Is this detonation damage?
Discussion
We tend to find the bhp suffers the moment we get detonation on NA and forced induction. With regard to senses, I cannot hear detonation anymore as I am getting too old. I am not sure many folk would ever hear the detonation at high rpm/torque WOT when the damage really gets done! When we get detonation we get a different smell to the exhaust emissions as well, this is probably due to the severe effects of the detonation. We are also working with the Plex Tuning knock monitor, slow progress as we never seem to have the spare time needed to set it up. Same with setting up our Superflow dyno, not enough hands on deck 
Never mind, one day we will sort it out, especially as we have a few folk queueing up to do some setting up/research.
It would be nice if there was a good one size fits all knock monitor which didn't need such specific calibrating for each type of engine tested.
Peter
Never mind, one day we will sort it out, especially as we have a few folk queueing up to do some setting up/research.It would be nice if there was a good one size fits all knock monitor which didn't need such specific calibrating for each type of engine tested.
Peter
PeterBurgess said:
We tend to find the bhp suffers the moment we get detonation on NA and forced induction.
Yes, it will. All that lost mechanical power is turned into heat and applied directly to the piston and cylinder head instead - ouch!That Plex thingy looks a bit fancy and not at all so difficult to set up. I'd have a crack at using it if it was lying on the shelf.
https://www.youtube.com/watch?v=v0ttwzR0kjs
I have used the turboXS Knocklite as a indicator. It works really well and detects even inaudible knock.
I confirmed this by pulling 1 deg of timing from those spots and the reproducible knock indication disappeared.
Since I wanted more, I changed to a J&S safeguard knock control system on my K incl. the gauge.
It reacts within one revolution, pulls up to 10 deg of timing on individual cylinders(or 20 in an other setting for slow combusing 2V engines with huge initial advance ), depending on knock intensity, ramps the retard down again in steps within a certain number of revolutions. The display allows you to also spot its operation and readjust the map at those spots.
On a Rover K in a Elise knock is very easy to hear, even under full boost. Sounds like electric sparks from the engine bay.
The J&S is a bit harsh as it sometimes pulls more timing than necessary, but this is better than the other way around. It also pulls timing getting knock from runing lean, e.g. Fuel starvation in long corners at low fuel levels, a known Elise issue.
The J&S is no OEM grade supersmooth fully integrated and tuned device, but it still works. And that is the most important. It saves your ass once knock gets out of hand and takes care about lighter knock until you fixed the map or whatever else was causing it.
Otherwise, I start rather retarded and they work my way up. The engine does not like too little advance, but as you know it is not enough, it is just a question of adding more advance until things clear up.
Starting in the middle is rather dangerous.
I confirmed this by pulling 1 deg of timing from those spots and the reproducible knock indication disappeared.
Since I wanted more, I changed to a J&S safeguard knock control system on my K incl. the gauge.
It reacts within one revolution, pulls up to 10 deg of timing on individual cylinders(or 20 in an other setting for slow combusing 2V engines with huge initial advance ), depending on knock intensity, ramps the retard down again in steps within a certain number of revolutions. The display allows you to also spot its operation and readjust the map at those spots.
On a Rover K in a Elise knock is very easy to hear, even under full boost. Sounds like electric sparks from the engine bay.
The J&S is a bit harsh as it sometimes pulls more timing than necessary, but this is better than the other way around. It also pulls timing getting knock from runing lean, e.g. Fuel starvation in long corners at low fuel levels, a known Elise issue.
The J&S is no OEM grade supersmooth fully integrated and tuned device, but it still works. And that is the most important. It saves your ass once knock gets out of hand and takes care about lighter knock until you fixed the map or whatever else was causing it.
Otherwise, I start rather retarded and they work my way up. The engine does not like too little advance, but as you know it is not enough, it is just a question of adding more advance until things clear up.
Starting in the middle is rather dangerous.
As others have mentioned the Plex is one device, and probably one of the better electronic units ( which can of course also go back to your ears, but will give visual/recorded info too on the device or laptop which can be used to trigger alarms )
If you're feeling flush, there is of course more fancy stuff like their combustion monitor system which uses in cylinder pressure transducers. Clearly this would be one of the better ways of detecting things like detonation and optimising performance.
