kung stew

Guys,

Time to poll the experts. I have Artguy's old car (custom T 60-2, PFC, Steve Kan tuned, water injection, all supporting mods, etc) and have been very pleased with it. After changing to an HKS Twin power last year, its been running great at 12-13psi on 9s all around. I noticed some higher than normal (for this car) knock in the no-no zone (p17, n14-18) during 3rd and 4th gear pulls about a month ago . Knock was in the neighborhood of 50s-80s, with 'baseline' knock in the 20s. I logged the runs, AFRs 10.4-10.9, and pulled 2 degrees of IGL out, and maintained a split of 15 in this area. This seemed to fix it for a couple of weeks, but now it is doing the same thing again. I pulled ANOTHER 2 degrees of IGL, and the problem persisted without change. I loaded the prior map (with 2 degrees retarded) and put 3 gallons of toluene into the tank to see if fuel was an issue, and to see if I needed to retard the ignition further (doesn't seem right to me), with a calculated octane of about 97. No change in knock at the same boost. AFRs are rich (<11), and the plugs have less than 1k miles on them. Car runs great, pulls strong, vacuum 16-17 inHg at 900rpm. No discernible knock or ping during these events. This all occurs w/ WI on, and I've confirmed that it is functioning. The kicker is that I also see knock occasionally in this rpm range (5500-7500) at zero boost. I would like to chalk this up to driveline noise, etc, but the area that the knock is occuring has me concerned. What do you guys think ?
-Stew

kung stew

I'm using Basic Knock, max values when I say "knock" above. I'd like to be able to post my maps and logged runs, but I'm computer retarded .

kung stew

anyone ever have a knock sensor go bad and do this (he asks hopefully)?

-Stew

kung stew

I seem to be carrying on a one sided thread discussion here. Oh well . I changed plugs (9s all around), and the numbers are about the same (improved top end power, though ). However, I now believe its just noise. The occasional 50-70 knock values I see under boost also occur at similar rpm at 0 boost or vacuum. I think that my knock seems more prominent in 4th gear b/c there is a long sampling time for the PFC to pick up whatever driveline/exhaust noise is causing it at those rpms, which is reproduced when I wind the rpms out in 0 boost/vacuum conditions. Its disconcerting to see knock numbers while boosting in n14-and up, but I think I've gone through pretty much all steps possible to rule out true detonation (increased octane, pulled timing, and rich afrs), and the car is showing no indication of knock (no stumbles, hesitations, noises, power loss, etc).

Why in the hell must I be so neurotic about my car?

-Stew

Narfle

iTrader: (16)

First, your knock reading isnt that high. Anything over 100 and I'd be worried.

Second, too much fuel and too little ignition can cause knock. 10.4 afr is really rich even for a rotary. 10.4 is rediculously rich with water injection.

Third, peak readings arent that important unless theyre way out of bounds. If your sustaining knock readings around 60-70 during steady WOT high gear pulls youre probably fine.

read this and do a search for "knock sensor" and "knock reading"

https://www.rx7club.com/showpost.php...83&postcount=8

pp13bnos

iTrader: (1)

While, its best not to ignore the knock numbers with the stock sensor.....I don't give it a lot of sway. My car with 110 octane fuel, would go 75+ as soon as you lifted your foot off the pedal. Turned out that the mid pipe was resting on one of the cross braces. I fixed that, but it still goes 60ish when you lift. As long as your a/f ratio is right on, and timing is safe, I would'nt worry about it to much.

The only thing that you might want to check too see when your injectors where last cleaned. While I don't think this is the problem, you might want to check just in case. My .02, CJ

RICE RACING

With those variables you wont get that amount of knock, with things being normal. Can you send me your log file so I can analyse it more for you?

I'll also give you some further tips by PM of what to look at/for

CrackHeadMel

iTrader: (2)

Thinking outside the box

are you sure your getting knock

What about other engine bay noise's that could be fooling the knocksensor into thinking its hearing knock, when its really just something ratteling, or somethings thats working its way loose

RICE RACING

Second point, No its not, I run on some cars AFR's in high 9's very low 10's WITH high levels of water injection to totaly supress knock.

Never have I tuned a rotary or piston car where knock has increased with AF's in 10.0:1 and low levels of timing.... but in every case it has reduce the knock greatly. Its only when you go below 9.0:1 or richer that you might start to find increases in detonation, but most AFR meters wont even show this, and most ignition systems wont fire it at high revs/boost settings sooo its not very relevant and thus I have nto found anyone in reality who has EVER experienced more knock running too rich. Lots however do quote mystical sources as this being a problem, nice story but sadly it has not much to do with reality.

