Engine Knock

This is the Forum for all your Peugeot Technical Questions, Problems or Advice.

Moderator: RichardW

User avatar
white exec
Moderating Team
Posts: 7445
Joined: 21 Dec 2015, 12:46
Location: Sayalonga, Malaga, Spain
My Cars: 1996 XM 2.5TD Exclusive hatch RHD
1992 BX19D Millesime hatch LHD
previously 1989 BX19RD, 1998 ZX 1.9D auto, 2001 Xantia 1.8i auto
and lots of Rovers before that: 1935 Ten, 1947 Sixteen, 1960 P5 3-litre, 1966 P6 2000, 1972 P6 2000TC, and 1975 P6B 3500S
x 1752

Re: Engine Knock

Post by white exec »

If the car had a DMF originally, and this has been replaced with a (conventional) SMF, I wonder if this has been done properly. The SMF should be solid, but the clutch friction plate should be centrally sprung. Wonder how well the job was done?

If an SMF kit (usually 4-part: solid flywheel, sprung friction plate, clutch cover and thrust race) was initially fitted, and then someone replaced just the friction plate, then I wonder too whether they took the trouble to obtain the identical friction plate that sat in that 4-part kit, or whether they might have used something else?

A judder on a new clutch (lining not yet polished, or careless oil/grease on the parts) is not unknown, but should disappear very quickly.

_________

Listening to your video soundtrack again, I get about 38 clonks in 20 secs, or 114 clonks/min.
At a tickover of 800rpm, that would be 400 power strokes/min for any one cylinder. Correct?
So there aren't enough clonks for one duff injector...
Has to be something else.
Could be wrong.
Chris
ekjdm14
(Donor 2020)
Posts: 1810
Joined: 19 Jan 2015, 17:42
Location: Congleton
Lexia Available: Yes
My Cars: '95 Xantia 1.9D automatic - 109k one of two remaining
'03 206 GTi180 - 97k in pieces being brought back up to snuff
'03 206 1.9TD - home-built veg-fuelled B-road bruiser
'99 306 SE 2.0i cabriolet - 92k awaiting service work ready for next summer
'97 306 XS 1.6i 3 door, Bianca White, from dead 12 years & seized solid to road legal in 7 days
x 212

Re: Engine Knock

Post by ekjdm14 »

Could you post a link to the web address for your video please, my browser's knackered and not letting me "see" any pics or videos in posts... If it's knocking rhythmically and about half engine speed then it's likely cam/valve related. From the above it sounds like it may be a bit slower than that though so not sure without listening.

One thing, you mention a recent cambelt change. Were the cam pulley/s disturbed to do the job? (think the crank pulley was eliminated already wasn't it? worth double checking though maybe) The reason that springs to mind is after replacing the head on a Mitsubishi 2.8TD, the cam sprocket worked loose after a couple of hours running and it sounded like it'd run a bearing!
'95 Xantia LX 1.9D-auto, Black, 109k
'03 206 GTi180 94k in surgery
'03 206 1.1S XUD9TE/veg project :mrgreen:
'99 306 2.0SE cabriolet 95k summer toy
'97 306 XS 1.6i 99k sat 12 years, fixed in 7 days, 150mile maiden voyage :rofl2:
User avatar
jimmymarsbar
(Donor 2016)
Posts: 139
Joined: 04 Apr 2016, 11:45
Location: Thackley, West Yorks
My Cars: 307 2.0 hdi 110
Megane Estate 1.5dCi
x 13
Contact:

Re: Engine Knock

Post by jimmymarsbar »

white exec wrote:If the car had a DMF originally, and this has been replaced with a (conventional) SMF, I wonder if this has been done properly. The SMF should be solid, but the clutch friction plate should be centrally sprung. Wonder how well the job was done?

If an SMF kit (usually 4-part: solid flywheel, sprung friction plate, clutch cover and thrust race) was initially fitted, and then someone replaced just the friction plate, then I wonder too whether they took the trouble to obtain the identical friction plate that sat in that 4-part kit, or whether they might have used something else?

A judder on a new clutch (lining not yet polished, or careless oil/grease on the parts) is not unknown, but should disappear very quickly.

