When I say knock sensor “hack” I mean HACK, ie. not a proper repair, just one that will work until the time comes when you are ready to remove the supercharger and possibly change out all the assorted valley hoses “while you are there” replacing the knock sensor.

After doing the requisite troubleshooting, and determining that all the wiring between the knock sensor and the ECU was not at fault, I knew I had to change or somehow eliminate the bank #2 knock sensor so the “restricted performance”, “check engine light”, and code 0333 would go back to wherever they came from and stop interfering with my happiness.


After some effort and with the help of the forum I located the connection for the bank 2 knock sensor. I unplugged the defective sensor and plugged in a replacement. It took a while to find a place to mount the replacement where it would stay put and not move around. There’s a scarcity of places on top of the engine where the sensor can be bolted down (none, zero) or otherwise fastened so it can’t flop around whenever the car makes a turn.

so I looked for ANY flat surface. What I settled on was a sheet metal bracket that holds a couple of wiring connectors on the bank 1 side of the engine. Since both the bracket and the knock sensor were composed of a ferrous material it seemed like a natural solution to use a strong magnet to hold the sensor securely to the sheet metal bracket.

200+ miles later, the sensor hasn’t moved at all, and all the error / warning messages have gone away. Now I can once more enjoy the car without interruption until the time is right to do the proper repair.


Z

PS.

a) this applies only to the 4.0 XKR, I don’t know enough about the later XKR’s to extrapolate.

b) having no supercharger in the way, the XK8’s have a much easier time of it when it comes to knock sensor replacement.

 

That is a genius solution. My hats off to you.

 

I’ll post photos tomorrow

Z

 

Photos attached


Z

Magnet on sheet metal bracket

Knock sensor held in place by the mysterious and invisible force of magnetism.

Unplugged defective bank 2 knock sensor

 

You do realize I hope that the detection of knocks is now compromised?

I can't tell you to what extend, that is a risk you take.

 

Thats why it’s called a hack fix, not intended to be permanent.

I use top tier 91 or 93 octane gas. And the bank 1 sensor is in the correct position & working fine.

I have zero concerns about engine knock, But thank you for the comment.


Z

 

I would describe it differently, what is hacked it that you disabled an engine safety feature that lets your engine go into "restricted performance" mode to prevent engine damage.

Knock control and the ECM relying on that has evolved from the AJ26 to AJ27, and more on the AJ34 engines, it is not just for WOT conditions, the ecu is enhanced enough to understand which cylinder has the knock and will retard the timing for that cylinder in steps.

Anyway its your choice, and I do way more dangerous things to my engine, my point is just that you (and or others that want to do the same) know what the potential effects are.

 

That’s the internet for you, one minute your brilliant, and the next minute you’re a reckless fool (my words ).

I’m not one who’s reluctant or afraid to admit a mistake, or hide any unintended consequences of my actions. If I experience any engine damage, changes in perceived performance, or unusual occurrences of any kind, I’ll try my hardest to determine if they can be traced to the knock sensor move. And post the results, if any.


Otherwise, expect reporting on the saga of the relocated knock sensor as time goes on. It might be spring (or later) before I get around to doing a supercharger R&R. Cold weather is especially hard on my back, so garage time is going to be self-limited for a while.

Until that time, I’ll be driving the car daily as usual. Living in a town of modest size allows me to enjoy the 2 lane rural roads and only endanger myself and / or the XKR when the speed limits are disregarded. There haven’t been too many days in the past 3 years of ownership when the car hasn’t seen a good workout; that will continue as well. If there turns out to be some unintended consequences from the knock sensor relocation my driving habits will surely bring those to the forefront. I’ve been pushing limits and boundaries all my life. Too late to change now.


Z


avos, I’d love to hear what “....more dangerous things you’ve done to your engine...”. And gotten away with it I’m presupposing.

 

As a guy that hopes to be driving one of these SC XKs soon, hats off to you sir... It is possible, likely, inevitable that in the ownership of this car I will deal with this issue, or a number of issues, and I, like you, am willing to be a little risky-er than some... Good temp fix and I'm sure everything is going to be just fine.

 

I’m not sure you understand the purpose of the knock sensor. As the previous poster was trying to say in the old days a knock sensor would wait for a knock at wot and then retard the timing or change the fuel trims to prevent the knock but in these cars they rely on the knock sensor as being one of the inputs to the equation across the rpm range. The ECM will adjust the timing of the cams the timing of the ignition the fuel mixture and other parameters do you get maximum power out of your car and then it uses that knocks sensor to listen for light knocks and then it cuts back slightly on the timing or some other parameter to prevent a knock. But with the knock sensor up all the way it’s going to keep advancing the timing because it’s never gonna hear a knock and you can guarantee that there will be damage. That’s why when the computer finds a fault with the knocks sensor it goes to a safe fuel mapping and ignition timing Routine so that you can get home using super low power mapping

 

i am doubting the ECU will be advancing the timing ad infinitum.. I’d have to see some Jaguar documentation to substantiate that, and your assertion of “...guaranteed damage...”. before I run out to the garage and rip off the supercharger. asap.

i didn’t get this idea out of nowhere after smoking some of Oklahoma’s fine medical marijuana. A forum searched me to another XKR owner who was faced with the same issue. They just laid the knock sensor on top fop of the engine with nothing keeping it from moving around loosely. And have been driving their car like that for months.

in any case, I’ve not posted this HACK in an attempt to get anyone else to follow me down the slippery slope of substandard repairs. If I’m guaranteed engine damage as you say, the internet will know about that as soon as I do.


thank you for your comments


Z

 

see if this helps

 

I had already read those pages prior to my adventure in temporary knock sensor relocation. But thank you for posting them nevertheless.

