Knock sensing: microphone sensors vs. ionisation current analysis
#1
Scooby Newbie
Thread Starter
Join Date: Feb 2004
Location: UK
Posts: 19
Likes: 0
Received 0 Likes
on
0 Posts
Knock sensing: microphone sensors vs. ionisation current analysis
AFAIK no Scoob ECUs use ionisation current analysis for determination of combustion chamber conditions. Det sensing can be performed using this technique, and is generally considered to be more reliable than a microphone-based knock sensor due to an immunity to mechanical noise.
A minority of car manufacturers have been developing their use of ionisation current analysis (notably Saab with their Trionic 5 system back in the early 1990s). Have any Scooby enthusiasts/tuners experimented with this approach in preference to conventional knock sensors?
Some explanatory links of interest:
http://www.mumintroll.org/axel/spark...tions/diss.pdf
http://www.fs.isy.liu.se/~larer/Projects/main.html
A minority of car manufacturers have been developing their use of ionisation current analysis (notably Saab with their Trionic 5 system back in the early 1990s). Have any Scooby enthusiasts/tuners experimented with this approach in preference to conventional knock sensors?
Some explanatory links of interest:
http://www.mumintroll.org/axel/spark...tions/diss.pdf
http://www.fs.isy.liu.se/~larer/Projects/main.html
#2
Scooby Senior
Join Date: Oct 2000
Location: Zurich, Switzerland
Posts: 3,105
Likes: 0
Received 0 Likes
on
0 Posts
Interesting, but appears to be a work in progress? Also mentions the fact that it's suitable for cars/ECU's currently monitoring spark plugs for misfires as a form of knock control - Scoobs use a separate sensor.
I suspect it's not that suitable for higher power turbo cars (mine can knock without me detecting any kind of misfire!), although the reference to Saab suggests it may be. Would be interested to hear Pats take on it
Richard
I suspect it's not that suitable for higher power turbo cars (mine can knock without me detecting any kind of misfire!), although the reference to Saab suggests it may be. Would be interested to hear Pats take on it
Richard
#3
Scooby Newbie
Thread Starter
Join Date: Feb 2004
Location: UK
Posts: 19
Likes: 0
Received 0 Likes
on
0 Posts
The ionisation current method of knock sensing has been used on 440BHP 2.3L turbo engines. I think the technique can yield so much data about combustion conditions (even crank angle and combustion phase) that inevitably some aspects remain in the lab undergoing research, while others have been perfected into consumer applications.
Isn't the Scoob's separate sensor a microphone fixed into the engine block? That is the more conventional method of detecting knock.
Isn't the Scoob's separate sensor a microphone fixed into the engine block? That is the more conventional method of detecting knock.
Last edited by Matthew Brockway; 18 May 2004 at 04:40 PM. Reason: Minor revision to text!
#4
Scooby Regular
Join Date: Nov 2001
Location: Leeds - It was 562.4bhp@28psi on Optimax, How much closer to 600 with race fuel and a bigger turbo?
Posts: 15,239
Likes: 0
Received 1 Like
on
1 Post
pat has cut heads up to work out where he can put pressure sensors into his cylinders to detect det.
Im sure he will find this thread..
David
Im sure he will find this thread..
David
#5
Scooby Senior
Join Date: Oct 2000
Location: Zurich, Switzerland
Posts: 3,105
Likes: 0
Received 0 Likes
on
0 Posts
Hi Matthew,
Yep - Scoob uses microphonic knock sensor bolted to the block. Don't get me wrong, I'm not suggesting it's a good system - I disable it on mine and rely and separate knock display and my right foot. It's better on the newer cars, but still has major drawbacks.
Any more details on the 2.3 turbo that was converted? What manufacturer, and were they already using the spark plugs to detect misfires as a knock correction system prior to the modification to ionisation method?
Some of the better aftermarket ECU's probably have inputs that can be configured for such a system, but I don't understand what needs to be added on the Scoob system for the ionisation current method to be feasible.
