How can I disconnect the knock sensor/retard system without reprogramming? Can I just unplug the knock sensor, or will that make it retard the timing? I am running a Zex150 shot and get a slight hesitation/flat spot/stumble in the acceleration in the middle of second gear. Doesn't happen N/A or with a 75 shot, but occurs with the 100, 125 or 150 shot. I think it is picking up false knock. I'd like to bypass the retard system and see if it changes. I'm running 4* timing and a pcmforless tune, NGK TR6's and 93 octane so I am sure it isn't real knock. It is either that or the fuel pump isn't keeping up. I'd like to try this for a run, then if it still stumbles I'll buy a new fuel pump since it is on my to do list anyway.

 

It can be done on some systems. Give me a hint what year car you're running.

 

Sorry, didn't realize my sig doesn't show up. It's an '89 L98 automatic convertible. I have Hooker headers, true dual exhaust, 1.6 RR's, B+M Shift Kit, Zex 150 shot with 15lb bottle and all the accessories, and a pcmforless tune. Many other small mods, but these are the major ones.

 

 

There is a knock sensor code, I do not know if having the code would make the car slower though.

Also there is an upgraded ECS module that might help.

 

Just an addition the upgraded ecs module per Gordon Kilibrew is pn# GM# 16038331 AC DELCO# 216-35. Its original application was 1986 caprice with 4.3 engine. I run one on my 1989 vette.

 

So if I just unplug mine it will pull timing out? I was hoping to disable the knock system completely to see if that was my problem or not. Then if it is I'll look into an upgraded ecs.

 

Looking at the part number I see you've got one of the early type ESC units (called the Half Function module) that mounts in the engine compartment. And, looking at the specs of the module, you already have one of, if not the, least noise sensitive modules made for that particular electrical design (identical to the stock calibration in my '84). The early ESC modules (and ECMs) will run fine with the knock sensor disconnected, as there is no electrical diagnostic circuitry in the module. IIRC, that diagnostic ability came on board when the ESC electronics were integrated onto the ECM memcal. The only monkey wrench I see would be if your software package had that (poor) attempt to exercise and test the ESC system by occasionally cranking in additional spark advance, and then looking to see if the ESC module signalled for the ECM to retard the spark. This system worked only modestly well, as differences in octane and ambient humidity would occasionally prevent intentional knock during this test. This caused several cases of false alarms regarding ESC system integrity. If you don't have this software package, then you can just disconnect the sensor, with no code issues. If you do have this software system, then it may try to exercise the ESC, and without an active sensor, it would then automatically set a malfunction code. I'm a hardware guy. You need to talk to a s/w guru who has a s/w listing that can confirm what you have in the box (ECM). In the meantime, you could just disconnect the sensor and see if you get a code while driving around. Just remember to not do anything stupid while you are not protected by the ESC system.
Keep us updated.
p.s.: A couple months ago I wrote a long winded set of posts describing the workings of ESC systems over the years. Perhaps you might be able to track that down if you are interested in more details about the ESC system. Just out of curiosity, what was the part number of your original module?

 

Just out of curiosity, what was the part number of your original module?[/QUOTE]

Hi the gm number on the factory installed esc module is 16052401 on the 89. ps good info by the way

 

This tech tip mentions that he would have the type that does the excersise.


Tech Tips
Search ::

Tech Tips >> C4 Related Tech Tips >> Upgrade your
'85-'89 ECS module.
Tech Tip

Upgrade your '85-'89 ECS module. -- 1 of 1
Date Published: 2001-09-09

Submitter's Name: James Kovitch
Email Address: jgkov@msn.com
Corvetteforum.com Member Alias: 65Z01

Background The '85-'89 L98 uses an Electronic
Spark Control (ECS) module to listen to the Knock
Sensor and report ping to the ECM (on '90-'96
Corvette motors the ECM listenes directly to the
Knock Sensor). When the ECM detects knock counts
it pulls timing till no knock is reported and then
gradually increases timing to the current
setpoint. Problem Even with base timing at +6deg
and running the stock MemCal chip, my '88 L98 was
plagued with false knock counts under WOT
operation (seen on my AutoXray scan tool). These
were causing timing to be pulled, which I felt was
hurting 1/4mi performance. Solution When I
discussed this issue with Gordon Kilebrew, at
Corvettes at Carlisle 2001, he told me to replace
my stock ECS module with the upgraded unit: GM P/N
16038331. Verification After doing this upgrade a
short WOT run on the highway showed no knock
counts on my AutoXray. Subsequent testing at the
local drag strip showed an improvement in trap
speed of about 1 mph. No other changes had been
made and trap speed data was normalized for
weather conditions to provide a valid comparison
with prior time slip data. Equipment Safety This
upgrade is safe for the motor since the ECM checks
the ECS module occasionally by advancing timing
till a knock count is registered; if no knock, the
ECM would record a code 43. Just to be sure all
was well, I advanced my base timing to about
+12deg and made a WOT run on the highway; 7 knock
counts had registered on my AutoXray.

