ECU with knock control
Discussion
Sorry for the basic questions, i'm an engine guy, not an ECU person!
I'd like to be able to run high CR to build a powerful turbo road car engine which has a modern feel and easy to drive. I'm thinking of getting an ECU with Knock control to help. Is it asking too much for the ECU to save the engine from destroying itself when such conditions arise? I mean as a road car, you are very rarely going to use (say) 400bhp regularly and constantly for 1/2 an hr at a time on the hottest day of the year after you've mistakenly filled it with 5yr old 95 Ron from Tesco, but...
I'm guessing it's down to how advanced the system is, but how will I know? It's quite new and the KC is described as having "2xDSP Knock inputs with programmable gain" and individual cylinder knock control.
More info is probably available If I ask.
The guy who does my mapping is familiar with the ECU range, but hasn't done one of these yet, although he says there shouldn't be any issues.
Just wondered what you thought really.
I'd like to be able to run high CR to build a powerful turbo road car engine which has a modern feel and easy to drive. I'm thinking of getting an ECU with Knock control to help. Is it asking too much for the ECU to save the engine from destroying itself when such conditions arise? I mean as a road car, you are very rarely going to use (say) 400bhp regularly and constantly for 1/2 an hr at a time on the hottest day of the year after you've mistakenly filled it with 5yr old 95 Ron from Tesco, but...
I'm guessing it's down to how advanced the system is, but how will I know? It's quite new and the KC is described as having "2xDSP Knock inputs with programmable gain" and individual cylinder knock control.
More info is probably available If I ask.
The guy who does my mapping is familiar with the ECU range, but hasn't done one of these yet, although he says there shouldn't be any issues.
Just wondered what you thought really.
Edited by Evoluzione on Wednesday 15th June 14:51
Evoluzione said:
Sorry for the basic questions, i'm an engine guy, not an ECU person!
I'd like to be able to run high CR to build a powerful turbo road car engine which has a modern feel and easy to drive. I'm thinking of getting an ECU with Knock control to help. Is it asking too much for the ECU to save the engine from destroying itself when such conditions arise? I mean as a road car, you are very rarely going to use (say) 400bhp regularly and constantly for 1/2 an hr at a time on the hottest day of the year after you've mistakenly filled it with 5yr old 95 Ron from Tesco, but...
I'm guessing it's down to how advanced the system is, but how will I know? It's quite new and the KC is described as having "2xDSP Knock inputs with programmable gain" and individual cylinder knock control.
More info is probably available If I ask.
The guy who does my mapping is familiar with the ECU range, but hasn't done one of these yet, although he says there shouldn't be any issues.
Just wondered what you thought really.
I cant see a scenario where you'd ever fill with the wrong fuel....I'd like to be able to run high CR to build a powerful turbo road car engine which has a modern feel and easy to drive. I'm thinking of getting an ECU with Knock control to help. Is it asking too much for the ECU to save the engine from destroying itself when such conditions arise? I mean as a road car, you are very rarely going to use (say) 400bhp regularly and constantly for 1/2 an hr at a time on the hottest day of the year after you've mistakenly filled it with 5yr old 95 Ron from Tesco, but...
I'm guessing it's down to how advanced the system is, but how will I know? It's quite new and the KC is described as having "2xDSP Knock inputs with programmable gain" and individual cylinder knock control.
More info is probably available If I ask.
The guy who does my mapping is familiar with the ECU range, but hasn't done one of these yet, although he says there shouldn't be any issues.
Just wondered what you thought really.
Edited by Evoluzione on Wednesday 15th June 14:51
But for the system to properly protect an engine against "stupidity" for want of a better description, then their knock control will not only need to be very very good, but also very flexible and it will take a while to configure so it isnt too sensitive and pulling timing far too often.
Just because an ecu claims knock control....doesnt mean all systems are equal.