A bit out of budget for most DIY'ers, but probably not unreasonable for a busy dyno ? I'd love one myself, but ass DIY, just out of my budget. Some day though
Although I think such pressure transducer setups are coming down in price thee days ? It's about time they were mainstream for all ecus.
Then there are some ecu's that detect detonation via crank acceleration or deceleration as the case may be, or via feedback from the coils. Again I'm surprised these arent more mainstream either in the aftermarket/tuning scene.
They'd be great additions and tools for the users.
But there are many methods over and above the basic analogue det cans ( which to be fair are bloody good if the ears/brain allow them to be )
And fair point about total timing on a K, it isnt an engine I've had the misfortune to work on, 10deg difference from std even with the lower CR would sound like a lot.
If you're feeling flush, there is of course more fancy stuff like their combustion monitor system which uses in cylinder pressure transducers. Clearly this would be one of the better ways of detecting things like detonation and optimising performance.
A bit out of budget for most DIY'ers, but probably not unreasonable for a busy dyno ? I'd love one myself, but ass DIY, just out of my budget. Some day though
Although I think such pressure transducer setups are coming down in price thee days ? It's about time they were mainstream for all ecus.
Then there are some ecu's that detect detonation via crank acceleration or deceleration as the case may be, or via feedback from the coils. Again I'm surprised these arent more mainstream either in the aftermarket/tuning scene.
They'd be great additions and tools for the users.
But there are many methods over and above the basic analogue det cans ( which to be fair are bloody good if the ears/brain allow them to be )
And fair point about total timing on a K, it isnt an engine I've had the misfortune to work on, 10deg difference from std even with the lower CR would sound like a lot.
Does anyone make an easy use knock sensing system for non ecu controlled cars? We tend to find our Olde Worlde engines are not particularly well catered for these days. Easier than the Plex Tune knock monitor that is
We have even tried to record the crankcase pressure rise associated with detonation but not getting very far!
My good mate Martin Faulks has been playing with a TFX pressure monitoring system he bought. He is finding it very difficult to get clean signals. If an electronics designer/engineer has problems what chance do I have?
http://www.tfxengine.com/
To be honest we find a lot of 'in car' diagnostic/recording stuff gets very upset with all the electronic noise we get in the dyno area and in the car!
Peter
We have even tried to record the crankcase pressure rise associated with detonation but not getting very far!
My good mate Martin Faulks has been playing with a TFX pressure monitoring system he bought. He is finding it very difficult to get clean signals. If an electronics designer/engineer has problems what chance do I have?
http://www.tfxengine.com/
To be honest we find a lot of 'in car' diagnostic/recording stuff gets very upset with all the electronic noise we get in the dyno area and in the car!
Peter
I've sometimes wondered if Saabs APC system could be rigged up to stand alone on say a V8.
https://ranwhenparked.net/2011/01/25/what-lies-ben...
Peter, it's tweakable to :-)
http://web.inter.nl.net/users/turbo-team-europe/ap...
https://ranwhenparked.net/2011/01/25/what-lies-ben...
Peter, it's tweakable to :-)
http://web.inter.nl.net/users/turbo-team-europe/ap...
Edited by Boosted LS1 on Monday 20th June 09:36
Problem is, engines are just noisy, and given there are so many noises anything has to listen for...I doubt it will ever be easy when the detector is listening for noise itself.
How much is that TFX gear ? What sort of transducers or spark plug arrangement do they use ?
Plex make their version
http://plex-tuning.com/products/plex-pcm-1000-comb...
Even one cylinder would be enough to start with, as you could simply tune that cylinder alone ( with suitable ecu of course ) and optimise it. You could repeat for all others, but if you got one working well, you could then apply that tuning/timing to all the others.
Or ION Sensing coils ? they dont seem new....but why arent they everywhere ?
http://vvnet.fi/ville/ion/DIY-Ion-Sensing-2.pdf
Hell, even basic misfire detection that most OEM cars run now, whether coil based or crank accel....why isnt this mainstream on aftermarket ecu's ? It would be a great tuning and diagnostic tool to be able to see and record misfire events.
How much is that TFX gear ? What sort of transducers or spark plug arrangement do they use ?
Plex make their version
http://plex-tuning.com/products/plex-pcm-1000-comb...
