Is it really knock?

[Kitch]

Along with all the other things I've been trying to get sorted with my van, spark curve was one of them.
I have a standard 13mile loop which is mostly motorway that I drive to make comparative logs on, most of it's driven at 60 Mph and when I get half way I accelerate out of a large roundabout on the return journey home, this gives me my W.O.T numbers.
Trying to tune with the knock sensor has been extremely frustrating and I ended up with a very weak looking spark curve to try and get a run with no knock counts (yet to achieve no knock counts). The van drives quite nicely around town in the 30-40MPH range driving along with 15-18 inches of vaccume on the gauge and no knock counts, once I get out onto the open road and into overdrive knock counts just take off.

I have change from NZ 95 octane to our 98 octane fuel to try and eliminate low fuel octane as the problem.

The last few last drives have been with the engine cover off so I could keep an eye on the fuel pressure gauge (lowest it went to was 15.0PSI) and with the engine is right beside me as I drive and I can't really say I'm hearing lots of knock but it's quite loud in general so I might be missing it.
The runs with my low knock spark table might generate 200-400 knock counts while I'm out on the motorway, virtually no counts while I'm cruising at 60MPH but sometimes I'd get some when trying to accelerate up an over bridge and always when accelerating hard out from the roundabout and sometime to annoy me it just throws a few as I lift of the throttle say after slowing down from an acceleration run and the spark is increasing as the load comes off.
Last night I copied a spark decent looking table from a post here and pasted it into my current BIN, my 40-60MPH accelerations times dropped by .5 of a second but knock counts while I'm on the motorway and accelerating have jumped up from 200-400 to 4500!
The spark table I'm currently running looks quite normal for a small block to me and I'm running 98 octane fuel, Wide band O2 (which I haven't connect to the Autoprom yet!) shows 12.0-12.2 at W.O.T., is this really knock that I'm seeing?
I've added images of my current spark table, my low knock spark table and my stock spark table. The XDL of my last run with the new spark table is there too if anyone would like to take a look.

[Fast355]

Along with all the other things I've been trying to get sorted with my van, spark curve was one of them.
I have a standard 13mile loop which is mostly motorway that I drive to make comparative logs on, most of it's driven at 60 Mph and when I get half way I accelerate out of a large roundabout on the return journey home, this gives me my W.O.T numbers.
Trying to tune with the knock sensor has been extremely frustrating and I ended up with a very weak looking spark curve to try and get a run with no knock counts (yet to achieve no knock counts). The van drives quite nicely around town in the 30-40MPH range driving along with 15-18 inches of vaccume on the gauge and no knock counts, once I get out onto the open road and into overdrive knock counts just take off.

I have change from NZ 95 octane to our 98 octane fuel to try and eliminate low fuel octane as the problem.

The last few last drives have been with the engine cover off so I could keep an eye on the fuel pressure gauge (lowest it went to was 15.0PSI) and with the engine is right beside me as I drive and I can't really say I'm hearing lots of knock but it's quite loud in general so I might be missing it.
The runs with my low knock spark table might generate 200-400 knock counts while I'm out on the motorway, virtually no counts while I'm cruising at 60MPH but sometimes I'd get some when trying to accelerate up an over bridge and always when accelerating hard out from the roundabout and sometime to annoy me it just throws a few as I lift of the throttle say after slowing down from an acceleration run and the spark is increasing as the load comes off.
Last night I copied a spark decent looking table from a post here and pasted it into my current BIN, my 40-60MPH accelerations times dropped by .5 of a second but knock counts while I'm on the motorway and accelerating have jumped up from 200-400 to 4500!
The spark table I'm currently running looks quite normal for a small block to me and I'm running 98 octane fuel, Wide band O2 (which I haven't connect to the Autoprom yet!) shows 12.0-12.2 at W.O.T., is this really knock that I'm seeing?
I've added images of my current spark table, my low knock spark table and my stock spark table. The XDL of my last run with the new spark table is there too if anyone would like to take a look.

I have been running the Express with the knock sensor disabled since the 4L60E days. It picks up knock even running the factory low octane spark tables. The knock sensor itself sometimes threw a code as well. The old Blackbox PCM had the LT4 knock module for years too which was rumored to disable knock retard on the 5.7 Vortecs as well.