And knock is something to take seriously, the knock monitor on a PFC is excellent and many people have turned this feature off or come up with excuses as to why its showing what it does rather than addressing the problem, most cars that come to me fit this bill and after solving the problem by miricale the knock goes away

General PFC's readings I target 30 as an acceptable level, 60 is a warning where the engien check light will illuminate and stuff north of that is dangerous and you should try to lower it if possible within the bounds of what your trying to extract.

In this persons case there is no reason with no other sources of false knock readings why his levels should not be below 20's on a properly functioning knock sensor.

RICE RACING

If he is logging it and not hearing any abnormal noise like exhaust hitting due to bad mounts or feelign some other vibration you can say then he is ontop of it being combustion induced shocks. You realy need to log it on a the datalogit and use the chart funtion to analyse the EXACT points at WOT, boost pressure and rpm ranges to see what is happening.

I had a FD here a few days ago (12000km on clock!) where it was registering 120 on the hand set but untill you data logged it and did a chart analysis you could verify that the knock was NOT happening at the important 100% throttle sites (average was high 30's) after retune this droped to low 20's but still had odd spikes of 120+ in odd zones even when crusing in neutral etc. You realy need to look at the log file of full power runs, generaly a 1st gear to 4th gear pull is great to see what is going on.

kung stew

Thanks for the responses, everyone! I've been wrapped up in work for the past couple of days, so haven't driven the car since I last posted .

Barban, although my numbers aren't that high, they did reflect a change in what I was used to seeing which is why it got my attention. Admittedely, I wasn't logging 4th gear runs previously due to the speeds that requires.

I usually pick a gear (2nd, 3rd or 4th), start at ~3k rpms and run to redline at WOT, then save that as a logged run individually. Is this the accepted method?

RICE, I'll zip the logged runs that I have and send them to you, hopefully this weekend. Thanks!

-Stew

sk8world

iTrader: (5)

FWI.. On the way home tonight I was running 20 psi and knock was around 40-50at wot. But while cruising I was seeing even higher sometimes. Try and log just a 3rd gear pull. That way you can eliminate some factors..

RETed

If it is knock...

Not enough fuel
Too much fuel
Not enough ignition timing
Too much ignition timing
Not high enough octane
Spark plug heat range too hot
Ignition misfire
Engine internals too much deposits
Intake air temps too hot


-Ted

kung stew

Plenty of fuel- afrs mid-high 10s, has been this way for over a year w/o probs
Timing is conservative
Added toluene to rule out low octane/bad tank of gas
Have been using 9s all around for 12-13psi boost w/o issue for over a year
Using HKS twin power for over a year, no upper rpm breakup/misfires
Not sure how to check for deposits, but this is a fresh engine, built with new housings by Steve Kan.
Intake air temps have been between 30-35 on the PFC while logging.

I'd be happy to show screenshots of my timing maps if anyone could tell me how to post them. Reference my computer-retarded comment above .

ZAN_TUNING

iTrader: (1)

use the "Print Screen" button and save it as a jpg.

and i'm with RICE all the way on this one!

kung stew

Screenshots of IGL, IGTsplit.

FWIW, I changed the oil last night, went out briefly today on the way to work and hammered it in 3rd several times...peak knock was 19. It was a little warmer out today, but water temps and intake temps weren't significantly different from before. I did move the spark plug wires and a SS oil line away from the knock sensor, but that just seems like voodoo .

kung stew

F!
The ign split map doesn't show up well. Here it is again.

Barry Bordes

What are your P row values on Settings 3 ?
P15= 5 psi ?
p16=
P17=
P18=
P19=
P20= 19.7 or 24 psi?
Barry

kung stew

Boost in P17 is about 13ish psi.
P20 is I believe 19psi.

-Stew

kung stew

Screenshots disappeared b/c I'm an idiot.

kung stew

First zipped file containing logs with the knock events, for anyone who wishes to view them (RICE RACING?).

thanks,
Stew

kung stew

Next set of logged runs.

Barry Bordes

I hope you can make this out with such a poor quality screen shot.

Check your Datalogit run for high knock (on left screen). If you find a particular

high cell or cells, usually for a couple of runs all you do is take a degree out on the

right screen IGL and IGT. Then do more runs to verify.

Notice on the lower chart where Knock is in yellow. This helps spotting problem

peaks while logging.

Barry

Note : Try 125% on the thumbnail.

kung stew

Do you usually log min basic knock? I've always used max values.

-S

Barry Bordes

I normally log on average to identify hot cells. Then drag the curser line along the chart watching the map watch box. I shouldn't have had it on min for the example.
Barry