I received my knocksenseMS from Boris today.

I'll install and start playing with it.

What settings are you using for stock pistons?

 

I'm using exactly half pot so halfway between low and high.....so far this seems to be accurate enough since I've noticed detonation slightly from cyl #4 plug after it shows its pinged from the LED.

I finally got the tip-out pinging under control for the most part.....My enemy now is simply High IAT's therefore I'm going to start using heat shielding tape or header wrap for the intake + charge pipes on top of making a colder air ram intake (might remove a highbeam since its an integra) and get a turbo blanket to get the IAT's as cool as possible

 

You could play with the iat compensation for timing also...

I think that's under advanced settings.

 

Oh indeed I have and have staved off enough detonation by timing being pulled at certain temperatures but too much timing pulled does nothing but shift heat toward the cooling system and the exhaust valves where they can get hot and trigger detonation too. I just need to address how the intake is fed to the turbo since its never been great in terms of IAT's in my car since I bought it.

 

Where on the block did you end up placing the Knock sensor?

 

^ not installed yet - I was going to emulate your install, just need to pick up a couple parts from Toyota, including that stud you posted above.

 

Gotcha....Ill take pics of a spare block I have in my garage to show exactly where I mounted it

 

If you are using the Toyota stud then you can just put it in the stock location. Detonation happens near the cylinder head and not in the at the crankshaft. (Yes I am well aware that detonation affects the rotating assembly as a whole.)

 

This isn't a bad idea either as common sense would say to place it where the OEM sensor was placed however I ended moving it to a lower position mid-cylinder #3 to try to get it further away from valvetrain noise which I read that some had issues with on other forums but hell I might try moving it to do comparisons or might just get another sensor and do the the dual sensor setup so ones uptop between Cyl #1-2 and one between Cyl# 3-4 mid-lower block.

Black R, here's where i placed mine....

 

I'm interested to see how that spot works out for you.

Post an audio file if you can once you get it working.

 

Me too, I have major issues with valvetrain noise myself on my built N/A B20/VTEC...

 

I will when I get the audio circuit working..... I put together a noise isolated circuit from the diagram you posted the link on and used the .01uf capacitors that were rated upto 50v which didnt specify if they were polarized but couldn't get it to work with the amp.

I found that it interfered with the knock detection too however instead of splicing in the circuit I simply tapped into the connections on the KS unit just as the signal and ground tap in since this should have offered a means of tapping the signal without cutting wires but I suppose this wasn't correct.

Also, on the circuit diagram, I set everything up as the polarity shows. The center wire (Shield) is ground and the Signal wire (Clear).

 

Wait, say that again...

You mean, center wire is the signal, outer wire is the shield, correct? Did you use a mono headphone jack or a stereo one? Are you sure the caps aren't polarized? The leads should be different lengths, the longer one is positive.

 

Sorry, just rechecked that, Clear Signal, Ground is the braided mesh wire......The packaging didnt specify if they were polarized and Radio shacks employees are no help since all they are anymore are phone salesman who've never picked up a soldering iron but I didnt know to check the lead lengths to identify which is the positive over negative side of the cap......I figured since the write-up said to get non-polarized that (assuming they weren't polarized) it didnt matter which lead was used.....

It was all really cheap and found the PERFECT solder for future projects that contained flux so when melted it instantly bounded with the surface of everything quickly.....

I'll redo this again and try to find caps that I know for sure aren't polarized if they were.....headphone jack was a stereo, not mono

 

Okay, that's correct.

But yeah, typically if it's polarized, one lead will be longer than the other or it will be marked:

http://www.learningaboutelectronics....onger_lead.jpg

Polarized are more common than unpolarized, and will have a lot of leakage if hooked up backwards.

Try hooking up your circuit without them and see if it works.

Also, make sure you have your headphone jack wired up correctly. If you are using a mono plug or jack with the headphone amp (which is stereo) you will short the ring to the sleeve and nothing will work. You can run the shield to the sleeve and the signal to either the tip or the ring and you will get one speaker output. If you bridge the tip and sleeve you will then have two speaker output.

And yeah, Radioshack stinks now. Unfortunately, it's really the only choice if you need something now and can't wait for a Digikey order. I'm lucky enough to have a pretty good surplus place locally, but it's a lot like fishing. Sometimes you find exactly what you want, sometimes something that will do, and sometimes you come home empty handed.

 

Thanks, I'll keep trying till I get it since it seems fairly straight forward so its likely the caps were probably polarized even though it never specified on the packaging.

When I get it working I'll post audio samples of low,mid, and high rpm engine noise and hopefully catch a glimpse of light detonation as bad as that sounds so that it can be identified. I've heard high frequency sound samples of Det from a piezoelectric sensor before posted on other websites which sounds like a weed eater making blunt contact with a wooden stake or something.

