Installing a later model knock sensor on a neon/earlier stuff

[Aries_Turbo]

so im setting up my sohc neon to do some knock box testing to make sure its detecting knock correctly before moving onto the next phase of the project.

i wanted to use a newer knock sensor that has a wide range and use the box to do the filtering for flexibility. something like the SRT-4 neon knock sensor. last i checked (today) you can get one on rock auto for less than 13$ plus then i can standardize the box recommendations on one sensor.

since the neon already has a knock sensor on the front of the block and i didnt want to remove that cause its my daily driver right now, i chose to put it on the back of the block and see what the box picks up.

problem is that all the holes on that block are tapped m10 x 1.5 and the sensor wants a m8x1.25 bolt to hold it down. the srt 2.4L blocks have a boss on the block under the intake, tapped at m8x1.25. i dont know what non srt 2.4's have but both dohc and sohc 2.0L neon motors that i have, dont have a tapped boss on the block that size.

now i could have drilled out the sensor like DJ did with his shadow build but i didnt want to do that for fear of messing up the sensor.

no problem though. got out the grinder and the tap/die set and made a custom stud. m10x1.5 on one end and m8x1.25

http://www.turbo-mopar.com/forums/photopost/data/522/IMG_1718.jpg

i ground off pretty much all the thread on a bolt till it was a smooth rod with m10 threads on it and then ran the die down over it. worked great.

here are the bits that im using.

http://www.turbo-mopar.com/forums/photopost/data/522/IMG_1716.jpg

the sensor in the pic isnt the actual srt sensor though it looks identical. i have an srt sensor in the garage. i have a feeling that the output will be the same.

i took a file to the centermost tapped boss on the back of the neon block to clean it up and make it flat, then ran a tap down it to clean up the threads.

then i greased up the new stud and screwed it in, slid the sensor over it, then the washer in the pic, then the nut and torqued it down to 15ftlbs to start. i believe the SRT torque spec is 195 in/lbs so im a touch under that. you dont want to overtorque these sensors.

ill see if i can get a pic of the sensor installed when my hands arent all greasy. :)

Brian

[Reaper1]

I'm interested to see how this goes. I'd like to see if these could be used on the 2.2/2.5 for better knock sensing.

[Mopar318]

How are you monitering the knock? I really want to use a moniter of some sort on my SRT4 swapped 1st gen. Megasquirt does not have any type of knock moniter thought as far as I know.

[Aries_Turbo]

risen, frank and i are developing a box that does all the decoding and spits out a digital signal to interface to a megasquirt or possibly the factory ecu though i need to check some stuff out on that front.

i want to check to see how often the stock ecu samples the a/d for the knock sensor.

we really have decent knock detection on our factory ecu's though compared to alot of aftermarket ecu's and other 80's factory turbo cars.

MS does have a knock input pin. it samples it pretty slow on the MS2 code.... like every 50ms which is an eternity compared to how fast this box can respond.

http://www.turbo-mopar.com/forums/showthread.php?t=33393&highlight=knock&page=4

you can always connect the sensor to the block and run shielded cable into the cabin and hook it to the mic/line in input on a laptop and record/listen to the signals with headphones like DJ does.

Brian

[risen]

Let's hope it works as well on car as it does on the bench. Fingers crossed.

[Mopar318]

Let's hope it works as well on car as it does on the bench. Fingers crossed.

Keep us informed. I see there is a box that works with megasquirt by sending a signal to one of the unused pin inputs, but it is not compatable or able to retard the timing with the waste spark system.

[risen]

Keep us informed. I see there is a box that works with megasquirt by sending a signal to one of the unused pin inputs, but it is not compatable or able to retard the timing with the waste spark system.

It shouldn't matter whether the setup is waste spark or not, if MS is controlling the spark it should be able to retard the spark. Once we feel like we've got some solid testing in I'll be able to do bench testing with my MS here to make sure the spark is getting retarded.

I can measure the spark advance angle externally, so I'll know if the MS is lying to us or not :)

[turbovanman²]

How are you monitering the knock? I really want to use a moniter of some sort on my SRT4 swapped 1st gen. Megasquirt does not have any type of knock moniter thought as far as I know.

DJ aka turboshad made something up using similiar sensors, its in his project log.

