Knock Modules - LT1

[Kurek, Larry]

I think I finally solved my false knock problem...and I didn't have to
hack into the wiring harness either. Here is what I did.

I have a 94 Z28 which had a major false knock problem between 4000 and
5500 rpm. In fact, the PCM was pulling out 8-12 degrees of advance
almost constantly in this range. On the dyno, I had a nice smooth power
curve up to 4000 rpm, then it started getting wavy and the rate of
increase dropped...costing me about 15-20 RWHP from my estimates. At the
track, I had knock retard almost right from the starting line throughout
the run...12 degrees tapering to 8 degrees, then 12 degrees again after
the shift....even with the 100 octane fuel. Obviously, I was pissed and
was losing power.

So, I figured there must be a better way to "tune" the sensor, rather
than going through the sensor desensitize circuit. The knock chips in
the PCM are tuned to a specific frequency and gain by using external
resistors between pins on the chip itself. While the desense circuit
does work, it doesn't really solve the problem, since the signal will
still be sent to the knock chip, but at a lower level. Since the chip
references from the base level of the sensor, you will not get a full
solution. Since I've added a Y-pipe, !cat, roller rockers, intake,
etc...I'm sure the knock signature of the engine, along with the base
harmonics themselves had changed over stock...so that the stock module
was interpreting my mods as knocks, although this wasn't true. To solve
this problem, and to figure out how to "retune" the knock module, I went
and bought two new knock modules from GM to tinker with. The first thing
I noticed is that GM had a new part number for the stock LT1 knock
module than the one that was installed in my PCM. So, I put that one in
first...just to see the difference. Wow! Huge difference. While I still
had knock retard...mostly at the shift points, it was only in for about
2-4 data packets, and topped out at 10 degrees! So, it seemed I found
something...although I cannot quantify what changes were made between
the two modules I tested. Second, knowing that GM produced an LT1 with
roller rockers, and noisier cam drive (i.e. the LT4), I figured I would
give that a try...since I am sure GM knows what they are doing when they
tune these things. Again, the LT4 module was a completely different part
number. Well, I put this module in and guess what...NO knock retard at
all. Actually, when the motor is cold in open loop, I get a peak here
and there of about 8 degrees retard, but it only lasts for ONE data
packet. Hop Damn! The false knock is gone! As a nice byproduct, the car
now pulls hard all of the way to redline. I used to feel the flattening
of the knock retard around 4000rpm, but it is gone. What exactly is
different between the knock modules...I have not researched yet.
However, I took one of them apart and found there is a surface mount
circuit board in there, with a lot of surface mount resistors, a diode,
and two IC's with NO description on them. One is an 8 pin, the other is
a 12 pin. They are of the surface mount kind where all you see is the
traces going in, and a blob of brown epoxy covering the IC itself.

Anybody have a schematic on a GM knock module, and what exactly does
what? I'd like to know for the future....

FWIW, the part numbers on the two new modules are:

LT1	16214671
LT4	16214681

Either are around $27 from Dal...

TTYL!

Larry