Richard
Yep - Scoob uses microphonic knock sensor bolted to the block. Don't get me wrong, I'm not suggesting it's a good system - I disable it on mine and rely and separate knock display and my right foot. It's better on the newer cars, but still has major drawbacks.
Any more details on the 2.3 turbo that was converted? What manufacturer, and were they already using the spark plugs to detect misfires as a knock correction system prior to the modification to ionisation method?
Some of the better aftermarket ECU's probably have inputs that can be configured for such a system, but I don't understand what needs to be added on the Scoob system for the ionisation current method to be feasible.
Richard
#6
Scooby Regular
Join Date: Sep 2002
Location: Glasgow
Posts: 429
Likes: 0
Received 0 Likes
on
0 Posts
Interesting stuff Matthew.
The VEMS (formerly Megasquirt-AVR) diy ecu project has ion-sensing as a development sub-project: http://www.vems-group.org/index.php?page=IonSense
but very much a work in progress.
I'd like to hear more of that 2.3 turbo engine.
Duncan
The VEMS (formerly Megasquirt-AVR) diy ecu project has ion-sensing as a development sub-project: http://www.vems-group.org/index.php?page=IonSense
but very much a work in progress.
I'd like to hear more of that 2.3 turbo engine.
Duncan
#7
Scooby Newbie
Thread Starter
Join Date: Feb 2004
Location: UK
Posts: 19
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by DuncanG
I'd like to hear more of that 2.3 turbo engine.
AFAIK there are no pressure sensors in the combustion chamber. The special resistive spark plugs are energised with a low current, with the level and nature of conductance across cathode/anode being analysed by the ECU. If the return current matches the pattern seen during knock, then countermeasures are taken. First timing is retarded, failing that the wastegate is opened.
Many owners of Saab 9000 owners have embarked on DIY tuning projects, with resulting power/torque well in excess of 300BHP/300ftlb. However, it's the kits supplied by the Swedish company Maptun that I referred to.
There's at least six stages, with Stage 6 being the 440BHP mark IIRC. Some have gone onto Stage 6+ which is around 500BHP. AFAIK the ion sense part of the ECU isn't altered by Maptun.
Maptun:
http://www.maptun.com
Saab 9000 2.3T 500 Hp, Stage 6+:
http://www.maptun.com/cars_9000_fredrik.php
Saab 9000 Aero +400Hp, Stage 5+
http://www.maptun.com/cars_9000_dareus.php
Saab 9000 2.3T CDE 475 Hp, Stage 6+
http://www.maptun.com/cars_9000_jonas.php
Saab 9000 CS Aero, MapTun individual, 440 Hp/580 Nm
http://www.maptun.com/cars_9000_magnus.php
Please note that I'm not on a Saab vs. Subaru thing here, and I fully acknowledge the limits of high-power with FWD. I just can't off the top of my head think of other highly-tuned turbo cars that use ion sensing for det protection.
Trending Topics
#8
Scooby Senior
Join Date: Oct 2000
Location: Zurich, Switzerland
Posts: 3,105
Likes: 0
Received 0 Likes
on
0 Posts
Sounds good - wonder why it's not used by other manufacturers...or is it and it's onl;y the Japanese that use cheap systems?
What system does BMW use, they have individual cylinder control I believe....which I guess the above is too.
Richard
What system does BMW use, they have individual cylinder control I believe....which I guess the above is too.
Richard
#9
Scooby Newbie
Thread Starter
Join Date: Feb 2004
Location: UK
Posts: 19
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by dowser
wonder why it's not used by other manufacturers
Maybe it's a Swedish thing LOL - what do Volvo use? I know Ford have their EDIS (Electronic Distributorless Ignition System), but I don't know if that can be used for ion sensing.
As you say Richard, Trionic has individual cylinder control (coil-pack instead of distributor).
#12
Scooby Regular
Join Date: Nov 2000
Location: 32 cylinders and many cats
Posts: 18,658
Likes: 0
Received 1 Like
on
1 Post
All looks very interesting.
