Odd Knock issue
Moderator: phpBB2 - Administrators
Odd Knock issue
Hi guys,
I have an L33 (LS1) with MS3X and the ExtraEFI "KnockTek" knock conditioner. Recently I've noticed some knock detection events from the KnockTek. It only seems to start when the car is almost warm (i.e. around 85c) and only happens at idle. Rev the car and it seems to go away. I haven't caught these in TS or anything as, for some reason, I can't get the MS unit to "see" the KnockTek. I presume I have messed up the pin wiring and it's on the wrong pin. I need to get the tools out and trace where the hell I have wired it to... Anyway, I first noticed it on Sunday morning, but then had to shut the car off as I had to be somewhere, so I never really looked into it, but I noticed AFR at the time seemed ok (around 14.5-15 or so) and the timing was at 15.6 advance.
Then, when I looked at it again yesterday it did the same thing. No knock at "cold" but starting just before it got up to "normal" operating temperature. I had already retarded the spark to around 14 degrees before I started the car, just in case, but all that seemed to do was make the idle lumpy, so I tried enriching the mix a little, and I also reduced the spark dwell a little in case a coil was firing early, but nothing made any difference, just random knock events.
I then decided it was time to sort out why the MS couldn't actually see it, so I shut the engine off and went to dig some tools out. I first thought I'd check the KnockTek output as this is supposed to send a steady 5v and drop to 0v on knock... so I exposed a little of the wire and put a multi meter over it which showed a steady 5v. Fine. Then I fired the engine back up and waited for some knock (While I expected the event to be too quick to get an accurate voltage reading on the MM I hoped I'd at least see a flicker or so to indicate that something was happening on the wire)... but I never got any more knock events! It's worth mentioning here that I don't really know what a "knock" sounds like to the human ear, and I'm not exactly bat-like in my hearing, but I couldn't hear anything obviously different when the red light was flickering... just the usual click-clack of moving parts.
Now, when I originally got the block one of the sensors was rather rusty, and I just cleaned it up and re-used it, but looking yesterday it seems that a rusty sensor is only fit for the bin... A few people on LS1-Tech and so on have said there is an issue on the truck engines where the rear sensor gets waterlogged and goes bad...
I did check the plugs to see what they showed up and most were fine, with one showing a bit of soot build up and looked a bit wet on the threads. I had checked and cleaned it up before I started the engine yesterday though, and then saw the knock was still there, so I don't think there was excessive soot buildup glowing or anything that may cause early ignition.
It's worth noting that I have two O2 sensors on this, and both banks are showing similar fueling (not always exactly identical, but usually within 0.2 or so of each other), also, the car only passed its emissions tests while the O2 sensors were showing the car running at or near stoich, as you'd expect, and it's only done around 350 or so miles since the test, so I don't think they've both gone bad or are massively out of calibration, so I don't *think* it's a dodgy O2 sensor and I'm running unknowingly lean.
I also checked the timing a few months ago and that was OK then, so again, I don't *think* I'm running the spark unknowingly advanced.
Finally, I've been working on this engine, with the bonnet up, quite a lot over the build period and I can only remember seeing the knock conditioner flicker red once or twice before during cranking... although that may have been back when I had the dodgy cam sensor and was having issues with sync so may have been a real misfire/knock event, I can't remember when, only that I've seen a knock event happen during cranking.