[Aries_Turbo]

DJ hooked the knock sensor up to a laptop with the mic/line in port and a resistor to knock the amplitude down some. then he listens himself with a pair of headphones.

we are making a box that does the knock determination itself.

brian

[turbovanman²]

DJ hooked the knock sensor up to a laptop with the mic/line in port and a resistor to knock the amplitude down some. then he listens himself with a pair of headphones.

we are making a box that does the knock determination itself.

brian

I know, which is damn cool, :hail: I was throwing that out as DJ usings Megasquirt, :p

[omnigoestohell]

How did you guys make out with this box? I am currently doing the same thing, although instead of making my own box I am trying both the KnocksenseMS and trying using a GM knock filter off any of the 80s/90s cars with a similar bore size. Currently it works with my stock sensor but I like the newer through-bolt type sensor better because its shielded and more servicable.

Do any of you guys know the thread/pitch on the stock knock sensor on the T1/T2/T3 head/intake ?

Instead of making your own stud you could use a metric step stud like this..
https://www.belmetric.com/s8x10x38-p-1399.html?zenid=5nlfh5lj5cd1lmp3hf03l8lei6&cPath=9_73

[Aries_Turbo]

we arent done testing. partly cause i have the only prototype and i have too much on my plate.

that said, id like to get outside today (ugh hot) and swap out the sensor and cable with the new one i made and give it a run down the road. its nice and hot and i can get trace knock on my neon going up hills lugging it with the A/C on.

Brian

---------- Post added at 02:55 PM ---------- Previous post was at 02:51 PM ----------

yeah that stud is almost exactly what i made. it only took me a few min with a factory bolt and a grinder and a die to make what i made rather than waiting for shipping but if you are already ordering a sensor, you already have to wait.

yeah i like the newer sensors too but you have to have bandpass filtering with them as the are pretty much a fairly wide frequency range microphone. thats what the TPIC chip does, filters out only the frequency that you want to listen for.

Brian

---------- Post added at 02:58 PM ---------- Previous post was at 02:55 PM ----------

just relooked at the thread.

i believe the box has picked up knock fairly well so far.

i was having some intermittent signal stuff though thats why i got another sensor and connector and some nice shielded microphone wire to use.

im going to set a goal for today to get a little bit of testing done on the setup. after i see if my k car seat brackets bolt up to my daytona enthusiast seats. :)

Brian

[Aries_Turbo]

well i got the sensor installed and the wiring cleaned up. now i need to do some testing. maybe tomorrow.

Brian

[omnigoestohell]

I figured out the knock threads appear to be standard 3/8 course pitch so I got impatient, didn't buy the step stud and instead made up my own stud as did aries_turbo this morning. i am using 6.5khz for the knock frequency, is this the same you guys are finding knock on if you analyze the audio in a spectrograph?

[Aries_Turbo]

http://www.phormula.co.uk/KnockCalculator.aspx

detonation resonant freq=900/(3.14*cylinder radius in meters) i found on another site. i have it written down home somewhere.

brian

[omnigoestohell]

Cool, doing that math it comes out to 6557hz for a 3.44" bore, which is just about what I have the box tuned to. I too just have the wiring cleaned up though and have to put the intake manifold and other pieces back together before I can test, we'll see how that goes by the weekend. Motec has some great instructions for recording engine knock and then visually displaying the recording so you can see exactally at what frequency your motor is knocking and how to decipher between knock and regular engine noise rather than relying on the theoretical calculation based totally on bore size. I will try this at some point too but need to probably do what DJ did and do something to knock down the amplitude of the signal a bit before i feed it into my microphone in port on my laptop..

[Austrian Dodge]

i bought the phormula ks-4 a couple weeks ago and will be testing it as soon as i have some time!

[Aries_Turbo]

i did some testing the other night on the knock box.

it is displaying good values for engine threshold noise with a few blips above the threshold indicating knock.

i put some 93 in the tank and ill do some baseline noise datalogs and set the device up with that data and then put the cheapest, crappiest gas in the car and make a hot air intake and run the ac and load it up the hill by my house and see what it picks up.

then with a little bit more testing of:

1. location of the sensor and check the noise distribution between cyls

2. audible testing vs actual knock detection with either a sound card or some detonation can earmuffs.

3. testing on my k car when its fixed.

4. adding the 3.0L distributor signal stuff and testing that.

it should be ready for use.

Brian

[Reaper1]

Sounds awesome! :thumb:

[Vigo]

woohoo!

[Aries_Turbo]

hmmmm. i had 93 in the car today and went for a spin.

im still getting little blips of signal above the overall noise curve of the motor.

i guess im going to have to hook up some listening devices sooner than i thought.

i also want to record the "sound" coming out of the sensor to see if there is any nasty electrical noise getting into the circuit.

maybe ill pick up a can of torco to ensure that the thing isnt knocking too.

Brian