_________

Listening to your video soundtrack again, I get about 38 clonks in 20 secs, or 114 clonks/min.
At a tickover of 800rpm, that would be 400 power strokes/min for any one cylinder. Correct?
So there aren't enough clonks for one duff injector...
Has to be something else.
Could be wrong.
Bit of a long story, but when I bought the car, there was no indication that the clutch had ever been changed. So, as the car had over 80k on the clock, I went by the assumption that it was the OE clutch and decided to change it anyway. I bought the kit for a DMF (307 2.0 hdi 110, DW10ATED engine), only for the mechanic who fitted it to find a single body flywheel. He then sourced a direct replacement (using part numbers from the removed friction plate), or so he tells me, and fits it, and that's when the judder starts. Unfortunately I only have his word to go on with this.

I agree, any "bedding in" should have happened quickly. It has/had improved, but never completely went away. Again, I was told that the O/S driveshaft was on it's way out, and bought one ready to be fitted until the knock started.

Onto the clonking...... on average, my engine idles at 802 rpm, which is bob-on. Even with the clonk, it idles at 802. The cam and crankshaft are in sync, according to PP. And there are no fault codes.

It's difficult to tell on the video because of the sound quality, but there seems to be a pattern where by the clonks are in sync with the engine, and then you get a double-tap. And, when you've got your ear in the engine bay, it's clearly from the gearbox end, which is why I've initially thought timing issues causing damage somewhere. But, conversely, the initial symptoms occurred over a 10-15 second period, rather than an almighty clatter, which is what I would expect if the timing belt failed.

I've not had chance to investigate anything yet, but hopefully tomorrow I can find out more.
User avatar
jimmymarsbar
(Donor 2016)
Posts: 139
Joined: 04 Apr 2016, 11:45
Location: Thackley, West Yorks
My Cars: 307 2.0 hdi 110
Megane Estate 1.5dCi
x 13
Contact:

Re: Engine Knock

Post by jimmymarsbar »

ekjdm14 wrote:Could you post a link to the web address for your video please, my browser's knackered and not letting me "see" any pics or videos in posts... If it's knocking rhythmically and about half engine speed then it's likely cam/valve related. From the above it sounds like it may be a bit slower than that though so not sure without listening.

One thing, you mention a recent cambelt change. Were the cam pulley/s disturbed to do the job? (think the crank pulley was eliminated already wasn't it? worth double checking though maybe) The reason that springs to mind is after replacing the head on a Mitsubishi 2.8TD, the cam sprocket worked loose after a couple of hours running and it sounded like it'd run a bearing!
The link is

I didn't do the timing belt myself - I prefer to leave the precision jobs to the professionals! Depending on what I do with the car, I won't know unless I take it apart.

The clonking sounds like it's coming from the gearbox end, which leaves me thinking it's related to the crankshaft. But, I'm poking about in the dark really. It is a bit slower than half engine speed, it's got a rhythm to it, but it's slightly out of sync with the engine and has a double tap every now and then.
User avatar
white exec
Moderating Team
Posts: 7445
Joined: 21 Dec 2015, 12:46
Location: Sayalonga, Malaga, Spain
My Cars: 1996 XM 2.5TD Exclusive hatch RHD
1992 BX19D Millesime hatch LHD
previously 1989 BX19RD, 1998 ZX 1.9D auto, 2001 Xantia 1.8i auto
and lots of Rovers before that: 1935 Ten, 1947 Sixteen, 1960 P5 3-litre, 1966 P6 2000, 1972 P6 2000TC, and 1975 P6B 3500S
x 1752

Re: Engine Knock

Post by white exec »

That info does help.

When taking a look -
We know the clonking occurs at constant-speed idling.
If you gently open the throttle, does the clonk reduce while the engine picks up speed?
At idle, if you press the clutch, does the clonk disappear?
At idle, if you engage a gear, and then lift the clutch slightly (to begin a bit of traction) does the clonk disappear?
Is the gearbox oil level correct, by the way?

