barks944

Ive been looking into the knock detection on the 951. So far my understanding is that the DME sends an advanced ignition signal to the KLR which retards the signal a given ammount depending on wether or not it detects knock. The signal goes back into the DME circuitry from the KLR and the coil is fired. My question is, does the microprocessor in the DME know this has happened or is it entirely oblivious?

gt37vgt

as the ignition runs retarded with the KLR disconnected i support the theory that the timing is advanced by the relationship between the 2 boxes but i kind of thought the KLR would advance the timing
i can't imagine why the DME needs to know .. unless it was to ad more fuel to kill the knock and as it doe'sn't add fuel i would suspect the DME is not being told ..
the KLR is listening to the knock sensor and it pulls timing and boost when it knocks ...

oh i guess the dme is given knock to be logged for the service light stuff

i did have a pretty long thread on this a year ago search my profile for it ...

barks944

I think by default the spark signal from the DME is 4 degrees advanced from when the DME actually wants the spark. So the KLR will retard that signal 4 degrees to give the spark when the DME intended. Should the KLR detect knock it will retard by say 10 degrees giving a spark 6 degrees retarded from what the DME wanted.

barks944

I noticed in the DME schematics that there are two inputs/outputs from the DME's digital board into the ignition circuitry on the analogue side of the board. I wondered if one was to trigger the output and one to measure when the signal was sent back from the KLR.

toddk911

This is something Rogue can answer in depth if he posts in.

Everything I have ever read in here about knock and the DME/KLR is that in current set up (no chip/board mods) the computers do not react quick enough to pull out timing or pull out enough to save any damage.

Rogue_Ant

Your pretty much right on. I would have to check my notes to find the exact 'normal' delay the KLR does, but 4degrees sounds appropriate.
The microprocessor never knows if there is knock or not. Once the ignition event signal is sent, then everything else is handled by the KLR and ignitor (on the DME power board side).

Which pins are you referring to?

Rogue_Ant

The KLR is more then sufficiently quick to do the job. Even at 7000rpm an engine is quite slow compared to a computer (even a 1980s one).
The KLR most certainly has saved many engines, but it is not a band-aid for poor tuning, and ridiculously advanced timing of certain aftermarket chips...

333pg333

I don't know too many people running these in here, but I've seen and heard good things elsewhere.

http://www.jandssafeguard.com/

Rogue_Ant

Yep, the J&S system is great. But hopefully this summer I'll dive into the KLR, and see if I can't add some real functionality...

Richgreenster

My motor has a noisy lifter. Much quieter now that I am running 20-50 weight oil but still some tapping. My question: would the knock sensor pick this up and consider it a engine knock?

Also What does the Vitesse chip board do to this?

Rogue_Ant

The DME does not know/understand knock, regardless of chip you are using. Knock is done completely by the KLR.

The KLR is purpose-built to filter noise and listen for true knock... I suppose it could pick up your lifter and mistake it as knock, but it is designed not to.

gt37vgt

i always wanted to run the KLR with my motec but as it ajust's a single coil trigger and i want to run 4 coils i thought it might not work but now I think there will be one coil trigger out put from the ecu sent to a multiplexer that then triggers the coils so perhaps i can put the KLR in between 2 ..
hmmm
i was realy intending to scope all the inputs and out puts from the KLLR one day to better understnad it but i pulled mine apart befor it happened

Renn 951

Motec has a knock sensor option - but it ain't cheap!!

TRP951

my TEC3r has a knock sensor option

333pg333

Worlds longest stethoscope?