 

So, if I simply unplug my knock sensor, what will happen? Will it retard the timing because no knock sensor is present? Or will it not retard any timing? Or will it run a limp mode? I don't care if it throws a code, just so it doesn't retard timing while the code is being thrown. I just want to bypass the esc system temporarily to see if my problem goes away or not. Then I'll hook it back up if the problem stays the same.

 

Thank you for this advice. I hope I don't have that software system.

 

I have a toggle switch on my knock sensor to shut if off for testing purposes. If you unplug it at the sensor a code will appear.
You must put a switch on the ground lead of the knock sensor module.
No codes will appear and there wont be any timing retard. If you manipulate the switch while the car is running it will throw a code. It must be manipulated with the engine off/key off.
Trust me it works. I have been using this setup for several years without a glitch.
I can provide wiring color and photos if you like.
John

 

Hi the gm number on the factory installed esc module is 16052401 on the 89. ps good info by the way[/QUOTE]

Well, guys, I'll admit a bit of confusion here. Looking up the electrical specs on the 16052401 (the stock module on my '84 also), and the 16038331 unit, they are identical. Same center frequency, amplification/gain, and noise spike ignore times. I'll have to assume that some internal components were changed, requiring a part number change. I'll send an e-mail to a former co-worker for some clarification.
As I mentioned earlier, these modules are not terribly sensitive. (There are bunches more sensitive, if someone was anxious to go that way.) I did find one that is slightly less sensitive on the bottom end, but probably not much difference on the top end (noise and rpm wise). This part number is 16131171/16131231 (code APLU/APLW), found on big block trucks. The lesser low rpm sensitivity, however, might make the advance/knock diagnostic problem worse though.
Sorry, but it keeps coming back to possibly having to disable the software ESC diagnostics. We'll keep working.

 

John,

I don't diagree that you are having success with your system, but I'll have to admit I can think of no way electrically that the ECM won't set a code, assuming you also have the software ESC diagnostics in your ECM. Grounding the sensor will imitate a non-knocking engine, but the ECM will look for a retard request from the ESC module if it temporarily advances the timing trying to incite knock. You've got me baffled.

 

 

I got a pm last night from a forum member, who I will leave anonymous unless they want me to say who it was, with something to try. I picked up another knock sensor, unplugged the one that is in my block, plugged in the new one, wrapped it in a rag and duct tape, and wire tied it to the top of the starter. This effectively made the knock sensor deaf. It worked. I ran the 150 shot with no hesitation at all, no pinging, nothing but huge horsepower. This idea worked perfectly for the test. Now I know false knock causing timing retard was my problem. I'm not sure yet if the upgraded knock module will be the permanent cure or not, I need more info before I spend $80 on something that won't work. I ran for about an hour and have no codes as of yet. I wish I knew how often the system advances timing and checks for the knock sensor, I would be fine with just clearing codes once a month. I also worry that it will advance timing in order to check for knock while I am blasting the 150 shot in there. Just what I need, a computer that thinks it is smart, advancing timing when timing needs to be retarded. I guess I have a retarded unretarder of a pcm.

 

Did you use just a one wire in, one wire out toggle switch just to disconnect the wire when the switch is off? Or is it a relay switch with power and ground? I assume the switch is on the wire coming from the knock sensor to the esc module. It is blue, I think, on my car. I could see putting the switch right beside the esc module easy enough. If it is just a disconnect switch, it would be the same as cutting the wire when the switch is off, which would be the same as unplugging the knock sensor completely. I did start my car with the knock sensor unplugged completely and didn't get a code, but I put the "deaf" sensor on before I drove the car.

 

Good idea! Glad you got the problem figured out. That is half the battle.

 

Yes, thanks for your help also. Now I know my problem isn't spark, fuel delivery, nitrous cut, or one of millions of other things. Diagnostics is definitely the hardest part. Now I'll wait to see what can be done so that I can reactivate the knock retard system. If nothing else I might play around with teflon tape and see if I can "desensitize" the sensor installed in the motor so that it works, but not pick up false knock. I have a header that bumps the heat shield once in a while, and there is no room to move it either way because of the floor on one side and the automatic shift linkage on the other. I barely hear it but obviously the knock sensor did. I also have Summit roller rockers that are noisy, along with other things that the sensor must pick up. I'll follow along with this thread and hopefully figure out exactly what to do for a permanent fix as I hate to run forever with a disabled esc system.