Even one cylinder would be enough to start with, as you could simply tune that cylinder alone ( with suitable ecu of course ) and optimise it. You could repeat for all others, but if you got one working well, you could then apply that tuning/timing to all the others.
Or ION Sensing coils ? they dont seem new....but why arent they everywhere ?
http://vvnet.fi/ville/ion/DIY-Ion-Sensing-2.pdf
Hell, even basic misfire detection that most OEM cars run now, whether coil based or crank accel....why isnt this mainstream on aftermarket ecu's ? It would be a great tuning and diagnostic tool to be able to see and record misfire events.
There's a dark art to cylinder pressure measurement. Even the guys with money and Kistler stuff have to work hard to make accurate data come alive. In a garage-type situation it's possible, but there will be some considerable hurdles. Easy to spend lots of time getting things to "work"
Also, so-called "spark plug transducers" generally suffer from a gas pipe resonance between the sensor and the opening to the combustion chamber. This is usually a long thin tube with a resonance close enough to knock frequency to really spoil the signal. Anything that uses piezo sensors in the "real world" will have them embedded into the cylinder head.
What I am playing about with is spark ion sensing. This can give knock signaling as well as Pmax data, allegedly. Currently fiddling with some SAAB trionic coils to see where it goes - that's if they don't zap me in the mean time...
Also, so-called "spark plug transducers" generally suffer from a gas pipe resonance between the sensor and the opening to the combustion chamber. This is usually a long thin tube with a resonance close enough to knock frequency to really spoil the signal. Anything that uses piezo sensors in the "real world" will have them embedded into the cylinder head.
What I am playing about with is spark ion sensing. This can give knock signaling as well as Pmax data, allegedly. Currently fiddling with some SAAB trionic coils to see where it goes - that's if they don't zap me in the mean time...
stevieturbo said:
Hell, even basic misfire detection that most OEM cars run now, whether coil based or crank accel....why isnt this mainstream on aftermarket ecu's ? It would be a great tuning and diagnostic tool to be able to see and record misfire events.
...there are a bunch of patents to spoil the party.Or just get a J&S safeguard knock control system. Runs of a BOSCH knock sensor and does the job. I did my map DIy with that, a wideband and my butt dyno. I can map it to within 1-2 deg of the knock limit. I was never near MBT, that made it easier. I always gained power adding advance until knock control tells me it is too much, regulates back and that is it.
Ive said:
Or just get a J&S safeguard knock control system. Runs of a BOSCH knock sensor and does the job. I did my map DIy with that, a wideband and my butt dyno. I can map it to within 1-2 deg of the knock limit. I was never near MBT, that made it easier. I always gained power adding advance until knock control tells me it is too much, regulates back and that is it.
but even it requires calibration, so in same boat as any detection device.If your ecu doesnt have good knock detection/control, then the J&S would be an option.
stevieturbo said:
but even it requires calibration, so in same boat as any detection device.
If your ecu doesnt have good knock detection/control, then the J&S would be an option.
It just has a sensitivity button. Turn it up until mechanical noise triggers it, then dial it back a tad. Easy as pie.If your ecu doesnt have good knock detection/control, then the J&S would be an option.
If it is too sensitive, you get too much retard upon knock and some retard from mechanical noises. If it is too sensitive, nothing bad happens, then engine just temporarily loses power.
Tell J&S the bore diameter when ordering. This is the main factor regarding the knock frequency.
I would not call this a calibration.
It is really simple.
A ECU with integrated knock control would be nice, indeed. I canot see patents being much of an issue until things get super sophisticated. Knock control is out there for more than 20 yrs. Many patents will be expired by now.
Ive said:
It just has a sensitivity button. Turn it up until mechanical noise triggers it, then dial it back a tad. Easy as pie.
If it is too sensitive, you get too much retard upon knock and some retard from mechanical noises. If it is too sensitive, nothing bad happens, then engine just temporarily loses power.
Tell J&S the bore diameter when ordering. This is the main factor regarding the knock frequency.
I would not call this a calibration.
It is really simple.
A ECU with integrated knock control would be nice, indeed. I canot see patents being much of an issue until things get super sophisticated. Knock control is out there for more than 20 yrs. Many patents will be expired by now.
