blackeg

does vtec create "noise" that shows up on the knocksense that may not actually be knock?

ive got a pretty mild 82x89mm lsvtec setup. itr cams, sk2 manifold, toda rep header, and a static comp ratio right about 11:1.

i have the low cam ignition tuned well, i dont have any knock. at like 3500 rpm its running 38 degrees around 20" down to around 24-25 degrees at atmospheric and the light stays off. as soon as i get into vtec the light goes nuts. ive tried running anywhere between 22 and 28 degrees with no difference. afr is just shy of 13:1 .

the plugs dont really look bad, im not really seeing any evidence of detonation down on the porcelain just some crap from additives in the fuel up toward the electrode. is the sensor getting false readings or am i knocking and need to redo either something mechanical or in the tune? this is my first go around using knock detection and not just relying on plug checking after pulls and monitoring afr changes in relation to timing changes

Muckman

iTrader: (2)

Yes any mechanical noise can be picked up by knock sensors. The trick is in the filtering. I've had piston to head and piston to oil jet contact show up on the knock sensor also. Is your light on just when VTEC activates or the entire hi cam map? Maybe you have some valve contact.

The great thing with the knocksense is you can easily turn down the sensativity. I calibrate it pretty high instead of idle rpm to get a better baseline. I also like the knock calibration table with AEM EMS too.

It sounds like you've already tried the obvious. If you significantly reduce ignition timing in the area of "knock" and the knock is still present then its not real knock - it could be nothing or mechanical noise.

blackeg

hey man, thanks for replying i know you have a lot of experience here. sadly it looks like here in the hudson valley we are going to get a snowstorm every few days im not sure when ill have the car out next. i guess what i really should do is put a fresh plug in there and do a pull and shut it down at max tq. id like to dyno this thing at some point but want to have all the bugs worked out so im not wasting my time and money.

as soon as it crosses onto high cam the light goes on hard for a second, it will flutter a bit up to around 6500 where it stays bright again. i guess i just thought it would be more inclined to knock on the low cam but its not. i find it weird i can run more timing on the low cam...i just turned the sensitivity down a bit yesterday with the adjustment inside the unit but no avail. it was originally calibrated on my bolt on b20z. i kept adding timing till i got it to light up at moderate load in the mid 2000 rpm range. i was able to get it up a few degrees above the p75 map which seemed reasonable to me. if i really go counterclockwise a lot on the adjuster will it still pick up only the worst knock or will it render it pretty much inactive?



the valve or internal contact is not out of the question sadly. these gsr pistons on ls rods sit pretty close to the deck height. im running a cometic .040" messing around on the zeal that gives me about .040 pth provided the gasket crushes like .003 as the calculator says the oem gasket does.

Muckman

iTrader: (2)

Try disabling VTEC or moving the engagement point to around 7k. See if it lights off on the low cam now but at the same high rpm.

blackeg

solid advice. ill get the car back out soon hopefully and ill give that a try. right now i have the crossover at i think 4800 ill move it up and see how it behaves in the 5000's rpm range.

i also had the plugs out today and took myself a look down into the cylinders. i didnt see any shiny specs on the piston tops. you know how like sometimes detonation can knock the carbon off the top of the pistons in random small speckles.

Muckman

iTrader: (2)

I dont think its detonation. Ive seen the light flicker when VTEC engages too. But I am worried that the light is solid on the hi cam. The sensitivity may be too high or it may be something mechanical like a vibration or contact. Once you confirm that the light is false you can turn the sensitivity down until it goes away. I have mine all the way down.

DC_Legacy

Hey man, I remember you bought your KS not long after I got mine. At first I really didn't have a clue about what was real vs. false with this unit since it picks up alot of mechanical noise and so I followed "spAdam" s recommendation and made my self one of these and this right here allowed me the means of really seeing what was truly knock over exterior noise that the KS was picking up on such as my BOV between shifts....

Electronic Det cans (to tune the KS) - https://drive.google.com/file/d/0B_3...RQN093c1E/view

Rightnow, for my setup I'd say I have the **** turn 2/3 - 3/4 to the right in which I no longer get blips for background noise or between shifts and I know its not knock because I've heard the difference and have heard real det up high when in vtec @ WOT (7-8 psi) since I'm still using Crome and with the use of this listening device + the KS I was able to fine tune my highcam map enough to where there is no Det brought on by the HighCam Ignition bugs of Crome which I found a work around for, for those here who still use it.


https://honda-tech.com/engine-manage...aided-3230463/

DC_Legacy

If you want, I can make one of these for ya since its pretty simple and cheap, just paypal me the shipping and I'll make one, test it, and send it your way....

theYBLEGAL

Before I buy a KS sensor, was thinking of connecting my stock knock sensor to an amplifier and some headphones, and just using my ears. Have you ever tried this DC_Legacy? Maybe that could help OP as well.

