"...hundreds of millions of dollars and thousands of man hours..." I doubt an OEM spends anywhere near that much, but way more than the aftermarket can afford. I've seen similar posts many times on different forums, as a way of saying an aftermarket "one size fits all" system can't work. True, most can't.
AutobahnPower installed my system on a customer's factory supercharged 2006 Mini Cooper, along with a Snow Water/Meth system and a 13% over speed pulley. The factory ECU was by Siemans, and employed a knock sensor.
The shop made a short video showing operation. I have it set to start where he leans into it, with the Snow kit turned off:
https://www.youtube.com/watch?v=tjA8xiPd3SU&t=19s
The top row of LED's on the J&S gauge show the amount of knock retard. The factory system does a pretty good job controlling knock, but the J&S still manages to make a few detections.
With the Snow kit activated, the unit detects no knock, as expected. It also demonstrates the system was responding to real knock in the first pull. If it were responding to engine noise, the LED's would have activated in the second run as well.
In this application, it's shown to be more sensitive than the factory system, while ignoring engine noise. Wire it in and go.
If using forged pistons, I recommend 0.006" clearance or less and letting them come to temperature before making a pull.
Knock sensing,
Moderator: Team
-
- Guru
- Posts: 9633
- Joined: Tue Nov 27, 2007 9:27 am
- Location: Chicago
- Contact:
Re: Knock sensing,
John, as yoy may recall, years ago we were running the J & S knock control system on a racing BMW.JohnP wrote: ↑Fri Jan 07, 2022 10:15 am Knock sensors are useful when you are using fuel with insufficient octane, for the conditions you are running. Then you are "knock limited".
If you are knock limited, you can map the timing with an ECU for RPM, load, water temp, etc. If no ECU you can do it with boost retard modules, etc. This is called "open loop" timing control.
Using pump gas, we over-advanced the timing and let the knock system retard the timing to the
"knock limited" curve setting. It worked great and we kicked ass. Track records tumbled.
Then we got greedy and began using race gas and promptly blew a head gasket.
The race gas prevented knock but the engine ignition was now over-timed, resulting in pre-ignition.
No amount of clever engineering can defeat the well determined fool.
Which we were.
Re: Knock sensing,
Thanks for the memories David.
Regarding pre-ignition, I didn't put on MY thinking cap. If you detect a pre-ignition event, retarding timing doesn't stop it. The fire is already lit before the spark happens.
Thinking cap back on...
Pre-ignition melts pistons and cylinder heads, right? How many consecutive pre-ignition events would it take to start damaging a piston or cylinder head?
I've heard it takes about two seconds, don't know if that's true. Assume pre-ignition at 4500 RPM for 2 seconds = 150 revs = 75 engine cycles.
Build a box that can "do something" within 10 or 20 cycles and you should be able save the engine. Sounds doable.
Regarding pre-ignition, I didn't put on MY thinking cap. If you detect a pre-ignition event, retarding timing doesn't stop it. The fire is already lit before the spark happens.
Thinking cap back on...
Pre-ignition melts pistons and cylinder heads, right? How many consecutive pre-ignition events would it take to start damaging a piston or cylinder head?
I've heard it takes about two seconds, don't know if that's true. Assume pre-ignition at 4500 RPM for 2 seconds = 150 revs = 75 engine cycles.
Build a box that can "do something" within 10 or 20 cycles and you should be able save the engine. Sounds doable.
Thinking is hard work.
Re: Knock sensing,
Hehehe, thanks John. Inductive is a great way and only maybe off by 2 rpm, and even 8 bit accomlished, luv itJohnP wrote: ↑Fri Jan 07, 2022 5:46 pm External cam and crank references are not needed to generate a knock window. Put on your thinking cap.
A relative cam "signal" can be generated internally from a software counter that counts ignition pulses. For an eight cylinder, the counter counts down from 7 to 0, then is reloaded to 7, every two revolutions. Then, "cylinder 0" can have one retard amount, "cylinder 1", another, etc. It doesn't matter the actual cylinder number.
For the Knock Finder unit, I used a timing light pickup on cylinder #1 spark plug wire, to synchronize the display and control knobs to the engine. The knock sensor portion of the unit did not need that, as it keeps track on a relative basis as described. It's only needed to tell the humans looking at the display which cylinders are knocking.
Thank you for your developments, awesmeness,
Heat is energy, energy is horsepower...but you gotta control the heat.
-Carl
-Carl
Re: Knock sensing,
Bill: I clicked on the link to Joe Sherman's interview but it doesn't go anywhere. I'll try to contact management.Bill Chase wrote: ↑Thu Jan 06, 2022 9:22 amBelieve he used a j&s setup, in the interviews forum here in the interview he did for winning engine masters he talks about it and mentions the model used
Oh I see I need to subscribe. Done.
Thinking is hard work.
-
- Guru
- Posts: 9633
- Joined: Tue Nov 27, 2007 9:27 am
- Location: Chicago
- Contact:
Re: Knock sensing,
As you well know, pre-ignition is very difficult to determine without in-cylinder pressure measurement.If you detect a pre-ignition event, retarding timing doesn't stop it.
Depends on which definition we elect to use.The fire is already lit before the spark happens.
Excessively early spark timing will cause a large increase in combustion temperature and pressure, closer to TDC
or even before TDC, and can be called pre-ignition.
Mixture that ignites without spark can be called auto ignition or pre-ignition as well.
And local mixture ignition, even after spark ignition, can cause the combustion pressure and temperature to
exceed design limits.
Yes, it does. And excessive rise in combustion pressure pushes head gaskets out of their sealing location.Pre-ignition melts pistons and cylinder heads, right?
This is often detected by acid formation in the coolant which is detected by pH indicators in the coolant.
Re: Knock sensing,
Oh I see I need to subscribe. Done.Bill Chase wrote: ↑Thu Jan 06, 2022 9:22 amBelieve he used a j&s setup, in the interviews forum here in the interview he did for winning engine masters he talks about it and mentions the model used
Bill, I just listened to part of the interview with Joe Sherman. At 5:35 he starts talking about the knock detector. I can tell from the description it was the Anson Knock Meter. I don't know anything about it or the guy that designed it. I'm guessing his name was Anson.
Thinking is hard work.