If it is too sensitive and pulling timing when it shouldnt...then it needs calibrating. That's what calibrating is ! setting things up so they work correctly.If it is too sensitive, you get too much retard upon knock and some retard from mechanical noises. If it is too sensitive, nothing bad happens, then engine just temporarily loses power.
Tell J&S the bore diameter when ordering. This is the main factor regarding the knock frequency.
I would not call this a calibration.
It is really simple.
A ECU with integrated knock control would be nice, indeed. I canot see patents being much of an issue until things get super sophisticated. Knock control is out there for more than 20 yrs. Many patents will be expired by now.
And there are plenty of ecu's out there already with knock control....and as has been said, some good, some crap !
I use Motec Standalone Knock Module (SKM), linked in to a M800 ECU. Calibration did take some time, with the following parameters needing calibrated.
Knock frequency- This involved making a sound recording of the engine whilst inducing light knock and identifying the frequency where the signal to noise was best. I used frequency analysis software by Goldwave, linked in to the SKM Audio out.
Knock window- This is the crank angle where the SKM will look for knock, ignoring all other noise at the Knock Frequency. This prevents false positives and is typically in the 10-30deg ATDC region. Once I had my knock frequency, again I induced light knock trying various windows from 0-5deg ATDC to 25-30deg ATDC. I found that the signal was reduced at less than 5deg ATDC and above 15deg ATDC so assumed it was being clipped. I ended up staying with the default 10-20deg ATDC setting.
Ind cylinder gain- As the sensor is not equidistant from all cylinders, the gain needs to be adjusted to get the same signal levels across all cylinders. This required taking a log and add/subtract, re-log until they were +-2-3%.
Knock Level- This is the threshold above which knock retard is activated. This was time consuming. It required an idle to redline pull at every load site, and logging the background noise levels. Once established, I set the knock level 5% above these levels, so any signals above normal background noise would trigger a retard.
Short Term Trim- This is the initial retard applied to the knocking cylinder. I forget the units now, something like 3deg per event.
Recovery rate- the speed of which ignition advance is added to get back. Something like 0.1deg/rev, can’t remember the details.
Long Term Trim- For every event there is a reduction in base advance.
Quite a lot of calibration for what I expect is a simple system compared to OEMs.
Knock frequency- This involved making a sound recording of the engine whilst inducing light knock and identifying the frequency where the signal to noise was best. I used frequency analysis software by Goldwave, linked in to the SKM Audio out.
Knock window- This is the crank angle where the SKM will look for knock, ignoring all other noise at the Knock Frequency. This prevents false positives and is typically in the 10-30deg ATDC region. Once I had my knock frequency, again I induced light knock trying various windows from 0-5deg ATDC to 25-30deg ATDC. I found that the signal was reduced at less than 5deg ATDC and above 15deg ATDC so assumed it was being clipped. I ended up staying with the default 10-20deg ATDC setting.
Ind cylinder gain- As the sensor is not equidistant from all cylinders, the gain needs to be adjusted to get the same signal levels across all cylinders. This required taking a log and add/subtract, re-log until they were +-2-3%.
Knock Level- This is the threshold above which knock retard is activated. This was time consuming. It required an idle to redline pull at every load site, and logging the background noise levels. Once established, I set the knock level 5% above these levels, so any signals above normal background noise would trigger a retard.
Short Term Trim- This is the initial retard applied to the knocking cylinder. I forget the units now, something like 3deg per event.
Recovery rate- the speed of which ignition advance is added to get back. Something like 0.1deg/rev, can’t remember the details.
Long Term Trim- For every event there is a reduction in base advance.
Quite a lot of calibration for what I expect is a simple system compared to OEMs.
The J&S's recover rate is fixed rate per numbers of revolutions.
Knock frequency is a function of bore diameter and speed of sound.
You can specifiy it upon ordering.
Windowing of a J&S is done internally as I it aware of the ignition timing being connected to the coil drive. This is how is does individual cylinder knock control.
If too sensitive, turn it down a notch.
It is simple and not that sophisitcated, but it is way more sophisticated than any of the "blow'm up" tuner's ears or what most DIY can do by any other as simple means.
It is a band aid. I'd not dial in 40 deg advance and drive around letting it preventing blowing the engine. IT is an indicator with immediate means of action.
