Programmable knock control systems
11-13-2007 | 03:02 PM
#74
Gold Member
Joined: Sep 2005
Posts: 1,987
Likes: 0
From: Missouri
11-14-2007 | 10:14 PM
#76
Registered
Joined: Jun 2007
Posts: 11
Likes: 0
RW? I read your last email, but it didn't click that you knew Ian.
Rage: I haven't responded, since the processors we were using are now very hard to get, though I have a contact in Asia that says they are still available.
Ian Whiteside is a calibration engineer at Ilmor Racing in the UK. He bought a system for his personal Formula 3 race car maybe three years ago. The engine is 2.0L, based on a 1.8L Eclipse engine, with 14:1 compression.
He said that with that much compression, the engine is severly knock limited, and can run no more than about 14° timing at 4000 RPM, with 98 RON fuel.
A MoTeC M400 was installed and adjusted on the dyno. Timing was optimized using knock sensing headphones, then was retarded two degrees as a safety margin.
The J&S was then installed, and timing was bumped two, four, then six degrees. The dyno operator could detect nothing more than occasional trace knock, yet the J&S was controlling timing on the knocking cylinders.
Remember, when you adjust for no knock, you are tuning to the worst cylinder, sacrificing power in the others.
Individual cylinder knock retard increases engine efficiency and power. Ian emailed that he was showing a four percent gain. Didn't sound like much to me, so I didn't bother plotting his spread sheet data for about a year. I was amazed to see the area under the curve when I did. The graph shows an eight percent gain when they switched to 102 RON, the spec fuel for his series. No other adjustments were made.
When you get to my site, click on the Vampire link, then scroll down to the dyno graph.
Rage: I haven't responded, since the processors we were using are now very hard to get, though I have a contact in Asia that says they are still available.
Ian Whiteside is a calibration engineer at Ilmor Racing in the UK. He bought a system for his personal Formula 3 race car maybe three years ago. The engine is 2.0L, based on a 1.8L Eclipse engine, with 14:1 compression.
He said that with that much compression, the engine is severly knock limited, and can run no more than about 14° timing at 4000 RPM, with 98 RON fuel.
A MoTeC M400 was installed and adjusted on the dyno. Timing was optimized using knock sensing headphones, then was retarded two degrees as a safety margin.
The J&S was then installed, and timing was bumped two, four, then six degrees. The dyno operator could detect nothing more than occasional trace knock, yet the J&S was controlling timing on the knocking cylinders.
Remember, when you adjust for no knock, you are tuning to the worst cylinder, sacrificing power in the others.
Individual cylinder knock retard increases engine efficiency and power. Ian emailed that he was showing a four percent gain. Didn't sound like much to me, so I didn't bother plotting his spread sheet data for about a year. I was amazed to see the area under the curve when I did. The graph shows an eight percent gain when they switched to 102 RON, the spec fuel for his series. No other adjustments were made.
When you get to my site, click on the Vampire link, then scroll down to the dyno graph.
11-15-2007 | 05:42 AM
#77
Registered
Joined: Oct 2007
Posts: 57
Likes: 0
Yep, Ian and I both used to work together at Cosworth and then Integral before he went on to Ilmor.
Individual knock control really is the correct way to go. Certainly calibrating Bosch systems when compared to the old method of open loop spark calibrated to worst pot and then retard 3 degrees, the ability of a proper multi channel knock control system to optimise every cylinder yields significant performance advantages.
I'm sure that with in vee intercooled pre SC carbed V8s ie likelihood of significant AFR changes from cylinder to cylinder down the blower, that a proper knock control system would give a major boost, as well as piece of mind.
I'll have to dig out some in-cylinder pressure traces and FFTs for everyone's amusement at this rate.
I'll get back to you separately on that sprint car question John.
Individual knock control really is the correct way to go. Certainly calibrating Bosch systems when compared to the old method of open loop spark calibrated to worst pot and then retard 3 degrees, the ability of a proper multi channel knock control system to optimise every cylinder yields significant performance advantages.
I'm sure that with in vee intercooled pre SC carbed V8s ie likelihood of significant AFR changes from cylinder to cylinder down the blower, that a proper knock control system would give a major boost, as well as piece of mind.
I'll have to dig out some in-cylinder pressure traces and FFTs for everyone's amusement at this rate.
I'll get back to you separately on that sprint car question John.
