? I didn't see a learning option in Forscan.
Kind regards,
Ronald

[yl05nto]

It's just the way they are don't waste money on this fault.just drive round it i spent.way too much trying to fix it and never did if u want my opinion it built into the factory map.

[DAZZ]

It's just the way they are don't waste money on this fault.just drive round it i spent.way too much trying to fix it and never did if u want my opinion it built into the factory map.

Yep, had my van remapped at weekend with an economy map, spoke to the chap doing it, he interrogated the ecu, my van had no stored or current faults since software update 2 (ish ) yrs ago.
I just drive round the issue, still does it remapped, but..... doesn't seem as bad???
He did say the vans 'program' is quite basic and crude and it did sound like an issue with the software having an issue calculating an algorithm with the set of perameters its had to deal with at that time.
strange it's worse in cold weather..... I do wonder if theres a temp sensor some where I could 'frig' with a resistor or something to say its warmer than it actually is....

[Shankie]

I was wondering whether a remap might make a difference as, if it's a debounce/hysteria issue, it all it might take is a slightly different torque request at a certain point to sort it out.