This section is not so active but thought I'd post here as well as ls1tech.

I’m hoping a few of you that have much more LS1 experience could give some advice. I have a basically stock 2001 C5 LS1 (K&N, trunions, ls6 pcv) with some minor tune massaging. My concern is knock.

I have false knock, so I added 50% e85 and dialed back the Global Gain. Love this tune and no worries, BUT I want to run pump when traveling to distant road tracks.

I’m in AZ and the 91oct here sucks *****… at least with my BMW turbo experience. So I’m expecting at least some knock even with good tune and cold temps.

My gain in the upper MAP is 1.75 compared with 1 stock. At 1.5 I get false knock. So I’m a little concerned that I have basically disabled my knock sensors.

Attached is my timing histogram with <50deg ambient and 91oct today. Common to never see knock with a good tune?
The scan is 186kb, so a lot of driving and at least 5-6 WOTs.
And I’ve disabled burst knock, so I’d think I should see some KR during transitions.

I’m guessing the knock threshold is basically = (Gain) x (Knock sensor level)
I don’t see how you’d ever gen knock at <50 MAP… the gain is so high.

Any thoughts would be great!

 

Well if you have real knock you can adjust the AFR to a richer a setting or add ethanol---
However to correct FALSE knock you are attacking it the wrong way--
With false knock you should never hear pinging or have to mess with the WOT fuel ratio
As an engine ages it becomes louder mechanically- those new noises are interpreted as "knock" by your knock sensors--
The only way to correct "false knock" is to "Desensitize" your knock tables in your tune so the sensors don't pick up those new noises
Also--make sure timing is NOT being removed 1st by IAT deducts or ECT deducts---take all those into consideration when comparing your actual timing to your commanded timing--
Example: say you are commanding 25*
You have an IAT temp of 130* that removes 4*
You have an ECT temp of 210 that removes 1*
so take 25 minus 5(deducts) and your actual should be20*
IF you have less than that it is indeed false knock
Most of the time the KR data logging will show all KR but you won't know exactly how much is false or how much is real--

 

thanks for the response, but I think you misunderstood my OP. To determine false knock I did add ethanol and re-tuned accordingly. I then started dialing back the knock sensor gain which ended up being 1.75 compared to 1.00 stock in the higher MAPs. My concern is that this new gain has basically disabled knock detection... because on pure 91oct I now have NO KR. But in my experiences on pump I would expect at least some KR during transitions at least.

Hopefully I'm just inexperienced with ls1 and NO knock is possible on 91. When it starts getting hotter I'll find out. I don't hear pinging, but I do not have much experience in audible detonation and 100k mi c5 is not a quiet car.

BTW: KR only registers that is perceived as knock. Any deducts for IAT, ECT will not show as KR. Maybe I didn't catch your meaning though.

 

I would not recommend going back and forth between essentially 10% ethanol and 45+% ethanol. Until your fuel trims adjust you are going to be running a bit lean when switching to the higher ethanol blend, and then will be richer right after switching back to the lower ethanol blend.

If you want to determine false knock, take 3 degrees or so of timing out where you think it might be occurring and see if you still get it.

At a gain of 1.75 on your knock sensors you are desensitizing them some, but you aren't disabling them. Get the gain settings just above where they need to be to eliminate false knock and then add the timing back. Unless you are adding a great deal of boost a little pinging isn't going to blow your engine although you do want to eliminate it. I've adjusted the gain on my blown LS-based truck engine to eliminate the false knock from its cam and will still get a little pinging now and then (got some this morning) as I'm dialing in the tune for colder weather after installing meth injection in June. Sometimes when I get pinging there is no knock retard, other times there is. I haven't yet looked at my scan data from this morning's drive to work so I'm not sure whether or not this morning I got any knock retard, but I do know there was less above 5000 rpm than there was from 4000-5000. May or may not be explained based upon whether knock retard kicked in. Unfortunately it's a fact of life that noisier engines will have less effective knock sensors...

 

Hey, thanks for the feedback. I have 2 tunes dialed in... 1 for e45 and pump.

I've only really seen false knock, so I have no comparison. The 1.75 is right at the threshold. Can you post your gain table and KR histo? Just curious.