You can also record the signal with your PC input, and run it through something like Audition or Audacity to slow it down, or filter out particular frequencies.

Thoughts?

DC_Legacy

I haven't tried this with a stock knock sensor but I don't see why it wouldn't work. With the KnockSenseMS and this circuit and the way this circuit taps into the polarity of the Anode and Cathode of the knock sensor inputs, It gets powered by the supply of current the KSMS is using to power the everything so it would be a slighly different approach I suppose but its doable I'd say, just haven't tried it yet with an OEM sensor.

And yes you can record the output with a PC MIC using what ever software you like, I used an older version of Sony sound Forge (SF7) here in this clip... (at 1:02-1:04 the sound your hearing is Det between 7500-8000 that I later realized in identifying it with this circuit however the recording itself distorts the intensity of the clip)

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

Filtering software could help to identify it....for my account I learned at the expense of real Det since I use Crome still but it doesnt bother me because now I know what I'M listening for and so this tax season I'm rebuilding my setup or maybe stepping upto an LS block but new pistons and other internals are going in this year..

Aside from recording it though, the mini AMP is a must when listening on your own. I didnt buy the exact model recommended in the PDF write-up, I bought a unit at roughly the same price from Radio Shack

spAdam

iTrader: (1)

I think you might be unimpressed since the stock sensor is resonant. It will work well over a very small bandwidth and will be pretty quiet elsewhere. Simply swapping it out for a Bosch one and hooking that up to your headphones will let you hear everything.

Here's some good info on knock sensors:

Turbo Triumph Web Site

Unfortunately "BigMoose's" page (referenced in the above) is no longer up. It had some really good analysis and even delved specifically into the Honda sensors.

I need to set up a LabView experiment and do some analysis on mine. It picks up so much stuff, especially with the noisy non-vtec valvetrain. Even catches the shift linkage going into reverse. I could get away with turning the sensitivity down but I haven't had time to experiment lately.

You can get real fancy with sliding windows and FFT translations and get real detailed with it. You can even pick out the non-knocking noise and specifically exclude those frequencies in your analysis... it's just a matter of how far you want to take it.

blackeg

so i noticed the wire from the knocksense to the sensor was just about lying on my alternator idk if that would make a difference. i had the car out the other day, the knock sensor just likes to go off. i drove though it like i shouldnt have and then killed the car at about 7000 rpm and checked the plugs. there is a tiny bit of what is likely light detonation just on the very end of the porcelain where it is still protruding. i always learned that the closer to the electrode the lower the load and the farther down the porcelain into the plug itself the higher load like wot. the color change was just past the bend on the ground strap but not near the base of the ground strap.

oh well, snow is here, dont see a day in the next week at the moment that ill have the car out again

ThE HeLLRaZoR

Don't disable Vtec and hit the rev limit EVER! Vtec Lobe can float on the lma and hit the Retainers(most common big aftmrket Cam).. As far as knock goes I would never count on it making mediocre adjustments... If your numbers are around 26-31* around 4-5krpm in vtec and wot, I would not blame the numbers.... Maybe check to see Timing is synced proper.. If your making a new tune or making big mods, I'd highly recommend you disable Knock Controls or watch its Trims% closely...

DC_Legacy

It's questionable since you found what looks to be Det.....I've mentioned before where I have the tuning **** at on mine but your's maybe louder or quieter per mechanical noise....

The wire near the Alt though I'd move assuming you already have since you found it there, I bought mine with an unshielded cable but I'm considering getting one since it offers that extra layer of isolation from any source of EMI since I have the wire going through the wirewall aside the main harness bundle...

Just wait though, once you get the opportunity to listen to everything you'll be able to get the KS noise floor tuning where it needs if its blinking on background noise half the time rightnow

OneBadTurboCRV

Well another problem is with today's fuel and additives, engine condition, as well as other variables reading plugs can be a huge PITA. Sometimes the specs on the end of the porcelain are just contamination and not detonation.