[billygraves]

..............

[Fast355]

I would suggest you begin with installing the cold spark plugs. I have no idea what engine fmaily you have nor year. Second, I'd use GM TOP ENGINE cleaner or the Delco. Shake the can well. It works best on a HOT HOT engine. Pur in 2/3 can slow and flood the last 1/3 to stall the engine. (I use 2 cans) Cover the engine and block air best you can. Let it sit for 24 hrs. Start up and run to 1200 rpm and drive a few miles or KMs. Change oil and filter. We found @ GM the Top Engine clnr works as well as the Walnut blast carbon removal in side by side tests back in 88. To verify this, we used a medical bore scope looking at the back of the valves and internal combustion chambers. It works.
Lastly separate plug wires 5 and 7 then 4 and 8. It makes a diff. I've seen the spark plug shields set off causing knock. With the engine cover off, have a passenger use a copper tube 1/2 to 3/4 inch ID to listen. It helps with all the other mechanical noise.
I use CR43TS plugs. Not the R45TS. They have been known to cause spark knock. Champions are junk as Bosch misfire at part throttle. (the C is a commercial grade)

I usually run MR42xx plugs. They are a stainless body marine plug.

[billygraves]

..........

[1project2many]

When diagnosing spark knock I always started with top eng clnr.

TEC worked great! But it's gone from GM now. At least in the US. They sell something else, kind of green, but nowhere near as good. http://trailerpartstogo.com/images/products/detail/88861314.jpg

I found out that TEC was gone when I got caught by surprise. I had some engine noise I was trying to diag and ordered the green stuff when the dealer parts guy said it was just as good. Bzzzt! Sorry. I found a number for a Motorcraft version but it's also gone. Mopar has a spray can (4318001AD) that's supposed to be close but my dealer wanted me to order a case in order to find out. That would have been $150. About the only thing I've found in a pour can that seems close is Motorvac cleaning solution. $45/quart. Smells the same and will lift carbon if left overnight, just like TEC. Again, not sure if this would be available in NZ.

Since you have a regular drive with a good stretch of load, you could install water / methanol injection. Either additive would supress knock and remve carbon at the same time.

[Fast355]

TEC worked great! But it's gone from GM now. At least in the US. They sell something else, kind of green, but nowhere near as good. http://trailerpartstogo.com/images/products/detail/88861314.jpg

I found out that TEC was gone when I got caught by surprise. I had some engine noise I was trying to diag and ordered the green stuff when the dealer parts guy said it was just as good. Bzzzt! Sorry. I found a number for a Motorcraft version but it's also gone. Mopar has a spray can (4318001AD) that's supposed to be close but my dealer wanted me to order a case in order to find out. That would have been $150. About the only thing I've found in a pour can that seems close is Motorvac cleaning solution. $45/quart. Smells the same and will lift carbon if left overnight, just like TEC. Again, not sure if this would be available in NZ.

Since you have a regular drive with a good stretch of load, you could install water / methanol injection. Either additive would supress knock and remve carbon at the same time.

Seafoam works great as does running the engine on straight B12 through the injectors.

[Kitch]

Thanks for all the advice.
It's a 94 G20 van, stock 39000 mile LO5 with coated full length headers, rebuilt TBI unit new EGR valve (but turned off), new MSD cap, rotor, leads and aftermarket coil.
I'll need to order the colder plugs from the US so it will be a couple of weeks before I can try them.

[Kitch]

New CR43TS plugs have been ordered, now to see if I can find a local supplier of seafoam.

[Kitch]

Spark plugs arrived today. I used an engine conditioning spray on the 350 that was supposed to loosen carbon deposits, a little bit of smoke came out the exhaust but it wasn't like the smoke screens I've seen generated by other engine cleaning products on some youtube clips.
I took the van for a run on the motorway with no noticeable reduction in knock counts. Came back home and changed the plugs for the new AC Delco CR43TS's (not very header friendly in length) and went for another run. Since the previous run a storm had come up and I was driving into a strong headwind which loaded up the engine (5" of vacuum instead of the normal 12") knock counts rising constantly, it was too disheartening to watch all the knock counts so I changed down a gear to reduce the load on the engine.
So a bit of an engine decoke and new colder plugs but no noticeable reduction in knock counts.
I've aded a couple of photos of a plug I removed, it does appear to have some small flecks on the porcelain but I can't discount the fact they are a result of the engine conditioning spray.