Kinda sucks cause I have to keep the pump timing conservative until I get confidence in my knock detection. Replaced sensors and harness with no improvement. Along with idler/tensioners. Spring time with hotter temps I should have a better idea of effectiveness at this gain. I guess if stock 4.0 gain can result in true knock detection then 1.75 should be fine... I'm curious if anyone gets KR in the lower MAPs. I've never seen it.

For your case, could prob estimate at what MAP you're getting audible ping and bet it's in the higher gain thus not registering as KR... although I'm not sure how this scales with FI. Theoretically the gain should be lower (or same) at reduced MAP... It just changes the amplitude threshold. Maybe GM does this so knock learn isn't so sensitive.

 

I can post the knock gain and a kr histogram but it probably won't be until tomorrow night at the earliest. My tuning laptop isn't hooked up to the internet so getting those on here is a minor PITA.

Big reason for my pinging today is I had just changed the tune (plus it was about 10 degrees cooler today than it had been for most of my previous post-meth injection tunes). I had noticed that before doing the meth injection I had scaled back the Power Enrichment spark advance at 100 kPa of MAP to 50% of what it was at 80 kPa - I'd forgotten that I'd done that who knows how long ago. I'd been getting some pinging now at part throttle (with the meth kicking in around 2 pounds of boost) and I figured the above tuning setting was the reason. Pulled a bit of timing where I was hearing the pinging and pushed the 100 kPa PE spark advance up to 75% which resulted in WOT pinging. Looks like I need to decrease my spark advance table by the amount that the PE spark advance is going to increase - I may want to push the 100 kPa to more than 100% of the 80 kPa.

There may be no point to sharing a kr histogram if I don't have any kr or if it just kicked in at one point. I "usually" don't have any kr. I've also found that having burst knock retard seems to eliminate some pinging due to either retarding the spark some in the delay before the meth sees the cylinders or due to the above mentioned PE spark advance setting.

 

Well, for some reason my data log from yesterday morning was absent from my HPTuners pro interface when I went to pull it off. However I did take a look at my most recent data log and can comment on that. I don't think there's much point in showing a kr histogram because I can describe what happened pretty easily as there were only 3 incidences of kr.

The log spanned 42 minutes of driving. All 3 kr events occurred between 4000-4500 rpm, around 80% throttle, and most likely in boost. (1-bar MAP sensor maxxed out.) The last kr event occurred before the previous event's retard had fully subsided. In my knock sensor's gain table I've got the two applicable cells set at a gain of 1.3.

The biggest changes in my sensor gain table are in the 2800 rpm column from 80-95 kPa where I've got the gain set at 1.7. It's set there because I was seeing kr at 1.6 even though I was running higher octane than the regular called for by GM for the 5.3 with less than stock spark advance. Seems my aftermarket cam must have some harmonics around 2800 rpm.

I've got a lot more experience tuning that truck's 5.3 than the vette's LS1, but it seems as though as stock as yours is having to go as high as 1.75 seems awfully high. Maybe it has something to do with an aluminum LS1 block versus the truck's iron block???

 

Thanks. There's not many LSx logs that I can find... it helps in correlating to your own results.

Yeah my 1.75 gain seems high to me also. I'm not going to worry until it gets hotter.

 

I was trying to help you understand how IAT and ECT deducts in timing works---It seems that you have already figured that part of tuning out---Most people don't understand how that works----I agree with the previous poster--If you keep changing fuel compositions you will never learn any history--I would 1st try to get the KR figured out with 91 or 93 oct fuel--- before messing with ethanol infused fuels
Basically---If you don't hear any detonation it's false knock---IF you desensitize the knock tables and you GET some pinging---It's REAL knock and you just have too much timing----I have tuned even LS6 engines that pinged at 19*---that is real knock-These LS engines do not require tons of timing to make max HP----
YES you can force the timing in and desensitize the sensors---But you really aren't making more HP--just shortening the life of your engine--

 

OP, is this a mod or does the '01 ls1 c5 come that way? The pcv setup on my '97 c5 doesn't look anything like the setup on my ls1 truck(which I'm considering switching to the ls6 pcv setup).

Thanks.

 

Mine was like your 97 c5. The ls6 runs from the valley cover to the manifold and you block the drivers side port. Cleaner engine bay and may pull more vac along with better oil separator.

There's DIY on ls1howto.

 

I'll check it out.

I take it you did your own swap; time and cost involved?

Thanks