11-15-2007 | 10:13 AM
#78
Gold Member
Joined: Sep 2005
Posts: 1,987
Likes: 0
From: Missouri
John, I guess back to my original question "What is involved to configure a J&S unit to work with the pair of Mercruiser "non-resonant frequency" knock sensors versus a single frequency tuned knock sensor?" May I assume from your comment below that the addition of a special (hard to get) processor to your basic J&S unit is required so that it will work with "non-resonant frequency" knock sensors? Would there be additional custom programing to taylor this additional processor to properly work with my unique engine or does adding the additional special processor alone produce a turnkey unit that will work properly with all "non-resonant frequency" knock sensors without any addition custom programing?
Thanks,
Bill
Thanks,
Bill
RW? I read your last email, but it didn't click that you knew Ian.
Rage: I haven't responded, since the processors we were using are now very hard to get, though I have a contact in Asia that says they are still available.
Ian Whiteside is a calibration engineer at Ilmor Racing in the UK. He bought a system for his personal Formula 3 race car maybe three years ago. The engine is 2.0L, based on a 1.8L Eclipse engine, with 14:1 compression.
He said that with that much compression, the engine is severly knock limited, and can run no more than about 14° timing at 4000 RPM, with 98 RON fuel.
A MoTeC M400 was installed and adjusted on the dyno. Timing was optimized using knock sensing headphones, then was retarded two degrees as a safety margin.
The J&S was then installed, and timing was bumped two, four, then six degrees. The dyno operator could detect nothing more than occasional trace knock, yet the J&S was controlling timing on the knocking cylinders.
Remember, when you adjust for no knock, you are tuning to the worst cylinder, sacrificing power in the others.
Individual cylinder knock retard increases engine efficiency and power. Ian emailed that he was showing a four percent gain. Didn't sound like much to me, so I didn't bother plotting his spread sheet data for about a year. I was amazed to see the area under the curve when I did. The graph shows an eight percent gain when they switched to 102 RON, the spec fuel for his series. No other adjustments were made.
When you get to my site, click on the Vampire link, then scroll down to the dyno graph.
Rage: I haven't responded, since the processors we were using are now very hard to get, though I have a contact in Asia that says they are still available.
Ian Whiteside is a calibration engineer at Ilmor Racing in the UK. He bought a system for his personal Formula 3 race car maybe three years ago. The engine is 2.0L, based on a 1.8L Eclipse engine, with 14:1 compression.
He said that with that much compression, the engine is severly knock limited, and can run no more than about 14° timing at 4000 RPM, with 98 RON fuel.
A MoTeC M400 was installed and adjusted on the dyno. Timing was optimized using knock sensing headphones, then was retarded two degrees as a safety margin.
The J&S was then installed, and timing was bumped two, four, then six degrees. The dyno operator could detect nothing more than occasional trace knock, yet the J&S was controlling timing on the knocking cylinders.
Remember, when you adjust for no knock, you are tuning to the worst cylinder, sacrificing power in the others.
Individual cylinder knock retard increases engine efficiency and power. Ian emailed that he was showing a four percent gain. Didn't sound like much to me, so I didn't bother plotting his spread sheet data for about a year. I was amazed to see the area under the curve when I did. The graph shows an eight percent gain when they switched to 102 RON, the spec fuel for his series. No other adjustments were made.
When you get to my site, click on the Vampire link, then scroll down to the dyno graph.
11-15-2007 | 11:03 AM
#79
Registered
Joined: Oct 2007
Posts: 57
Likes: 0
No, I think he's saying that the processor he uses is hard to get full stop. Irrespective of input type.
The main difference between the input stage for resonant and non- resonant is the level. Just turn the gain up. That's all there is to it.
The main difference between the input stage for resonant and non- resonant is the level. Just turn the gain up. That's all there is to it.
11-15-2007 | 01:29 PM
#80
Gold Member
Joined: Sep 2005
Posts: 1,987
Likes: 0
From: Missouri
Really. I have the general specs on my Delphi knock sensors, copy attached, but have no idea what the resonant knock sensor output versus engine vibration specification is. I am getting the impression that the resonant sensor output is low (filtered) until the "tuned resonant frequency" is reached by at which point the sensor output signal amplitude increases significantly, like a step function, and then returns to a low amplitude above the "tuned resonant frequency". Is that how it works or other?
