Hi all, I have a 2002 corvette with Gen3 engine wiring and PCM etc but with a Gen4 LS3 engine.
I currently have LS6 sensors on with the wiring adapters/extensions and as most of you guys know I had to retap the block to accommodate the larger LS6 KS.

Which knock sensor gives the best/most accurate knock readings on HP Tuners? I hear KS can be sensitive. THX Brian!

 

With a Gen3 computer you really need to use the Gen3 knock sensors. They won't be very accurate when they are relocated, but they are better then not working at all with the Gen4 sensors.

 

I'm in the same boat. 2002 Corvette, even. It's embarrassing that I still haven't reached any conclusions about this.

What I have now is flat-response sensors hooked up to a Plex Knock Monitor, which works pretty well, but at some point I need to get off my *** and do some wiring so that the PCM will pull timing if there's knock.

Someone makes an adapter harness that adapts "flat-response" (aka Gen 4) knock sensors to the Gen 3 main harness, and I've heard that it works fairly well with no changes to the tune. I haven't tested that myself yet though.

The P59 PCMs have inputs for flat-response knock sensors, which in theory should work after some minor changes to the tune. I have not tested that yet either.

 

The P01/P59 computers have several tables that are not in the common editors like HP Tuners, EFILive, etc but you would have to change the frequency/filter/initial voltage thresholds as well as build your own sensitivity tables if you wanted to try to use the flat-response 2 wire sensors with that PCM. Just use the commonly available XDFs to track them down and compare against other calibrations that came with them. Many have tried, heck I've tracked down the exact frequency and filter tables for the flat response sensors in other calibrations with the same bore diameter and tried that but it still just didn't work right. That led me to believe that even though it's selectable in the tune with a switch and even with the proper frequency/filter/voltages the code just isn't there to work properly with them.

Basically that adapter harness that "converts" it to use the flat response sensors has never worked for me, even with the proper changes in both the P01 and P59s.

 

I was sent a bin file from a V6 that used flat-response sensors with the 7603 operating system and, and the guy who made (what I think is) the best 7603 XDF defined the tables for tuning them. I made an adjustment to the fundamental frequency based on an online calculator and in theory that oughtta work. In theory.

When I had my motor swapped, I have the installer detailed written instructions about how to wire the sensors, and one of those converter harnesses just for the connectors and he installed the sensors according to their instructions, not mine. That was the one miss in a fairly complicated motor swap (dry sump, etc) so I can't complain too much.

For the tuning (and ongoing monitoring/logging the Plex Knock Monitor has dedicated knock sensors stacked atop the flat-response sensors that are in the gen-4 locations. The calibration process worked as they described. It's only logged one knock event so far, but I've just been tuning for drivability so I can be confident that it's running right before I take it to a tuner for a dyno tuning session.

Option A: sensors wired with one signal wire each going to one of the two flat-response inputs and the other grounded at the block. This seemed like a less-bad idea than wiring it as per the instructions that came with the harness, but I suspect it's actually the worst thing possible. This approach did not work for me. The PCM didn't register the knock event that the Plex captured.

Option B: flat-response sensors on the sides of the block (gen-4 locations), wired to the resonant input pins, and a tune with the same knock settings as the 2004 Corvette. That totally doesn't seem like it should work. I assumed that the conversion harness was snake oil and just bought it for the connectors... but like I said, some guy on Facebook said it worked for him. I just need to move the signal lines back to the resonant sensor inputs on the PCM. Edit: this did not work for me either.

Option C: flat-response sensors with each sensor wired to both of the flat-response pins on the PCM, tuned as per the V6 tune that I mentioned earlier. Hopefully use the Plex to guide changes to the tables in the PCM. This is what I originally had in mind, but will require running some wires. It's kind of discouraging to hear that this didn't work for you though. Edit: this approach worked for me!

If I can't get any of the above to work, I'll still have the Plex and will be logging with it. I would like to have the PCM able to pull timing as needed though.

What sensors are wired to your PCM now, and which inputs are you using?