So, what do we think? Considering the AFR should have been "safe" enough, and if anything the spark was probably a little retarded for idle, could this be:
A. Oil from a leaky stem seal or something in one cylinder igniting and causing actual knock
B. Something else causing actual knock, such as timing or fueling
C. a dodgy knock sensor showing knock when there isn't really anything
D. the KnockTek interpreting a noisy drive train or something as knock
E. Idle timing is too retarded at the original 15.6 and maybe causing its own issues?
F. f**k knows...
I have an L33 (LS1) with MS3X and the ExtraEFI "KnockTek" knock conditioner. Recently I've noticed some knock detection events from the KnockTek. It only seems to start when the car is almost warm (i.e. around 85c) and only happens at idle. Rev the car and it seems to go away. I haven't caught these in TS or anything as, for some reason, I can't get the MS unit to "see" the KnockTek. I presume I have messed up the pin wiring and it's on the wrong pin. I need to get the tools out and trace where the hell I have wired it to... Anyway, I first noticed it on Sunday morning, but then had to shut the car off as I had to be somewhere, so I never really looked into it, but I noticed AFR at the time seemed ok (around 14.5-15 or so) and the timing was at 15.6 advance.
Then, when I looked at it again yesterday it did the same thing. No knock at "cold" but starting just before it got up to "normal" operating temperature. I had already retarded the spark to around 14 degrees before I started the car, just in case, but all that seemed to do was make the idle lumpy, so I tried enriching the mix a little, and I also reduced the spark dwell a little in case a coil was firing early, but nothing made any difference, just random knock events.
I then decided it was time to sort out why the MS couldn't actually see it, so I shut the engine off and went to dig some tools out. I first thought I'd check the KnockTek output as this is supposed to send a steady 5v and drop to 0v on knock... so I exposed a little of the wire and put a multi meter over it which showed a steady 5v. Fine. Then I fired the engine back up and waited for some knock (While I expected the event to be too quick to get an accurate voltage reading on the MM I hoped I'd at least see a flicker or so to indicate that something was happening on the wire)... but I never got any more knock events! It's worth mentioning here that I don't really know what a "knock" sounds like to the human ear, and I'm not exactly bat-like in my hearing, but I couldn't hear anything obviously different when the red light was flickering... just the usual click-clack of moving parts.
Now, when I originally got the block one of the sensors was rather rusty, and I just cleaned it up and re-used it, but looking yesterday it seems that a rusty sensor is only fit for the bin... A few people on LS1-Tech and so on have said there is an issue on the truck engines where the rear sensor gets waterlogged and goes bad...
I did check the plugs to see what they showed up and most were fine, with one showing a bit of soot build up and looked a bit wet on the threads. I had checked and cleaned it up before I started the engine yesterday though, and then saw the knock was still there, so I don't think there was excessive soot buildup glowing or anything that may cause early ignition.
It's worth noting that I have two O2 sensors on this, and both banks are showing similar fueling (not always exactly identical, but usually within 0.2 or so of each other), also, the car only passed its emissions tests while the O2 sensors were showing the car running at or near stoich, as you'd expect, and it's only done around 350 or so miles since the test, so I don't think they've both gone bad or are massively out of calibration, so I don't *think* it's a dodgy O2 sensor and I'm running unknowingly lean.
I also checked the timing a few months ago and that was OK then, so again, I don't *think* I'm running the spark unknowingly advanced.
Finally, I've been working on this engine, with the bonnet up, quite a lot over the build period and I can only remember seeing the knock conditioner flicker red once or twice before during cranking... although that may have been back when I had the dodgy cam sensor and was having issues with sync so may have been a real misfire/knock event, I can't remember when, only that I've seen a knock event happen during cranking.
So, what do we think? Considering the AFR should have been "safe" enough, and if anything the spark was probably a little retarded for idle, could this be:
A. Oil from a leaky stem seal or something in one cylinder igniting and causing actual knock
B. Something else causing actual knock, such as timing or fueling
C. a dodgy knock sensor showing knock when there isn't really anything
D. the KnockTek interpreting a noisy drive train or something as knock
E. Idle timing is too retarded at the original 15.6 and maybe causing its own issues?
F. f**k knows...
Good point, I haven't really rooted around to check if something else is knocking around against the block... worth a look, and I didn't know the KnockTek was adjustable... I assumed the knock detectors themselves only gave a signal at certain frequencies... I'll have to have a look into that, although "loose items" or whatever is more what I actually meant in option D.