It's like boost control.....the w/g itself is "boost control", and so is an all singing/dancing ecu based boost control that can be configured around many other parameters.. Same name, very different abilities.
Really you need to speak to people familiar with whatever system you're talking about, and getting real world experience of it.
Two (major) factors limit the overall performance of knock control systems:
1) The complexity / suitability of the control strategy
2) The accuracy / repeat ability of the calibration of that strategy
IME, very few aftermarket mappers understand the complexities of knock control beyond the basics, and hence struggle to properly, and robustly calibrate these systems. Added to which, because of the lack of knowledge, the strategies themselves often leave a lot to be desired.
A proper, OE style knock control calibration will start with hardware validation and signal processing validation, focusing on signal to noise ratio and frequency distribution. It will then move onto control response and ignition angle compensation (including adaption) and finally validate the chosen parameters against a wide range of limit hardware and under a huge gamut of environmental conditions. Typically, KCS cal/val for a single engine variant will absorb two engineeers full time for something like 6 months for the first test bed cal, and then 1 enigneer for another 6months on validation (inc in-territory testing)
If you have no knowledge of robust data acquisition and high speed signal processing, then chances are the KCS will not be optimally calibrated under all conditions. Hence, the more aggressive you are with your base ignition tables, the more chance you have of wrecking things when the KCS fails to respond in a suitable fashion.
Also remember that you will need to calibrate the system to compensate for changes in ignition angle (false or real) and that uncontrolled retard can be as damaging (due to exhaust line temperatures) as excessive advance.
1) The complexity / suitability of the control strategy
2) The accuracy / repeat ability of the calibration of that strategy
IME, very few aftermarket mappers understand the complexities of knock control beyond the basics, and hence struggle to properly, and robustly calibrate these systems. Added to which, because of the lack of knowledge, the strategies themselves often leave a lot to be desired.
A proper, OE style knock control calibration will start with hardware validation and signal processing validation, focusing on signal to noise ratio and frequency distribution. It will then move onto control response and ignition angle compensation (including adaption) and finally validate the chosen parameters against a wide range of limit hardware and under a huge gamut of environmental conditions. Typically, KCS cal/val for a single engine variant will absorb two engineeers full time for something like 6 months for the first test bed cal, and then 1 enigneer for another 6months on validation (inc in-territory testing)
If you have no knowledge of robust data acquisition and high speed signal processing, then chances are the KCS will not be optimally calibrated under all conditions. Hence, the more aggressive you are with your base ignition tables, the more chance you have of wrecking things when the KCS fails to respond in a suitable fashion.
Also remember that you will need to calibrate the system to compensate for changes in ignition angle (false or real) and that uncontrolled retard can be as damaging (due to exhaust line temperatures) as excessive advance.
How many cylinders does the proposed engine have?
If 4-cyls you could do far worse than a loom and ECU transplant from a Trionic 5 managed car (Saab 900 '04-'99 or Saab 9000 late '93 to '98)
Lots of Vauxhall owners are swapping the engines into their cars complete, but no reason why the ECU etc won't run another 4 cylinder engine with all the sensors swapped in.
Trionic 5 is rather mappable with ion-sensing knock detection for each cylinder and variable ignition, boost and fuelling according to any knock detected.
As an asides, it's also very tweakable and is used to run some cars at 450bhp+, with fuelling, ignition and boost all programmable using a laptop and a cable. There may be an electronic niggle on earlier cars (16MHz ECU) that limits revs to 7,500rpm, but I think the later 20MHz ECU's don't suffer with this.
If 4-cyls you could do far worse than a loom and ECU transplant from a Trionic 5 managed car (Saab 900 '04-'99 or Saab 9000 late '93 to '98)
Lots of Vauxhall owners are swapping the engines into their cars complete, but no reason why the ECU etc won't run another 4 cylinder engine with all the sensors swapped in.
Trionic 5 is rather mappable with ion-sensing knock detection for each cylinder and variable ignition, boost and fuelling according to any knock detected.
