Mercedes CLA Forum banner

41 - 60 of 64 Posts

·
Registered
Joined
·
277 Posts
simplification

"Remove the 3 torx screws from the airbox neck (green circle) and loosen the two hose clamps (red circle)" :

You do not need to do disassemble the neck joint, as you will have to remove the air filter box top anyway.
Just release the hose clamp and remove the 4 torx screws holding the filter box top in place,
and remove the filter box top, and the assembled neck joint as an unit.

While you are at it, check the air filter element, chances are it needs to be replaced, and it can be easily done
at this point. So it is good to have a new filter on hand.
 

·
Registered
Joined
·
39 Posts
I've been struggling with my ebay knock off scanner to check on timing as it only shows realtime data. Is there any OBD2 adapter and software to make a real datalog?

regards.
I have been using the obdlink mx bluetooth unit and its obdlink software for years now ... lots of real time data (time density can be 100ms but not all pids respond at that rate)
I got my obdlink and downloaded the apps. How exactly can I log knock?

Regards.
 

·
Registered
Joined
·
12,121 Posts
I got my obdlink and downloaded the apps. How exactly can I log knock?
Regards.
Wow this is nontrivial. The baseline is the knock sensor (essentially just a piezo microphone)
which listens to the cylinders/pistons for the "ping" signature and sends a blip to the ECU to
pull power. Problem is by the time the sensor "hears" the ping it has thus already occurred as
that acoustic signature stems from an improper ignition profile which slams into the piston and
all those sounds show up. It's like pushing a kid on a swing ... the proper way is to add power
to the swing by applying force with a correct time profile, instead of just slamming a fist against
it at the top of the swing's travel, or worse hitting it on its way up.

So a sidebar is if any CELs show up with codes reflecting a dead knock sensor, fix it now!

Anyway you can't "track" knock as it is not a continuous parameter (and if it were your engine
is surely dead or on its way out) but under the best conditions the knock sensor is working so
the ECU pulls power when pings occur, so tracking data over time (power torque boost timing
etc) will show how often this "pull" is occurring ... tied to what you "hear" while driving so you
know when to look at the logs. Then one would track this "behavior" over time ... and if this
shows up more and more often, then something is wrong. All those GB of "data" is just that,
data, but looking for "information" is the critical exercise. Just a bit of Excel elbow grease.

In my case I log every trip with data points at 200ms intervals. On certain trips I will have an
unexpected power requirement due to conditions, or I make a deliberate WOT run. I find and
plot those segments and archive them, and make any notes of "odd" ECU responses. So far
when power data >220hp and torque data >310ftlbs, no data plot trends show the "snapping"
back of parameters like ignition timing or boost and I have not "heard" any pinging. Though
of course I am presuming pinging would occur under heavy loads, as the engine is healthy,
so I home in on those data segments. I hope I never hear any pinging regardless of engine
status. It will be a sad day when she pings just backing out of the garage!

Hope this makes sense?
 

·
Registered
Joined
·
39 Posts
Sure makes complete sense. I do understand everything you say. I'm very used to work with data logging softwares for stand alone and open source ECU flashing (ecuflash, evoscan, romraider, etc...) And that's exactly why I was asking if there's a "knock" datalog as there is in the softwares I mentioned. Usually there's a "knock retard", "feedback knock" or "knock sum" data input to check for possible knock or even engine noise.
I guess there's no knock input on obdlink software and I'll just have to track it through timing curves.

Best regards.
I really appreciate your time explaining me this
 

·
Registered
Joined
·
12,121 Posts
No worries ... hope I got it right though
Since you are savvy with the software side perhaps you can look into an extended real time data parser which can “count” a designated (preknock) event occurrence and plot/log them over time ... then notify when counts vs elapsed time hits a threshold ... and gee that will be better than any knock sensor
 

·
Registered
Joined
·
277 Posts
Torque

This is a 14mm thread plug.
The NGK site below gives the torque as 25-30 NM (18-22 foot pounds).
PLUG STUDIO / NGK

Where does the 17 foot pounds come from? (that is what I used as per Picard and it seemed to work fine, but curious)

What is the recommendation on the torque when reinstalling a plug?

NGK gives an alternative method of : 1/2 turn after fingertight when new, and 1/12 (30 degrees) turn on reinstall.
 

·
Registered
Joined
·
12,121 Posts
Yeah the fault was mine
M270 engine docs note 23nm/17ftlb
Not too keen with generic torque specs or methods as those do not account for specific head alloy specs nor thread cut specs not to mention engine specific thermal ranges at the plug threads so over stretching them is asking for trouble ... aside from the typical monkey wrench bit about stripping threads at room temp
As for reinstalling if there are crush washers (not just a conical seat) I would toss and use new plug not worth the headaches related to an improper seal at such a critical point
 

·
Registered
Joined
·
1,176 Posts
Discussion Starter · #49 ·
This is a 14mm thread plug.
The NGK site below gives the torque as 25-30 NM (18-22 foot pounds).
PLUG STUDIO / NGK

Where does the 17 foot pounds come from? (that is what I used as per Picard and it seemed to work fine, but curious)

What is the recommendation on the torque when reinstalling a plug?