[sturgillbd]

I reviewed your log and even though the knock counts are increasing under load, there is no knock retard in the log. The knock sensor is hearing something but it isn't enough to trigger knock retard.

Brian

[Kitch]

I've had the van dynoed previously and even with all the stock settings in the PCM the dyno operator was getting knock related spark retard and it was effectively taking more timing out each run.
Doing the dyno runs sitting right next to the engine with the cover off, he said he couldn't actually hear any knock so he disabled the knock retard and thats the way I've left it.

[Fast355]

Try increasing the 02 voltage for closed loop from the stock 350mv (.350v) to the mid 550mv (.550v) range across the board. You will find several tables with the voltage. Then reduce the PE entry TPS to about 40% to bring in the enrichment sooner.

[Fast355]

I've had the van dynoed previously and even with all the stock settings in the PCM the dyno operator was getting knock related spark retard and it was effectively taking more timing out each run.
Doing the dyno runs sitting right next to the engine with the cover off, he said he couldn't actually hear any knock so he disabled the knock retard and thats the way I've left it.

I had a buddy with a 1995 C1500 5.7 pickup truck, ALL stock PCM settings with long tubes, edelbrock 3704 intake, and tbi mods years ago that would knock retard so badly that the PCM would go into Low Octane Spark mode (pulls 5* timing from the calibration) on any highway trip more than about 8-10 miles.

[Kitch]

Last year when we went away for a family trip in the van (when it was dead stock) and while I knew the van was heavy we were going up a steep hill in it and I had my foot hard to the floor in first gear and the van was slowing down! Got up the hill but I was left wondering how people manage to tow anything behind one of these, I suspect knock retard was playing a big part in it's lack for performance back then.

The brand new L31 Vortec crate motor for the van is away at the engine rebuilders at the moment. I'm getting him to degree the new Comp Cams 252XFI HR-13 roller cam and check all the clearances, when I ported the Votrec heads the cylinder head guy that reassembled them commented he had to lightly machine the seats because from new they needed more than just lapping in to seat properly! so I figured I might as well see what other clearances are a bit out and if the compression doesn't end up too high zero deck it for nice and tight quench.

[Kitch]

What would be the expected result of increasing the O2 mV for closed loop?

[Kitch]

I've added a photo of my current PE settings for your comments on :-)

[Kitch]

Hi Fast, I was looking for the O2 closed loop table and I stumbled across the TPS for PE vs RPM table. I take it that was what you were previously referring to?

[Fast355]

Hi Fast, I was looking for the O2 closed loop table and I stumbled across the TPS for PE vs RPM table. I take it that was what you were previously referring to?

There should be two that look like that. PE and PE Fast Entry. Set the PE table to ~40% across the board. Set the PE Fast Entry to ~60%. Like the powervalve of a carb, you want to make sure when the engine is under load it has fuel.

[Fast355]

What would be the expected result of increasing the O2 mV for closed loop?

Slightly richer mixture, cooler cylinder temps, less spark knock, more torque, and better performance. The change you will see first is a BLM increase. I have run the o2 MV as rich as 650mv (.650v) on my old G20 when performance was the goal. Cruise air/fuel mixture at the time with my mild cam was in the 13.5-14:1 range but my 3.08 geared G20 really liked the extra fuel to help hold overdrive at 85 mph.

[Fast355]

Last year when we went away for a family trip in the van (when it was dead stock) and while I knew the van was heavy we were going up a steep hill in it and I had my foot hard to the floor in first gear and the van was slowing down! Got up the hill but I was left wondering how people manage to tow anything behind one of these, I suspect knock retard was playing a big part in it's lack for performance back then.