I have seen this on a couple cars I have installed the KnockSenseMS on. We over retarded, left stock, and advanced the entire map and did light driving as well as WOT pulls to experience the same phenomenon. We even tried switching to E85 and added a healthy amount of good Torco octane booster (can't buy it at parts stores) with same effect.

After trying everything we could think of including bore scoping each plug pull we concluded this is what was happening - impurity contamination. We went through literally hundreds of plugs between two cars trying to figure it out over the course of a year. The solution to our woes was electronic amplification. We just thought I had lost my mind and couldn't tune worth a **** anymore haha.

In my opinion it is a must to have electronic amplification as by the time you tune out the noise on the KnockSense's potentiometer with a built engine, the light is pretty much useless. On stock engines it works pretty decent though it still throws false positives, so it really comes down to your specific situation.

***DISCLAIMER*** I never said I was a professional tuner, just a hobbiest that doesn't trust anyone else with my things. So take it for what it's worth.

blackeg

haha I like your disclaimer. im in the same boat just a hobbyist I actually teach high school as my day job.

so it seems that unless you basically tune the knokcsenae attempting to read plugs for det and timing mark on the groundstrap is a slightly better bet ?

ive seen the additives on the plug I think because they aren't shiny more dull in color.

im currently in the 23-26 degree range at wot. I did some map smoothing last night and moved the vtec up to I think 5200 just from the way the motor wanted fuel.

I did move the wire away from the alternator I have it following the brake booster line now but idk when ill be oit again with it.

i also know the right exhaust valve is pissing out like 15% leakdown maybe I forgot to lap it. that cylinder is only aboit 215 psi the others are 235-240. I hate my life ...

spAdam

iTrader: (1)

All of his cables are EMI shielded. The option is for a Teflon sleeved cable for high temp use.

Still, I agree that it's far from ideal to have that thing anywhere near an alternator.

Black, if you've got a leaky exhaust valve it's a total wildcard. If that thing is getting hot it will do all kinds of weird things.

blackeg

yeah sadly i do have a bit of a leak out of that valve. ill go back and double check the lash too at least make sure its not hanging in the cyl too long. i have another head at the machine shop as we speak getting a full overhaul and some 70lb supertechs and k20 retainers. i dont know how rush i am to fix it currently. this motor is just in the car for a few months so i can have it driveable as a spare car while i do a bit of freshening up on my built boosted engine. (that thing was going low 11s and id like to crack 10.9 with it this year)...after that i think this motor is shed bound but who knows. i did move the wire away from the alt. it follows the brake booster line and goes through the firewall through the spot where the other throttle cable that was for the cruise control went.


had a heat wave today, 31 degrees. had the car out for about 45 minutes. spun the tires at 7k in 3rd so i upshifted and logged 4th gear too. egt was sitting 680-700c making me happy. i finally stopped being a bitch and turned the sensitivity like 4/5 down. i also swapped some lightly used bkr7e's in and took the 6's out. im up to 25-27 degrees with the light staying off with the lower sensitivity but the porcelain was a bit hard to see with the slightly used plug. i know im ballpark but id like to get it tight then maybe dyno it who knows...

OneBadTurboCRV

Yes, in my opinion the KnockSense is just a visual aid as it comes from Boris but it depends on your situation. My semi built B20 VTEC is so damn noisy especially at 9k RPM it rendered the KnockSense useless. In my opinion adding a jumper resistor is extreme but sometimes necessary. I would rather rely on plug reading as the KS is just an extra tool in the tuning tool box.

I was trying to use the KSMS features for automatic timing retard with both the P30 and P72 ECUs I have using CROME at first and then tried the S300. I then tried same thing on a internally stock B16A2 with same results - severe power loss from random KnockSense output. What I noticed on my datalogs is that different engines have different "starting levels" or baselines of knock. What I mean by that is between the two engines the B20V creates enough noise for timing retard output off of the KSMS unit to activate ECU timing retard at idle... Given, the stock knock board is over sensitive anyways but you should be able to control knock only output from the KSMS anyways.

It makes it a major PITA, but I have hundreds if not thousands of logs from these two specific project cars that would make you want to burn your car to the ground. Knock being logged at different points back to back - same gear, same road, same WOT, etc but have noise at 6k and then 7k but maybe only 5.5k another time. It's frustrating to say the least. Maybe I am too much of a perfectionist but if you drove my B20V you would never know I have S2 Pro 2 cams, 12.5:1 comp, 70mm TB, Ultra Intake, etc. Aside from maybe the 900 - 950 RPM idle. You can put it in 5th and have the RPM down to 1.2 - 1.5k RPM without bucking. It drives like factory but I always shoot for longevity in any build I do.