As an asides, it's also very tweakable and is used to run some cars at 450bhp+, with fuelling, ignition and boost all programmable using a laptop and a cable. There may be an electronic niggle on earlier cars (16MHz ECU) that limits revs to 7,500rpm, but I think the later 20MHz ECU's don't suffer with this.
Edited by McVities on Friday 1st July 14:27
Max_Torque said:
Two (major) factors limit the overall performance of knock control systems:
1) The complexity / suitability of the control strategy
2) The accuracy / repeat ability of the calibration of that strategy
IME, very few aftermarket mappers understand the complexities of knock control beyond the basics, and hence struggle to properly, and robustly calibrate these systems. Added to which, because of the lack of knowledge, the strategies themselves often leave a lot to be desired.
A proper, OE style knock control calibration will start with hardware validation and signal processing validation, focusing on signal to noise ratio and frequency distribution. It will then move onto control response and ignition angle compensation (including adaption) and finally validate the chosen parameters against a wide range of limit hardware and under a huge gamut of environmental conditions. Typically, KCS cal/val for a single engine variant will absorb two engineeers full time for something like 6 months for the first test bed cal, and then 1 enigneer for another 6months on validation (inc in-territory testing)
If you have no knowledge of robust data acquisition and high speed signal processing, then chances are the KCS will not be optimally calibrated under all conditions. Hence, the more aggressive you are with your base ignition tables, the more chance you have of wrecking things when the KCS fails to respond in a suitable fashion.
Also remember that you will need to calibrate the system to compensate for changes in ignition angle (false or real) and that uncontrolled retard can be as damaging (due to exhaust line temperatures) as excessive advance.
This1) The complexity / suitability of the control strategy
2) The accuracy / repeat ability of the calibration of that strategy
IME, very few aftermarket mappers understand the complexities of knock control beyond the basics, and hence struggle to properly, and robustly calibrate these systems. Added to which, because of the lack of knowledge, the strategies themselves often leave a lot to be desired.
A proper, OE style knock control calibration will start with hardware validation and signal processing validation, focusing on signal to noise ratio and frequency distribution. It will then move onto control response and ignition angle compensation (including adaption) and finally validate the chosen parameters against a wide range of limit hardware and under a huge gamut of environmental conditions. Typically, KCS cal/val for a single engine variant will absorb two engineeers full time for something like 6 months for the first test bed cal, and then 1 enigneer for another 6months on validation (inc in-territory testing)
If you have no knowledge of robust data acquisition and high speed signal processing, then chances are the KCS will not be optimally calibrated under all conditions. Hence, the more aggressive you are with your base ignition tables, the more chance you have of wrecking things when the KCS fails to respond in a suitable fashion.
Also remember that you will need to calibrate the system to compensate for changes in ignition angle (false or real) and that uncontrolled retard can be as damaging (due to exhaust line temperatures) as excessive advance.
Interesting reading as I'm just about to do an engine swap in my MX5 and drop in a 4L 1UZ lexus engine
The 1UZ has a knock sensor with the stock ECU but the emerald ECU I was running in the MX5 with the supercharger doesn't have a knock sensor and I was considering whether or not it was worth selling the Emerald and replacing it with an aftermarket standalone ECU that did support knock sensors
I'm now thinking it's probably overkill and, as you say, how can we be sure it's totally accurate
I think I might stick with the Emerald and just rely on getting it properly mapped
(The 1UZ has a 12 tooth trigger wheel on the crank and a cam phase sensor at TDC which the emerald can handle)
The 1UZ has a knock sensor with the stock ECU but the emerald ECU I was running in the MX5 with the supercharger doesn't have a knock sensor and I was considering whether or not it was worth selling the Emerald and replacing it with an aftermarket standalone ECU that did support knock sensors
I'm now thinking it's probably overkill and, as you say, how can we be sure it's totally accurate
I think I might stick with the Emerald and just rely on getting it properly mapped
(The 1UZ has a 12 tooth trigger wheel on the crank and a cam phase sensor at TDC which the emerald can handle)
As the others have said and you have found out, it's time consuming to set-up proper knock control though basic settings can be applied if you know the knock frequency and you have someone that is experienced with setting it up
McVities said:
How many cylinders does the proposed engine have?