The brand new L31 Vortec crate motor for the van is away at the engine rebuilders at the moment. I'm getting him to degree the new Comp Cams 252XFI HR-13 roller cam and check all the clearances, when I ported the Votrec heads the cylinder head guy that reassembled them commented he had to lightly machine the seats because from new they needed more than just lapping in to seat properly! so I figured I might as well see what other clearances are a bit out and if the compression doesn't end up too high zero deck it for nice and tight quench.

I never had a problem getting up switch backs in the mountains even with my stock 305 almost 2 miles above sea level. 6-8% grade at 55-65 mph and pedal to spare. Now pulling a car on a trailer had me down to 45 mph in 2nd gear, foot to the floor into the secondaries (had a Q-Jet at the time).

Proper quench more than makes up for the added compression with a vortec head. A poor quench distance on a vortec head will ping more at 9.5:1 than a 10.5:1 with the right quench. Probably the reason I am having to fight my L31 head Express at the moment. The quench distance is just not there. When I tried to run a thinner head gasket, I had an internal coolant leak.

[Kitch]

I've changed both the PE tables, for some reason when I entered 40.0 in the "TPS for PE vs RPM" table it changed to 39.8 when I saved it? Close enough.
With the "TPS For Fast PE vs RPM" table I only had to change the value at 1200 RPM from 75 to 60. Looks like I had previously changed most of the % from 84.8 to 50 apart from the one at 1200RPM which it looks like I left at 75 (probably by mistake). I've now changed them to 60 as you recommended or 60.2 as it's been automatically rounded to.

A couple of how does this work questions:

I still don't understand the O2 mV change.....if the BLM increases hasn't the change made it go lean and eventually won't the Integrator go rich and try and trim the BLM back to 128 in closed loop off setting the change in the O2 mv, or am I missing something obvious here?

I see there is a "Stoichiometric AFR" scalar for open loop, if open loop doesn't use input from the O2 sensor for correction why does it need to know the air fuel ratio?

Also my "Base PE AFR" has a value of 1.5:1 in it what does that represent? Does it subtract 1.5 from 14.7 to give a new AFR of 13.2? I'm guessing that not how it works because I have a "PE AFR vs RPM" table that has values set in it for the various RPM points.

[Fast355]

I've changed both the PE tables, for some reason when I entered 40.0 in the "TPS for PE vs RPM" table it changed to 39.8 when I saved it? Close enough.
With the "TPS For Fast PE vs RPM" table I only had to change the value at 1200 RPM from 75 to 60. Looks like I had previously changed most of the % from 84.8 to 50 apart from the one at 1200RPM which it looks like I left at 75 (probably by mistake). I've now changed them to 60 as you recommended or 60.2 as it's been automatically rounded to.

A couple of how does this work questions:

I still don't understand the O2 mV change.....if the BLM increases hasn't the change made it go lean and eventually won't the Integrator go rich and try and trim the BLM back to 128 in closed loop off setting the change in the O2 mv, or am I missing something obvious here?

I see there is a "Stoichiometric AFR" scalar for open loop, if open loop doesn't use input from the O2 sensor for correction why does it need to know the air fuel ratio?

Also my "Base PE AFR" has a value of 1.5:1 in it what does that represent? Does it subtract 1.5 from 14.7 to give a new AFR of 13.2? I'm guessing that not how it works because I have a "PE AFR vs RPM" table that has values set in it for the various RPM points.

It has been years since I looked at the $OD Hack and code but I seem to remember PE working something like this. Log the Commanded Air/Fuel ratio to double check. If you open the throttle to 40% (above PE Entry) you will have to wait for a time delay or RPM bypass to happen for PE to start. Once PE starts the air/fuel ratio drops by 1.5 afr from stoichiometric. As the vehicle stays in PE there is a time adder that further subtracts afr until the desired PE air/fuel ratio is met. On the otherhand if you cross the fast PE threshold it just drops straight into the PE table. I maybe a touch off in places but should be pretty close. The air/fuel ratio is still being adjusted for even in open loop by the PCM. There is also a catalyst overheat protection routine that richens the mixture if the cat starts overheating that I generally disable.

The o2 crossover counts are what drive the integrator and the INT drives the BLMs. If you increase the switch point your overall tune will go richer due to lower o2 voltage and fewer cross counts resulting in a lean INT and BLM.