My needs have changed a little recently - I was chasing breathing and ignition timing to get more and more power on cheap, effectively disposable EJ205 or EJ257, until I reached the limits of two items in my EJ257 motor - piston and headgasket, and not I think from detonation, although they will not help when trying to run nearly 200 BHP per litre. If I strengthen those items I think I will have a succession of further failures of different items for each 10-20 BHP increase in power and I can't be bothered with it, so my future tunes on my ROAD engine will run very sensible ignition timing - although the engines are disposable, the down time and hassle are not.
For a maximum effort tuned engine I think this setup could be superb. For an emissions and economy minded factory car similarly, along with variable compression ratio which looks very interesting.
My needs have changed a little recently - I was chasing breathing and ignition timing to get more and more power on cheap, effectively disposable EJ205 or EJ257, until I reached the limits of two items in my EJ257 motor - piston and headgasket, and not I think from detonation, although they will not help when trying to run nearly 200 BHP per litre. If I strengthen those items I think I will have a succession of further failures of different items for each 10-20 BHP increase in power and I can't be bothered with it, so my future tunes on my ROAD engine will run very sensible ignition timing - although the engines are disposable, the down time and hassle are not.
For a maximum effort tuned engine I think this setup could be superb. For an emissions and economy minded factory car similarly, along with variable compression ratio which looks very interesting.
#13
Scooby Regular
Join Date: Nov 2001
Location: Leeds - It was 562.4bhp@28psi on Optimax, How much closer to 600 with race fuel and a bigger turbo?
Posts: 15,239
Likes: 0
Received 1 Like
on
1 Post
so john, would cylinder pressure sensors not be beneficial to you..
I'm beginning to think the ideal setup is something that combines..
Per cylinder EGT, Ion Sensing and Cylinder Pressure Sensors.. could be pricey.. but maybe its time to do something a little different to the norm.
MBT every stroke would be nice.
now question is do you use ombt or smbt
optimax minimum best timing
scottimax minimum best timing
David
I'm beginning to think the ideal setup is something that combines..
Per cylinder EGT, Ion Sensing and Cylinder Pressure Sensors.. could be pricey.. but maybe its time to do something a little different to the norm.
MBT every stroke would be nice.
now question is do you use ombt or smbt
optimax minimum best timing
scottimax minimum best timing
David
#15
Scooby Regular
Join Date: Nov 2001
Location: Leeds - It was 562.4bhp@28psi on Optimax, How much closer to 600 with race fuel and a bigger turbo?
Posts: 15,239
Likes: 0
Received 1 Like
on
1 Post
yeah, but in theory with a fast processor and some drivers a bit of code, to watch the crank and cam signal to work out which cylinder is firing I could confuse the hell out of my MSD to retard timing
#17
On our cars, individual cylinder Missfire is derived from crank signal "instantaneous" acceleration and cam signal. I wonder wether knock detection couldn't be detected using the same kind of processing ?
#19
I don't know Dave, I would imagine that rather than measuring time differences between pikes and zero crossing, the ECU performs a full analog to digital conversion on cranck signals (cam signal is used for cylinder resolution only). Then what kind of algorythm is used...no idea...but some patents could be found.
#20
Scooby Regular
Join Date: Nov 2001
Location: Leeds - It was 562.4bhp@28psi on Optimax, How much closer to 600 with race fuel and a bigger turbo?
Posts: 15,239
Likes: 0
Received 1 Like
on
1 Post
Im just meaning thart most cars run 36-1, where as the subaru is a little odd..
You can work out which cylinder is firing, but im just wondering if it needs more resolution on the crank to detect them?
David
You can work out which cylinder is firing, but im just wondering if it needs more resolution on the crank to detect them?
David
#21
I don't think more resolution is required, even with our poor crank sensors. Just think in analog signal and not in "detected levels". the waveform is well predictable. When knock occurs, the signal shape should drastically and ponctually be altered.
Thread
Thread Starter
Forum
Replies
Last Post
Mattybr5@MB Developments
Full Cars Breaking For Spares
12
18 November 2015 07:03 AM
Sam Witwicky
Engine Management and ECU Remapping
17
13 November 2015 10:49 AM