If 4-cyls you could do far worse than a loom and ECU transplant from a Trionic 5 managed car (Saab 900 '04-'99 or Saab 9000 late '93 to '98)
Lots of Vauxhall owners are swapping the engines into their cars complete, but no reason why the ECU etc won't run another 4 cylinder engine with all the sensors swapped in.
Trionic 5 is rather mappable with ion-sensing knock detection for each cylinder and variable ignition, boost and fuelling according to any knock detected.
As an asides, it's also very tweakable and is used to run some cars at 450bhp+, with fuelling, ignition and boost all programmable using a laptop and a cable. There may be an electronic niggle on earlier cars (16MHz ECU) that limits revs to 7,500rpm, but I think the later 20MHz ECU's don't suffer with this.
There's a high possibility that the knock frequencies that the ECU is programmed to listen for are different, not to mention that ionic knock sensing requires either special spark plugs or pressure sensors within the combustion chamber themselvesIf 4-cyls you could do far worse than a loom and ECU transplant from a Trionic 5 managed car (Saab 900 '04-'99 or Saab 9000 late '93 to '98)
Lots of Vauxhall owners are swapping the engines into their cars complete, but no reason why the ECU etc won't run another 4 cylinder engine with all the sensors swapped in.
Trionic 5 is rather mappable with ion-sensing knock detection for each cylinder and variable ignition, boost and fuelling according to any knock detected.
As an asides, it's also very tweakable and is used to run some cars at 450bhp+, with fuelling, ignition and boost all programmable using a laptop and a cable. There may be an electronic niggle on earlier cars (16MHz ECU) that limits revs to 7,500rpm, but I think the later 20MHz ECU's don't suffer with this.
feef said:
Interesting reading as I'm just about to do an engine swap in my MX5 and drop in a 4L 1UZ lexus engine
The 1UZ has a knock sensor with the stock ECU but the emerald ECU I was running in the MX5 with the supercharger doesn't have a knock sensor and I was considering whether or not it was worth selling the Emerald and replacing it with an aftermarket standalone ECU that did support knock sensors
I'm now thinking it's probably overkill and, as you say, how can we be sure it's totally accurate
I think I might stick with the Emerald and just rely on getting it properly mapped
(The 1UZ has a 12 tooth trigger wheel on the crank and a cam phase sensor at TDC which the emerald can handle)
Probably not needed with a stock N/A engineThe 1UZ has a knock sensor with the stock ECU but the emerald ECU I was running in the MX5 with the supercharger doesn't have a knock sensor and I was considering whether or not it was worth selling the Emerald and replacing it with an aftermarket standalone ECU that did support knock sensors
I'm now thinking it's probably overkill and, as you say, how can we be sure it's totally accurate
I think I might stick with the Emerald and just rely on getting it properly mapped
(The 1UZ has a 12 tooth trigger wheel on the crank and a cam phase sensor at TDC which the emerald can handle)
Si1295 said:
There's a high possibility that the knock frequencies that the ECU is programmed to listen for are different, not to mention that ionic knock sensing requires either special spark plugs or pressure sensors within the combustion chamber themselves
The Trionic 5 ECU runs current accross the plugs post-sparking event. The plugs used are NGK BCPRES 6 or 7 depending on level of tune (and very resonably priced).Point taken about different engines having differing characteristics in terms of frequencies, but that should make much less of an issue with ion-sensing.
Gassing Station | Engines & Drivetrain | Top of Page | What's New | My Stuff