I'm thinking a loose or badly fitted flywheel, which could cause erratic flywheel position (sensor) readings, and also poor clutch action. The above tests would give some clues.
Chris
User avatar
jimmymarsbar
(Donor 2016)
Posts: 139
Joined: 04 Apr 2016, 11:45
Location: Thackley, West Yorks
My Cars: 307 2.0 hdi 110
Megane Estate 1.5dCi
x 13
Contact:

Re: Engine Knock

Post by jimmymarsbar »

I'll test it tonight, from what I remember on Sunday, the clonks are apparent throughout the rev range and clutch action. Whether there is any change when it is engaging I can't say right now, but hopefully I'll come back with more info later today
User avatar
white exec
Moderating Team
Posts: 7445
Joined: 21 Dec 2015, 12:46
Location: Sayalonga, Malaga, Spain
My Cars: 1996 XM 2.5TD Exclusive hatch RHD
1992 BX19D Millesime hatch LHD
previously 1989 BX19RD, 1998 ZX 1.9D auto, 2001 Xantia 1.8i auto
and lots of Rovers before that: 1935 Ten, 1947 Sixteen, 1960 P5 3-litre, 1966 P6 2000, 1972 P6 2000TC, and 1975 P6B 3500S
x 1752

Re: Engine Knock

Post by white exec »

Basically we need to know whether asking the clutch/flywheel to do a bit of work under light load "takes up the slack" in what might be a loose component.
Hope all goes usefully tonight.
Chris
User avatar
jimmymarsbar
(Donor 2016)
Posts: 139
Joined: 04 Apr 2016, 11:45
Location: Thackley, West Yorks
My Cars: 307 2.0 hdi 110
Megane Estate 1.5dCi
x 13
Contact:

Re: Engine Knock

Post by jimmymarsbar »

Ok, clutch/flywheel update.

At idle, the engine is the same. Ticks over at a steady 800ish rpm, with the knock doing it's thing.

Rev the engine, the knock frequency increases as expected.

Put the car into first, engage the clutch, and as it hits biting point, the clunk becomes muffled (ie. it quietens and is a duller, less sharp sound). Driving off, the clonk picks up volume again slightly.

If I start moving off under extra load (ie. uphill), the car seriously struggles to get moving - the first time I tried moving up this slight incline, I stalled. It was very tangible as to how much it struggled - the proverbial square peg in a round hole comes to mind.

So, at the end of that little experiment, I can quite confidently say that the issue may lay in the crank somewhere, almost certainly the clutch/flywheel.
User avatar
white exec
Moderating Team
Posts: 7445
Joined: 21 Dec 2015, 12:46
Location: Sayalonga, Malaga, Spain
My Cars: 1996 XM 2.5TD Exclusive hatch RHD
1992 BX19D Millesime hatch LHD
previously 1989 BX19RD, 1998 ZX 1.9D auto, 2001 Xantia 1.8i auto
and lots of Rovers before that: 1935 Ten, 1947 Sixteen, 1960 P5 3-litre, 1966 P6 2000, 1972 P6 2000TC, and 1975 P6B 3500S
x 1752

Re: Engine Knock

Post by white exec »

I would guess a loose flywheel, or maybe clutch cover - the latter a bit unlikely, as it's usually located by pegs as well as peripheral bolts.

Not sure whether you can get to inspect the flywheel edge anywhere, to try levering it, and ascertaining whether it's loose or not. Take starter motor out? Access via crank sensor hole? Maybe someone will know this engine better than I do.

Looks like it's going to have to be opened up to establish exact cause. Probably best not to drive the car as it is.
Chris
User avatar
jimmymarsbar
(Donor 2016)
Posts: 139
Joined: 04 Apr 2016, 11:45
Location: Thackley, West Yorks
My Cars: 307 2.0 hdi 110
Megane Estate 1.5dCi
x 13
Contact:

Re: Engine Knock

Post by jimmymarsbar »

Thanks for all your feedback and comments.

This engine, DW10ATED in a 307, is possibly the most inaccessible engine in the history or motoring. It is a case of gearbox out and checking from there.

The flywheel on it's arse is making everything make much more sense
bigjl2
Posts: 73
Joined: 12 Jul 2007, 15:53
Location: essex
My Cars:
x 2

Re: Engine Knock

Post by bigjl2 »

When people started putting SMF on Transit diesels instead of a DMF there where more than a few bottom end failures.