I highly recommend you download the Virtual Dyno program. I have compared it to my silly GTECH Pro RR (whatever the best one is) and it seems much more accurate and within the ball park of the random number generating GTECH. It takes your datalogs and plots it out on a chart. Use a smoothing factor of 3 or 4 at minimum and it seems damn close if you do calculations based on fuel usage.

spAdam

iTrader: (1)

'Nother vote for VD. It's a highly valuable tool for any street tuner. I do need to find a good stretch of road though. That's my biggest hangup. Of course, getting off my *** and finding someone to rent me some time on a steady state dyno would be the real best solution.

blackeg

so virtual dyno is worth it huh? i probably should mess around with it. adam is right though, really should get on a real dyno.

i have a 5pm traffic court on monday next week. 72 in a 55 trying to tune out a rich spot at 3800 rpm. it was at 615am on my way to work one day in december, in the middle of no where through a corn field no other cars around. probably cost me around 300 bucks all said and done i could have2-3 had 2-3 hours on a dyno...


here's a random question: around when is a switch from 6 to 7 heat range plugs on an all motor b series a good idea? is it the power level? compression ratio etc?

OneBadTurboCRV

If you are that lucky to have a dyno that close then I am not sure why you wouldn't. I would have to drive hours to get to one... let alone a few hours more to someone that I would actually consider letting touch my car. I have seen more than a handful of " reputable t00ner" tunes that I have collected from people over the years that would make you cringe.

Why didn't you just datalog in a lower gear and simulate the engine load needed with the brakes if need be. I do this only when I have to or as a last resort, but it is much better than getting a ticket.

As as far as plugs go, you essentially want to run the hottest you can while keeping detonation and / or preignition at bay. Some people use horsepower ranges, but over the years in my experience it has been based strictly on setup. Every aspect of the engine factors into the equation as well as the fuel you are running. There are mechanical as well as chemical thresholds to be breached for detonation / preignition to happen. From what you said about your set up, 7's would be playing it safe but shouldn't affect drivability.

spAdam

iTrader: (1)

Agreed, BUT... generally I've switched to 7's on anything at or above ~11.2:1, and I've never had any driveability or fouling issues observing that rule. I'm sure that some of the setups would have been fine or even marginally better with a hotter plug, but in the absence of data I'd rather err on the safe side. I tried a set of 6's in the B20 a year or so ago though and it was not happy at all. BKR7E for life

OneBadTurboCRV

I hear you, I currently run 7's with my 12.5:1 compression. I forgot to mention that before. I tried 8's in the name of science, but when I switched to E85 it wasn't happy haha. I was trying to utilize the S300 v3 "Flex Fuel" feature but found it to be utterly useless being everything is globally applied. You would need AEM v2 at minimum to have a true flex fuel setup. My motor wants WAY more fuel than normal with E85... but I like the safety factor it provides and I can experiment more without fear! Although during the winter I usually switch back to 93 if the turbo is off.

blackeg

well i just want to thank you guys for keeping this thread going i feel its a good discussion and myself and others are learning from it.

onebad - do you suffer mpg with that e85 since its stoich ratio is substantially lower than gasoline? i am a complete ethanol noob

i have quantumraceworks (synapse's old dynapack) an hour north of me and xenocron is an hour south. either way i would be tuning it this would just be a rent the dyno scenario.

i should have loaded it in a lower gear. it was literally 615 am and i was only speeding for 4-6 seconds no other cars for over a mile from me. i was wideband watching and not rookie state trooper out to write tickets watching. i also had a wicked handlebar mustache going as a bet with a friend and im sure they didnt help me look like a clean cut adult lol.

i noticed the other day that after some pulls with the 6's the knock led was more active just regular driving, i think it had a bit of extra cyl temp or slight hot spot on the plug for a bit not sure though. thats why i went back to 7's just to be safe. i think they are gapped at .038.

my internally stock b20z i put 6's in it it seems happy running pretty much a p75 timing map with a few degrees out at low rpm load and high vacuum i have at like 38 degrees not 40+ .

my 84.5x89mm boosted gsr is 10:1 and i run 8's in it (440whp on pump). the gap is super low like .018 on the stock ignition but it runs fine in vacuum i even get over 30mpg with it.

happy friday everyone