The part about the ECM retarding timing during “acceleration at critical engine speeds”, which I take to be high rpm and possibly high load situations, INDEPENDENT of the knock sensor signal is what sealed the deal for me to comfortably go ahead with my temporary hack. In other words, during times when when engine knock is most likely, and most dangerous, the ECM is already dialing back the timing regardless of any signal received or not received from the knock sensor(s).

Thank you again for the contribution to the topic. I’ll be in touch with the forum on this subject should my joy ride become the train wreck you suggest.

Z

 

I think the the sensitivity of the systems in these machines can spin in a direction of being overestimated,,, or the efficiency and functioning of the systems in these cars is overstated to the point of bordering on the unrealistic. That's just me. Lots of it is nuts and bolts, some technology - which is pretty old at this point that granted, was and still is pretty advanced - to an extent... I think Jaguar and manufacturers WISH that these systems worked as well. Especially 20yrs in.

I resist approaching all of this stuff like the machine is so sensitive it can't be touched,,, without fear. Attributing qualities and characteristics that don't match what's being dealt with. I mean,,, it's a car... That's just me. Fear sucks.

In the event that I end up with a knock sensor code or for some reason need to do this, I'll do it. I think it's an excellent temporary solution - considering what 100% fix means on the spot. Then, pay attention and tone the driving down a bit - for a while.

 

Accidentally came across this info for a AJ26 ECU on Knock control, as mentioned the AJ27 and AJ33/4 engines are more refined. Most important here is that you understand its not just at WOT knock is checked/handled, also at " tone the driving down a bit" as you say:

3.3 Knock (Detonation) Control Engine ignition is normally performed close to the detonation limit in order to improve fuel economy and increase power output. Knock sensing allows calibration of the ignition timing for normal operating conditions with the risk of detonation eliminated by the use of knock control. Two accelerometers (knock sensors) (1 per bank) mounted on the engine block are used to determine the onset of detonation by sensing engine vibrations. The sensors output signals to the ECM, which ascertains which cylinder has detonation present and retards the ignition timing on the affected cylinder to prevent this. The reference level (background noise) for the knock detection is calculated from the average value of sensor signal when detonation is not taking place.

You might want t use a higher octane fuel for the time being, though I can't guarantee that will safeguard you, it sure adds some extra safety.

 

I always use 91 octane pure gas, which is the highest octane in my area. Am also limiting my top speed to under 125 mph and am limiting full throttle acceleration to no more than 2 or 3 miles at a tone.

it is painful to hold the car back so drastically, but in the interest of prudence etc etc etc



Z

 

You can't source 100 octane racefuel? Sure, its expensive, but it should help prevent knocks.

 

Doesn't it seem funny that the ECU was satisfied with the new KS?
The one effort just laying one somewhere may not have been the best idea, but it seemed to solve the problem at that time.
Now Z has made a positive connection to the block by the strong magnet which really allows it to give some feedback to the ECU. After all, it's just a nut holding the KS to the block.
A magnet holding it to the block will still in effect connect it to that area.
Agree, don't push it and it sounds like Z has already gave it some excessive (to me) effort.
Magnets are a great tool to have in your box, when welding they give you freedom of both hands.

 

No, not at all, you just deprive the ECU of knock events.

Only thing that was solved was to block the ECU from protecting the engine, see my post.

Its not that simple, remember, it is already important to exactly torque the knocksensot to the block at 20nm, and that is at the spot where it was calibrated for detecting the right amplitude of frequencies to understand what a knock is.

I guess the above answers this flaw right?

My extra post was to ensure you understand it is not at WOT alone, also under normal driving.

Definitely true for ferrous material, I missed something like that for my aluminium welding.

 

the 91 octane pure gasoline (no ethanol) is working very well. I’m not going to be finding race fuel when pleasure driving day trips or going cross country either. But thanks for the suggestion.

I have zero evIdence of engine knock. It’s just not happening. I do appreciate the suggestions from forum members, but the sky isn’t falling. If and when it does fall, I’ll be the first to admit I was wrong.

So far I’ve put about 1,800 miles on the knock sensor hack; a variety of short trips, 7 hour day trips, and countless full throttle passing events. The car’s performance is identical to what is was before the KS relocation.

According to Jaguars own publications, the ECU is already retarding the timing when the engine is at a high rpm / high power event. Those are times when engine knock is most likely to occur. It’s not like there is currently no ECU spark control with one of the two KS’s relocated.

the XKR is my daily driver. I’m logging about 20,000 miles yearly. A lot of that is day trips of 400-500 miles round trip. The balance is a mix of local driving’s and 3-4 yearly cross country trips. If I was
going to have a problem I think it would already have shown some sign of happening .

in any case it’s an interesting topic. Stay posted, if the sky falls on me I’ll have a double helping of crow and humble pie.


Z