Not sure if that is relevant but just one of the many pointless facts stored in the old grey matter.
Xantia 2.0SX estate (now sold)
Xantia 1.9 TD SX hatch 102K (all old MOT'S) and working A/C.(Sold to forum member, time to save for a C5)
2007 Berlingo Multispace 1.4i Forte, bought in November with genuine 27k miles, now on 39799 miles.
User avatar
white exec
Moderating Team
Posts: 7445
Joined: 21 Dec 2015, 12:46
Location: Sayalonga, Malaga, Spain
My Cars: 1996 XM 2.5TD Exclusive hatch RHD
1992 BX19D Millesime hatch LHD
previously 1989 BX19RD, 1998 ZX 1.9D auto, 2001 Xantia 1.8i auto
and lots of Rovers before that: 1935 Ten, 1947 Sixteen, 1960 P5 3-litre, 1966 P6 2000, 1972 P6 2000TC, and 1975 P6B 3500S
x 1752

Re: Engine Knock

Post by white exec »

The thinking behind DMF was that it would reduce rotational vibration and stress on the transmission, and could permit the use of less-heavy driveline components, thus reducing weight, consumption and costs. Some manufacturers did not change the spec of their drivelines when they fitted DMF - Toyota was one of these on its D4D engines - and so the retrofitting of SMF did not cause problems. (Avensis taxis, worldwide, received plenty of SMF conversions.) It is possible other manufacturers took a different line.
Chris
User avatar
jimmymarsbar
(Donor 2016)
Posts: 139
Joined: 04 Apr 2016, 11:45
Location: Thackley, West Yorks
My Cars: 307 2.0 hdi 110
Megane Estate 1.5dCi
x 13
Contact:

Re: Engine Knock

Post by jimmymarsbar »

The DMF aren't there for the funnies - they serve a purpose, and dampening any vibration caused by more powerful engines is necessary. So it does make me wonder why people voluntarily convert the drive to a single body flywheel.

I've heard someone describe it as "an expensive way to compensate for poor driving".

Even though the DMF has it's own inherent faults, perhaps they are better left alone?
User avatar
white exec
Moderating Team
Posts: 7445
Joined: 21 Dec 2015, 12:46
Location: Sayalonga, Malaga, Spain
My Cars: 1996 XM 2.5TD Exclusive hatch RHD
1992 BX19D Millesime hatch LHD
previously 1989 BX19RD, 1998 ZX 1.9D auto, 2001 Xantia 1.8i auto
and lots of Rovers before that: 1935 Ten, 1947 Sixteen, 1960 P5 3-litre, 1966 P6 2000, 1972 P6 2000TC, and 1975 P6B 3500S
x 1752

Re: Engine Knock

Post by white exec »

DMF is fine when it works well, but they do have a somewhat chequered history, right across the piece. Later incarnations are better, but from the mid-90s to around 2008 or so, they caused a fair bit of mayhem.

The circumferential springs inside the flywheel (polymer infill on some versions) acquire slop, allowing the two parts of the flywheel to become only loosely connected, with vast amounts of rotational backlash. This can make gear changes, and particularly hill starts, difficult. In the case of hill starts, high engine revs have to be used to deter clutch snatch, no matter how skilfully you drive. Poor driving it is not, I can assure you.

Wear inside the flywheel shortens the life of the (unsprung) clutch friction plate, and can cause the modern epoxy-and-twine friction material to wear rapidly, and actually strip and disintegrate, leaving the car suddenly without drive.

Symptoms of DMF wear are poor clutch control, unexplained knocks and tinkles, and problematic hill-starts.

DMF flywheels are now regarded as consumable items, along with clutch plates, brake pads, and the like - but they are not cheap. It may cost €600 to replace one, in parts cost alone.

Life can be alarmingly short. Our RAV4 got through two (one at 60k km, and its 'improved' genuine Toyota replacement just 24k km later). It now runs on a Blueprint SMF kit (with 110k km under its belt), and has all the proper clutch action and control that you would expect from a new vehicle.

SMF kits are available from most of the major clutch manufacturers, including Valeo, Borg+Beck, etc. Some converted vehicles exhibit slightly increased vibration at idle; most do not. Modern DI diesels can be harsh things, especially when deprived of a meaty crank, damper and flywheel. Unfortunately, when DMF flywheels are described as "absorbing unwanted vibrations", this absorption does come at zero cost - all that energy has to go somewhere, and the device can simply give up at some point, later or sooner.

My own opinion is that if a DMF fails, it should be replaced with another (latest-type) one. If that fails, then it's time to call it a day, and fit an SMF, at a fraction of the parts cost.
Chris
Post Reply