rrich

Yesterday a friend brought me his Renix Jeep. His complaint was just no power.
I drove it - to me it felt like it was not getting any timing advance.
Checked it with a light - rear wheels off the ground, top gear - it started to advance, then pulled back.

Looking at it - he had a bracket for something hanging loose, rattling. The knock sensor is just a little microphone listening for spark knock. I took off the unused bracket - timing went back to normal, drove fine.

Low power complaints - start with a visual inspection!

cruiser54

That's funny. Good advice.

DFlintstone

I'll be darned Rich, that's brilliant! So somebody driving around with 263K and a ghost of a rod knock might be perpetually retarded. (with a Renix Jeep),(lol in advance Cruiser).

Wonder what happens if you unplug the thing?

cruiser54

The ECU won't retard the timing when knock occurs.

DFlintstone

So it should otherwise function normal. Is it necessary to be in drive to look at the timing curve?

cruiser54

Yes, and No

DFlintstone

Much thanks again. HOWLING storm here 3 inches last night. Maybe I'll clear the Shop, (barn), enough to get the Jeep in there and get out the timing light.

cruiser54

That'll only take you til the spring thaw............

DFlintstone

Yea! Idk. WAS clear one spring day.....28 years ago....

rrich

I lifted the rear wheels and put it in drive to make sure there is nothing that will inhibit advance. Some vehicles inhibit advance until in top gear.

The computer is constantly trying to find the "sweet spot" for advance (EST). That point is just shy of detonation. It slowly advances timing until it "hears" detonation - the knock sensor. When it detects detonation it bumps the timing back down 2 or 3 degrees on each "knock." Then it tries again to advance. It drops down faster than it advances, preventing surging from the searching. It's a constant search for the best advance spot. That's the purpose of the knock sensor.

OBDII systems don't have one. The advance curve it mapped into the computer.

It's not just Jeep, many other vehicles use the same method.
If a rattle sounds like a knock, then the rattle constantly bumps the timing down to initial - resulting in low power.
It's not a common thing, but not unusual either. Through the years I've encountered several with that condition - mostly GM with a loose alternator bracket.

For the disbelievers - try it - use a timing light, run up the R's, tap near the knock sensor - you'll see it back down.
BTW -that's how you test a knock sensor anyway.

When I drove it it felt like others with the condition - runs smooth, idles fine, tries to accelerate but falls on it's face, no stumble or misfire - still sounded good, but sounded retarded (you get good at hearing timing over 45 years), just a gross lack of power. I didn't have a manual handy - so I tried it before investigating other troubles. Bingo! Simple fix - I removed the loose bracket.

I've never seen a rod knock do it - but I suppose it could.

DFlintstone

Much thanks Rich. Great info there. Last year I had the "gang" on my 4.0 short on the heat shield up front. Don't know if it was knock, temp or 02, (or combo). Ran like ****. (fired the mouse nest out the front, drove through confetti!)(no sign of the mouse ) Could well have been the knock sensor it sounds like.

Took me a long time to find that. Could start and drive all day. Turn it off for 5 min's and the shield would expand and contact the sensor wire('s).

A_Jeepr

First off Rich, Knock sensors are used on ODB II and Newer BUS/LAN systems.

Second, you can unplug the knock sensor as I did for a long time, and the computer will go into a closed loop (run like a static old school engine). The engine runs fine without the knock connected, and as I found ran better while I had a dead O2 sensor (didn't care at the time, work was 1 miles down the road).

Knock sensor listens for knock and then adjusts the timing slightly with-in a fraction of a second. It does not adjust it all the time (only when pre-ignition id detected). A loose bracket ratting with in about 3 inches away from the sensor will cause a false reading, you are 100% right about that. However that bracket had to have been VERY loose.

rrich

Some OBDII systems use them, some not. The 91+ Jeep does not. No knock sensor is used. The knock sensor with EST was used on the less sophisticated systems - ie: Renix, GM CCC etc..

The computer always searches for the sweet spot slowly advancing till it knocks,then backs off a tad, then starts over again. That way it stays as far advanced as possible, for the most efficient burning.

I don't know if unplugging the knock sensor will allow it to advance too far creating detonation. Unplugging it the computer may detect it's unplugged and may cause the computer to limit the total advance - never had a reason to try to defeat the purpose of it

I do know that if it's still connected but it no longer works as a microphone (defective) you will experience ping.

If the rattle noise travels to the sensor, it will affect it. You never can tell how noises will travel. The GM'S I've encountered where the alternator bracket was the usual culprit - about 8 inches away from the sensor. A loose air cleaner can do it!

Loose - loose enough to make the noise.

A_Jeepr

sorry I thought you were referencing all makes/models concerning the ODB II knock sensor.

I experienced no PING or knock when I had the sensor unplugged. I do run only ethanol free 87 octane fuel however.

1988cherokee

if you unpug the knock sensor it makes the computer go into default mode and it runs off of stored known good values. Thats why they usually run a little better with it unplugeed but for the 02 sensor when they go bad on the renix you get crappy fuel mileage and it defaults your short term fuel to 128 all the time. As far as the Alt brackets on gms messing with the knock sensor ive run into that before and he is right the same goes for a losses bracket hitting the block will mess with the knock sensor and retard timing