steady 9 counts of knock above idle and off boost

[PC5505]

okay, so for the last couple days, my 1g has been showing a steady 9 counts of knock at any rpm, under all throttle and boost(or lack there of) conditions except idle(even cruising in 5th), basically any time its in gear and under load. Obviously this has resulted in a ton of timing being pulled and a total lack of power. This doesn't seem to match the descriptions of phantom knock I've read about so I'm at a loss.

Background:

-I've been running 14 lbs on an e16g w/ a 190 w no knock for a year.
-Using MMCD
-There is a CEL, but its for my temp sensor, (21) this is because my wiring harness is shot and the wire cracked off the connector. Its been like that for some time, with no ill effects. It used to only be an intermittent CEL when the wire was just loose, but now its full time. I don't believe them to be related and will fix it when I find a local partout I can scavenge from.
-ECU Caps replaced.
-Timing belt replaced less than 6 months ago by Mitsu dealer
-My positive battery terminal has a bad connection and fairly often I have to jiggle it around until it starts. I have a new terminal in hand I will put on soon.
-The last couple of weeks my logger has been showing strange readings, ie; momentary knock count of 255 off boost, timing that is way out of the range of possiblity(like 63 advanced) intermitently for just a second then returning to normal. I'd never have noticed, except I have it set to record peak numbers. Sounds like ECU? Maybe doing something wierd when I hit bumps and the terminal breaks/makes contact?


Thanks in advance.

 

[absolute_DSM]

Well if you think its an ECU, you could find a buddy and swap real quick to check.

 

[PC5505]

Yeah, unfortunately I dont really have anybody nearby that I can swap with...

Has anyone ever had readings like that on mmcd? Its been flawless up until now. I was gonna reinstall, but that wont fix my knock issue, since those readings are definitely accurate and I can feel the timing being pulled almost as soon as the car starts moving. I'm sure its not a boost leak, because it knocks while off boost the same as on.

 

[yokotabrat]

Usually when this happens, the ecu can no longer tell that the knock sensor is there anymore because the ground connection has been lost. It goes into a type of limp mode that defaults to 9 counts of knock to protect the engine. I would say it sounds like either the knock sensor has gone bad or the wiring to the knock sensor is bad.

 

[PC5505]

thanks yoko, I'll take a look at the wiring tonight and get back to you. Hopefully its that simple. Maybe I'll have time to do the terminal at the same time and elimintate two possibilities at once. Shouldn't it throw a CEL if the sensor was out?

 

[PC5505]

Well, I found a disconnected wire. It looks like it was held on to another wire by wrapping them around each other and covered with electrical tape. I really hate the previous owner. The knock sensor only has 1 wire coming out of it. Should it have two? I'm assuming this wire would connect to it somehow, but its not long enough to reach it and I can't find another disconnected wire around there, so I'm not sure where it went to. Just because I cant find it doesn't mean its isn't there, its very tight back there and like I said, my wiring is a mess.

Do knock sensors come with a connector of some kind? Or are they like bosh O2 sensors with just the wire ends?

 

[yokotabrat]

Knock sensors do come with a connector. The harness side should also have a connector.

The wire for the knock sensor is shielded similar to a coaxial type cable. The inner wire is for the signal and the outer wire is the ground braid to shield the signal from noise.