pcmhacking.net

So if I read correctly the gen3 are resonant, but gen4 are piezo. Piezo make their own voltage, so could scope the signal to see when it's showing something. Can even just tap on block while scoped to see what it does. Not sure computer can do anything with signal, but that's what you're trying to solve.

aaronc7 wrote:

jlvaldez wrote:HPT has nothing to log a live knock level thing. You just get to record KR.

Its one of the things I miss about my old Honda I tuned... It had a live "knock level" PID that was read at 60 hz and was a 0-100% value. You could see knocks as large spikes. You would also see the relative "volume" get "louder" with load (As you'd expect).

I've found no such functionality in the tech 2 nor HPT. Haven't played with EFI Live, so I don't honestly know.

Never saw anything like that (other than standard knock retard) on EFI.

Didn't think so. I'd have been surprised.

and LOL looks like we're all on the same forums... racechrono, Corvette forum, here.

Cincinnatus wrote:o if I read correctly the gen3 are resonant, but gen4 are piezo. Piezo make their own voltage, so could scope the signal to see when it's showing something. Can even just tap on block while scoped to see what it does. Not sure computer can do anything with signal, but that's what you're trying to solve.

The issue we're tryign to solve is a bit different. We want to see that the ECU is seeing activity on the knock sensors.
Supposedly the P01/411/p59 can all handle the flat response sesnors, but some pins must be moved around. Issue is without any sort of ECU live reading (or even historical), it's tough to determine if the ECU sees the input and other things like gain are set right.

I get it. If you can get KR by tapping near sensor, then PCM is working from the signal. Just saying the scope will assure that there is a signal to process.

I think the largest issue is mapping all of the parameters for the knock sensors in general. It's likely that the few of the parameters and tables need to be populated differently. I have a tune file from a S10 or Blazer from memory that uses a single flat response sensor. I have most of the tables mapped out to try just have never gotten around to messing with it anymore to finish mapping parameters and physically testing.

ShorTuning wrote:I think the largest issue is mapping all of the parameters for the knock sensors in general. It's likely that the few of the parameters and tables need to be populated differently. I have a tune file from a S10 or Blazer from memory that uses a single flat response sensor. I have most of the tables mapped out to try just have never gotten around to messing with it anymore to finish mapping parameters and physically testing.

What format is it in? Binary? HPT? EFILive? HPT's tune repository is kinda janky after they redid their website. Seems like some of it is back now but it's not as populated as it was.

Hptuners removed any non stock tunes to avoid EPA fines for emissions equipment violations.

jlvaldez wrote:

ShorTuning wrote:I think the largest issue is mapping all of the parameters for the knock sensors in general. It's likely that the few of the parameters and tables need to be populated differently. I have a tune file from a S10 or Blazer from memory that uses a single flat response sensor. I have most of the tables mapped out to try just have never gotten around to messing with it anymore to finish mapping parameters and physically testing.

What format is it in? Binary? HPT? EFILive? HPT's tune repository is kinda janky after they redid their website. Seems like some of it is back now but it's not as populated as it was.

HPT but can be converted to BIN by flashing to a PCM and ripping it back out in BIN. It's a 2002 S10 V6. Uses one flat response sensor. Here is the HPT file.

here's a bin.

Curious to see how this turns out

I just did a bunch more Googling and Binging on this, with no new findings. There's mostly just a ton of threads that end with something like "no, the P01 hardware is not compatible with LS2-LS7 knock sensors" which seems entirely wrong in light of the checkbox and the pinout from EFI Live. There's a couple that say to just rewire and tune for them, which is promising but completely vague.

The comments from Mike454 in this thread give me hope:
https://ls1tech.com/forums/pcm-diagnost ... 1-pcm.html

And the replies from CalEditor@PcmCalibrators, here:
https://ls1tech.com/forums/pcm-diagnost ... puter.html
I'm pretty sure that's AJ Whiteley, who invited me to call him a couple weeks ago after we got to talking in the comments on a Facebook post... I need to actually do that.

There's a thread on Holley's site about how to tune knock sensors (using their ECU of course) and when you look at it from that perspective... any tuner that has successfully set up knock sensors for a couple of standalone installations would probably look at knock sensor tuning as just another step in the tuning process. Those of us who are used to tuning factory ECUs/PCMs have just been fortunate to be able to skip this step in most cases.

But anyway...

Thanks for V6 bin file. A quick summary of relevant differences:

B6226 - The S-10 V6 file has a higher knock sensor frequency (10.8khz) than my stock C5 tune (5.5khz) but that's almost exactly a 2:1 ratio so maybe the bore sizes are similar and the V6 tune is just looking for the first harmonic of the knock sound.

B6227 - The knock sensor filter table for the V6 is all 1s, where as in the C5 tune it's 2.

B6228 - Knock sensor sensitivity - This table only exists in the V6 tune.

B6240-B6242 - Learned knock sensor theshold, initial/min/max are different.

B6244 - Knock sensor multiplier A - these tables 'slope' in different directions, which is kind of interesting. The V6 tables are 8 at idle and increase to 10 or 12 by 3000-4000 RPM, whereas the LS tables are 11ish at idle and 9 at high RPM.

B6245 - Knock sensor multiplier B - pretty similar in both

B6248 - Knock sensor background noise filter - 1 for the V6, 0.1 for the LS

B6250 - Knock sensor multiplier TPS - For the LS it's 15 in all cells from idle to 3200, 13 from 3400-8000, whereas the V6 has 24 in all cells.

B6255 - Knock sensor multiplier RPM - 24 in all cells for the V6, 17 in all cells for the LS.

The thresholds for TPS knock mode and RPM knock mode are also different but I'm guessing those are based on how the engine behaves rather than how the sensors behave.

NSFW, are the differences you posted from the p01 bins or p59 bins? Curious, so the OS that uses them is 3x for the v6, not the ls 24x OS, or is the OS the same with different segments? If it's the same I would think it could work. If different OS then maybe tables are utilized differently and can't work. Someone I'm sure knows from disassembly of some OS's for xdf or tuning.