 

Yeah it's easy to pick out.

 

No update with sound clips yet since I still have to make another audio circuit however I did want to update that I fabricated a cold air intake for the turbo inlet since what was in place wasn't cutting it.

The turbo compressor is facing the passenger side like all B series AC friendly manifolds however this manifold will be swapped in the long from for a SPA topmount cast manifold which will flip the compressor to driverside facing but in the mean while my biggest issue has been high IAT's especially on hot Florida summer days....

Since the CAI has been added I've seen fewer knock blips between shifts except on a few hard pulls after the motors fully warmed up (during daytime temps, nights are fewer).

I've been aiding what I can in addition to the CAI by adjusting the temperature thresholds for when ignition retard begins for IAT's and ECT's which seems to be getting me in the right direction now.

Will post further updates.....

 

As of yesterday I think I've made my biggest break-through yet.....A well known issue I've been having is pinging unheard to the naked ear that my KS unit keeps detecting (on tip-out, not tip-in) between shifts on med/hard pulls mainly in 1st and 2nd gear....I've played around with many of the compensations for Ign, Fueling, and have tried enriching areas of the lowcam map where they might be coming from in addition to other changes with testing in street tuning.

With all that has been tried up till this point I've only been able to reduce knock ever so much until the motor is fully warmed up of which the frequency of detection was still not acceptable in my book until I tried an alternate MAP sensor configuration.....

I'm currently using Crome free version....my ecu had a badly soldered datalogging header soldered in of which the previous owner did and I'm unable to really log anything in real-time until I switch out ECU's otherwise I know this would have been realized sooner.

I'm now using the stock Honda MAP "test" configuration in Crome... I have revised my fueling for the differences is pressures for the "test" MAP calibration and I've been able to kill almost any presence of knock detected after the engine is fully warmed up and even heat soaked. All knock detection seen between shifts has seized and any knock seen has only happened while on the throttle during a hard pull in second gear (before shifting) momentarily so I know there's still further refinements to make but all in all I'm very pleased this was recognized by use of this device since the previous owner did not have this technology on hand as he had the vehicle tuned.

 

Wait, you're using crome free?

I'm using pro, not sure how you're doing so well with the free version...

Anyway, was it the map sensor or the configuration for it in crome that ended up being the culprit?

 

Yes....not using it to be cheap but this aspect will soon change when my new ecu comes in but as I said, the datalogging header is fried from poor soldering attempts by the past owner so I can't datalog therefore theirs not too much I saw reason to jump into cromepro with just yet....I have a wideband + the KS so its not as though im completely in the dark...

And yes, at the expensive of harsh criticism im sure, the culprit was the MAP config it seems however the image the car was tuned on by the past owner was the non "test" Honda stock sensor and he said he dyno tuned and daily drove the car with upto 7psi for 8yrs so im not pleased to know that the car was running under these conditions for that long but with next year's taxes the motors being rebuilt anyhow..

 

My last comment wasn't intended to sound critical.

I pulled the trigger way back when and paid for crome pro just so that I could datalog for tuning.

I understand what you're saying about the header on the ecu being fucked up though - that sucks.

 

No offense taken....I really should have installed a newer P30 usdm board by this point so I could be using atleast CromePro.....I have kids, wife, with bills up the *** *sigh* so as much as I want to put my car ontop of my income expenses sometimes it has to wait....

 

Here's a link to a sound sample I recorded on a WOT pull I made on my way home from getting smokes last night. Took me a while to get this working with the correct "Metalized" capacitors and all but now I guess I have to learn how to analyze the output i'm hearing.

https://drive.google.com/file/d/0B_3...it?usp=sharing

In this clip I used the amp and recorded with SoundForge 7 however I probably shouldn't have used the amp since the volume of the output was so high that it clipped the sound at certain volume dB where as when I tested w/out the amp the entire spectrum of the sample was visible. I also converted this sample to Mp3 from a WAV file for upload on my G-Drive.

upto 8K, 7psi, 4th gear....

@1:29-1:30 theirs a slight static like pop but I cant tell if that was a ping or just interference on the signal or output.

On a side note, I'm really curious to see how a different ground would do since its currently grounded through the chassis as the install for the KS recommended. I'll test the ECU ground and maybe buy the shielded cable for the KS since I'm using the std cable.

 

Just replied to your PM, glad to see you got it up and running.

Changing the ground location probably won't make a difference since you're just powering the unit and not relying on it to provide a calibrated signal to the ecu. As long as the ground is good you should be fine.

I think what I'm hearing in there is just noise or the signal clipping, it doesn't have the distinct sound of detonation. You'll know it when you hear it.

*Note: I only listened up to about the two minute mark.

BTW, vtec heads sound sooo much better than non vtec heads. Your clip makes my motor sound like a total turd, lol.