NGK gives an alternative method of : 1/2 turn after fingertight when new, and 1/12 (30 degrees) turn on reinstall.
Here you go. From the CLA250 service manual. m270.JPG
 

·
Registered
Joined
·
277 Posts
Dielectric grease

Thanks for posting the plug torque specs.

Question about the Benz dielectric compound: I see it on Ebay, but no info whatsoever as to what it is or
what its properties are. Is it something special, or can we use some generic
contact protector/enhancer?

AFAIK dielectric grease is non-conductive so if used, only a very thin layer should be used.
Its role would be to seal out moisture, help future disassembly and protect terminals from
corrosion. Is there a reason to believe the MB grease is better/different than Permatex, or Versachem?

Deoxit has a "contact enhancer" L260cp grease that is conductive, and removes oxidation.
The cp indicates it contains copper particles to enhance conductivity.

Goood up to 260 celsius, lithium based. Looks to me it does everything a dielectric grease would do
and a lot more. I have some experience using Deoxit products with Hifi or guitar amps, and their products are
awesome for those applications.

What do you all think?
 

·
Registered
Joined
·
12,121 Posts
Always have though high energy e connections should be dry but then these conductive compounds show up and not only make connect and disconnect easier they are supposed to enhance those critical contacts but of course if the compound flows at temp then ... oops
The nonconductive stuff is just to relief the boot friction for easier mount and dismount and seal out moisture ... which makes sense
My compromise would be a tiny tiny tiny smear of conductive on the spark plug contact and a small bit of nonconductive around the boot lip ... get dual benefits
 

·
Registered
Joined
·
277 Posts
The Deoxit grease is rated up to 250 Celsius (I think that is 550 Fahrenheit) .. that is comparable to the Permatex, and Versachem Dielectrics.
Too late for the Benz ( I am not taking it apart again! ) , but I have a couple of Hyundai Sonatas due (my kids cars) , will try the Deoxit on those.

I thought you were not supposed to put grease on the rubber boot. it would certainly smear all over the plug
I am reading that we are not supposed to get greasy fingerprints on the porcelain part of the plug.

I would be curious to learn more about the OEM dielectric grease, if someone has some, or has some info
on this please post.

I will let you guys know how well my DeoxIt experiment works out.
 

·
Registered
Joined
·
12,121 Posts
Re: grease ... yeah two schools of thought
I know dry/dry works and thats a baseline and may just stick with that (... if it aint broke ...)
 

·
Registered
Joined
·
12,121 Posts
was watching a timing chain video ... and came across a couple frames which shed some more light on the intake removal thing (well at least according to the germans)
what's left dangling hints at how they prefer to execute the move

IntakeRemoval1.jpg

IntakeRemoval2.jpg
 

·
Registered
Joined
·
39 Posts
No worries ... hope I got it right though
Since you are savvy with the software side perhaps you can look into an extended real time data parser which can “count” a designated (preknock) event occurrence and plot/log them over time ... then notify when counts vs elapsed time hits a threshold ... and gee that will be better than any knock sensor
Well after trying to datalog my car lots of times, it started runing on 3 cylinders. I thought it was another busted spark plugs but no. I had a "small" issue with a piston. Luckily cylinder wall had zero damage.

Finally I fixed it with a new piston and traded it for a C450 AMG. Cant be happier after the hell I lived with the cla.

Guys, don't mess with your ECU. Not worth the risk.

Regards.
 

Attachments

·
Registered
Joined
·
12,121 Posts
Ouch!!!!!
Some surface microreliefs must have literally tore up that piston ... or a speck of dirt within the oil spray and caught between the piston and cylinder wall at just the right moment
Doubt anything to do with the tune etc else a lot of us will have busted pistons ... more like contamination in the air and/or oil channels

Hows the new C ?
 

·
Registered
Joined
·
12,121 Posts
Also the damage zone is orthogonal to the thrust plane ... if you have pix of that part of the cylinder wall it will be very telling
 

·
Registered
Joined
·
39 Posts
You had a ECU Tune or PiggyBack ?
ECU tune by Brperformance

Also the damage zone is orthogonal to the thrust plane ... if you have pix of that part of the cylinder wall it will be very telling
These are the best pictures i could find.

20190319_165609.jpg 20190319_165615.jpg

I really dont know if it was an oil issue or a tune one. I've seen lots of pistons with busted ringlands due to bad tunes (knock) so honestly no big surprise for me although piston crown looks intact and cylinder wall also.

The new C is awesome. V6 twin turbo, 367 HP, 4 matic. Love it!

Regards.
 

·
Registered
Joined
·
12,121 Posts
Great pix to see!

And I stand corrected ... you have a good point cuz a tough ping (bad tune) at the wrong spot in the flame front could do a number on the piston and not on its crown either

So the C is not quite the grandpa mobile ay?
 
41 - 60 of 64 Posts
Top