 

In the Gen3 vehicles that I've tuned over the years the only way the sensors kind of worked was by moving the Gen3 sensors to the side of the Gen4 block and just leaving them wired into the original inputs. Some false knock from time to time and not registering real knock all the time no matter what you do to the sensitivity and initial voltage tables. Not the best solution at all, but the only one that kind of works with what we have to work with. Tried on several occasions to wire the flat response Gen4 sensors to the appropriate pins and changing frequency to the 2nd harmonic (these are wideband sensors and hear everything unlike the Gen3 resonant sensors) as well as changing the filters, sensitivity, etc to match the other calibrations that came with those sensors and nothing really worked. Basically it was a lot of time invested to essentially figure out that something else must be missing in the code to truly make them operational in that fashion.

 

LS3 knock sensors will work just fine with the 0411 PCM. Using hptuners, copy and paste all tables related to knock sensors from a 2002 Chevy S-10 file into your file. The '02 S-10 used an 0411 PCM with flat response sensors. As long as you have the adapter harness or have a custom harness with the correct connectors out of the box, you're in business. I'm running a LS3 with an 0411 PCM and the sensors are working perfectly.

 

Can you post a link to an adapter harness that's known to work?

The only one I'm aware of uses 1 wire per sensor, with the sensor's other wire grounded at the block, and I'm wondering if that's actually viable.

Thanks!

 

I don't know of any specific adapter harness, sorry. When I did my swap I used a BP Automotive harness with LS3 knock sensor connectors already integrated into the harness.

 

Seeing as how I've tried this, and there are quite a few more tables required than what are available in HP Tuners, they are definitely not working just fine. Here are just a few of the ones that I had to define myself (not available in the editor) way back when I was trying to make the newer sensors work on the Gen3s, and they most definitely need to be changed to have them be accurate at all. BTW, the correct frequency isn't even selectable for flat response sensors on the 2nd harmonic which is where you want them to be. 1st harmonic on flat response is NOISEY. **edit** I lied, some of the cals have the correct frequency you can select. It's been a while since I've looked but it still doesn't work...

 

Ive only experienced knock retard once since doing the LS3 swap and I was actually stoked to see it and get validation that the system was working. There were a few days in September where I was driving in 108° heat, running California 91 octane. Under high load, low rpm conditions (highway driving at 60mph) the scanner would consistently show anywhere from 0.5 to 5 degrees of knock retard. Haven't seen it before or since. So in that scenario where you would absolutely expect to see predetonation, my engine was protected and the sensors were working just fine.

 

Not trying to be a wise guy here, but I see this in print more than I should, with "this" being predetonation. No such terminology when it comes to engines. Maybe when setting off dynamite, but doesn't apply to engines. Apparently, people mix pre-ignition and detonation, and come up with predetonation. Detonation is detonation. Pre-ignition is pre-ignition. Just trying for accurate terminology here......

 

Got my car dyno'd today, with flat-response sensors hooked up to both the PCM and a Plex Knock Monitor v3, and when the Plex saw knock, so did the PCM. I posted a P59 .bin file (operating system 12587603) and a few notes here:

https://pcmhacking.net/forums/viewto...113096#p113096

It's pretty much what AdsoYo described above: GM made a couple vehicles that used 411 PCMs and flat response knock sensors, and if you copy the right tables from those vehicles, and modify them a bit to suit your own, it just works. The vehicles with flat-response sensors were V6s, but they used the same operating system as gen-3 V8s, so it's pretty straightforward.

Edit: HP Tuners might not provide access to all of the needed tables. I'm using Tuner Pro with the 7603 XDF from PcmHacking.net, and reading/writing with PCM Hammer.

 

Interesting results. This has been the cause of years of conflicting speculation with no real answers over on the HPT boards as well.

 

Yeah, I wasn't sure what to expect. Really happy that it worked out, though.

It was helpful to have another method to verify knock, because at first the PCM was pulling timing due to background noise at high RPM. Plex's Knock Monitor was nice, but something like Haltech's Knock Ears would probably suffice if you know what knock sounds like. And my tuner did (Turbogixxer in Washington state) but I spent a fair mount of time troubleshooting other issues before I got to the dyno and I wanted all the help I could get.

 

Funny that the OP (Rockitansky) never came back here