It works really well for what it is, is readily available, simple to install and set-up.
Is it OEM smooth, uninstructed and 110% idiot bullet proof running 20PSI boost on 80 octane sh...e fuel? No.
Is is of tremendous help mapping an engine and keeping it alive under use or during adverse conditions? Yes.
You can make things more complicated that they have necessarily to be to do something useful. Especially if you convince people of not doing any knock control as they believe you need to do is as sophistaced as above or by a OEM.
If you choice is J&S safeguard or nothing, I'd do and did J&S.
If my choice is J&S safeguard or the above Motec or Syvecs or Pectel knock system, I'd use the latter. (For 10x the price of a J&S unit).
The J&S has an analog output that you can connect to a MS3 ECU to implement knock trim.
I have asked Emerald, if they would implement that, but it was declined.
Knock frequency is a function of bore diameter and speed of sound.
You can specifiy it upon ordering.
Windowing of a J&S is done internally as I it aware of the ignition timing being connected to the coil drive. This is how is does individual cylinder knock control.
If too sensitive, turn it down a notch.
It is simple and not that sophisitcated, but it is way more sophisticated than any of the "blow'm up" tuner's ears or what most DIY can do by any other as simple means.
It is a band aid. I'd not dial in 40 deg advance and drive around letting it preventing blowing the engine. IT is an indicator with immediate means of action.
It works really well for what it is, is readily available, simple to install and set-up.
Is it OEM smooth, uninstructed and 110% idiot bullet proof running 20PSI boost on 80 octane sh...e fuel? No.
Is is of tremendous help mapping an engine and keeping it alive under use or during adverse conditions? Yes.
You can make things more complicated that they have necessarily to be to do something useful. Especially if you convince people of not doing any knock control as they believe you need to do is as sophistaced as above or by a OEM.
If you choice is J&S safeguard or nothing, I'd do and did J&S.
If my choice is J&S safeguard or the above Motec or Syvecs or Pectel knock system, I'd use the latter. (For 10x the price of a J&S unit).
The J&S has an analog output that you can connect to a MS3 ECU to implement knock trim.
I have asked Emerald, if they would implement that, but it was declined.
Boosted LS1 said:
I've sometimes wondered if Saabs APC system could be rigged up to stand alone on say a V8.
https://ranwhenparked.net/2011/01/25/what-lies-ben...
Peter, it's tweakable to :-)
http://web.inter.nl.net/users/turbo-team-europe/ap...
The Saab ECU's have come a long way since the APC boxes on the classic 900.https://ranwhenparked.net/2011/01/25/what-lies-ben...
Peter, it's tweakable to :-)
http://web.inter.nl.net/users/turbo-team-europe/ap...
Edited by Boosted LS1 on Monday 20th June 09:36
Trionic 5 as fitted to all post 1994 9000's and all turbo new generation 900's combined ion-sensing knock detection in the ignition cassette (four coils all in one handy cartridge - great till they break!) with variable boost APC valves, variable ignition timing and extra fuelling if knock is detected. The ECU map runs the usual sensors plus a manifold pressure (MAP) boost-request based program.
The later Trionic 7 - fitted to all first generation 9-5's (up to 2009-10-ish) and 9-3's up to 2003 has a greater scope of variables and parameters, still using the ion-sensing knock detection but uses a metered airflow torque-request based programming.
The added bonus of both Trionic 5 and 7 is that for quite some time, the ECU's have been mappable at home with basic electronic kit and a laptop. There are some who use a generic wizard to stage tune, and a few who have the necessary deep understanding to custom map to high bhp levels (a few cars run 400bhp+) and not destroy the engines. Gearboxes are another matter!
I had always wondered if two engine looms and ECU's from four cylinder engines plus the appropriate sensors could be used to run a V8 - each ECU and loom running one bank........
McVities said:
I had always wondered if two engine looms and ECU's from four cylinder engines plus the appropriate sensors could be used to run a V8 - each ECU and loom running one bank........
Not sure firing orders would lend themselves to that.But I think some of the old V12's, whether BMW or Jag were run pretty much as 2 x 6cyl engines as opposed to a single 12.
Gassing Station | Engines & Drivetrain | Top of Page | What's New | My Stuff