Option B. I was thinking more of something else actually causing "pinging" or "detonation" or "pre-ignition" or whatever else people know as "knock" that I haven't thought about...
I've ordered some new knock sensors anyway. As I'd read about them going dodgy if they rust I thought I may as well change them just in case.
Option B. I was thinking more of something else actually causing "pinging" or "detonation" or "pre-ignition" or whatever else people know as "knock" that I haven't thought about...
I've ordered some new knock sensors anyway. As I'd read about them going dodgy if they rust I thought I may as well change them just in case.
No V8 at present
http://www.ms2tuning.com/uploads/3/0/5/ ... truct2.pdf
There's an adjuster on the side
There's an adjuster on the side
So thats where it went !
-
- Forum Contributor
- Posts: 3979
- Joined: Sat Nov 18, 2006 6:22 pm
- Location: Northern Ireland
-
- Forum Contributor
- Posts: 3979
- Joined: Sat Nov 18, 2006 6:22 pm
- Location: Northern Ireland
If you are referring to knock, as uncontrolled combustion, detonation, pinking etc etc whatever you want to call it.volospian wrote:I'm saying i don't know if it's the device or some sort of knock.
It only seems to be happening at idle, and only when the engine is just up to temp, but it then appeared to go away once the engine had been hot for a while.
There is NO way whatsoever that is occurring at idle. So that totally resolves that aspect.
If you are suggesting knock is some other mechanical engine noise it is picking up, then certainly that is possible.
Hence the need for correct calibration of such devices for them to be of any use. Best way of that is to confirm via ears/stethoscope type device what noises may or may not being heard.
Lol, for the avoidance of doubt I'll call pinging, det, pre-ig, etc. "knock" and engine rattles, bangs and vibrations as "clatter"...
I thought it couldn't be "knock", but I never like to simply assume in case someone with more knowledge and experience comes along and says "oh, yes, that can be caused by ...."
So, if we're ruling out knock, it's probably just something clattering. I'll look around and see if anything is tapping the block/sump. If not it's probably something internal.
In the meantime I'll look into calibrating the KnockTec and assume it's just something rattling a bit while it warms up
Cheers,
Rich.
I thought it couldn't be "knock", but I never like to simply assume in case someone with more knowledge and experience comes along and says "oh, yes, that can be caused by ...."
So, if we're ruling out knock, it's probably just something clattering. I'll look around and see if anything is tapping the block/sump. If not it's probably something internal.
In the meantime I'll look into calibrating the KnockTec and assume it's just something rattling a bit while it warms up
Cheers,
Rich.
No V8 at present
Pinking is usually at its most obvious when you put a high load on the engine at low to moderate revs.
Caused by too low an octane petrol, too much ignition advance and or a too weak mixture.
If it is pinking, changing from 95 to 98 octane petrol - or the other way round - should make it better or worse.
The actual knock sensor is a form of narrow band contact microphone. The electronics also usually further filter the bandwidth of the audio to isolate the wanted frequencies. As with all audio, it can be difficult filtering out the wanted from unwanted. Which is why it can be tricky fitting a knock sensor to an engine which didn't have one.
Caused by too low an octane petrol, too much ignition advance and or a too weak mixture.
If it is pinking, changing from 95 to 98 octane petrol - or the other way round - should make it better or worse.
The actual knock sensor is a form of narrow band contact microphone. The electronics also usually further filter the bandwidth of the audio to isolate the wanted frequencies. As with all audio, it can be difficult filtering out the wanted from unwanted. Which is why it can be tricky fitting a knock sensor to an engine which didn't have one.
Dave
London SW
Rover SD1 VDP EFI
MegaSquirt2 V3
EDIS8
Tech Edge 2Y
London SW
Rover SD1 VDP EFI
MegaSquirt2 V3
EDIS8
Tech Edge 2Y