Which variable is more useful to watch, knock counts or knock retard deg.? When I do a medium pull from a standing start I see 22-29 knock counts. The knock retard degrees will jump from 0 to 55ish at around 2500-3000 rpm then go back to 0. I have done many pulls like this and see similar results, but never hear a knock and the plugs look fine. I understand that the knock sensor is just a listening device, is this just normal operation?

 

Both. You can adjust the level of knock retard that "adds" to the count. But I like to look at both.

Don't be lulled into the belief that JUST because you are not hearing knock that the knock is false. This is a common mistake that guys make when they get a scan tool and notice this. What is REALLY happening is the knock sensor is doing its job and pulling out enough timing to stop you from hearing it.

The decreasing levels of knock is due to the "recovery" routine reducing the knock retard after the initial "knock incident".

I have been doing a lot of experimentation in this area and I hope to have a "Discussion article" on Knock. My general conclusion is that TPI engines are very susceptible to knock retard, probably just by the design of the heads. They were designed around a lower spark advance as suggested MANY TIMES by Grumpy.

I have tested my Knock Sensor and even basically "neutered" it by capping it to very low levels. Guess what happened? That inaudible knock became VERY REAL at surprisingly low spark advance levels. Remove the cap, and then my knock sensor behaves like everyone's elses and pulls out a ton of timing.

Grumpy is absolutely correct when he says "Ignore the magazine articles, these heads were not designed to run 36-38* of spark advance". Those are different heads and a different induction system.

To further compound things, most of our engines have a "few miles" on them. While they may not be an "oil burner", there is enough wear in the engine to leak small amounts of oil that trigger detonation (also mentioned by Grumpy many times).

I have a lot more to say (and discuss) on this subject in the near future, but I want to do more testing. But, EVERYTHING Grumpy has been telling everyone is correct - these engines (induction/heads) were not designed around a large amount of spark advance @ WOT...and the engine DOES NOT LIKE IT.

On my L98, I am finding that getting 30* of EFFECTIVE Spark Advance is about the maximum. Remember also, that when knock retard occurs, the Total Spark Advance displayed by a scan tool is NOT the real Effective Spark Advance you are getting. If you reconcile the Total Spark Advance to the Relative Spark Advance (plus base), the Total Spark Advance is higher...by the amount of knock retard you are detecting.

The Total Spark Advance should really be called "Desired Total Spark Advance". People will see "39* Spark Advance" for their Total Spark Advance and forget to subract the 12* of Knock Retard to get their effective Spark Advance of 27*.

FYI, 30* of EFFECTIVE Spark Advance is a lot of advance. A typical L98 is lucky to get 26* of EFFECTIVE Spark Advance when knock retard occurs.

 

Very good points! I guess I am living in my carb tuning days, I didn't realize that the knock sensor works so quickly. So if I see a consistant knock retard and counts, should I work to tune out the knock as a result raise power that would be lost in the retard function, or is this just someting to live with? I have seen 87 degs. in knock retard a few times. Just for reference I am running Vette heads, so I am still playing with the timing. Thanks!

 

I suspect you are using Craig's software and it is really 8.7* of retard. If you got 87* of retard, your engine wouldn't even run.

According to Grumpy, he's found 28* of EFFECTIVE Spark Advance is what Vette Aluminum heads like. I am finding similar for the Iron Heads.

Be VERY CAREFUL if you "cap" the Max Knock Retard. I have been experimenting with this, but I am aware of the dangers and have a rebuild planned. Capping the Max Knock Retard vs RPM will eliminate any retard...but you can rapidly turn that "inaudible" knock into "audible" knock...followed by "bang and boom".

 

What about adjusting air/fuel ratio or fuel pressure to reduce knock instead of capping the amount of retard? Right now I'm in the data gathering process, I am trying to build a play of attack when the tuning really starts.

 

Try experimenting. I found that I would that the knock sensor would detect "knock" at very low spark advance settings too...and adding fuel never eliminated it.

It tried LOTS of combinations. But I am still doing some analysis and don't want to "let the cat out of the bag" just yet.

Let's just say that key is the proper mixture with the correct spark advance yields the most performance. After much testing, I found the 28*-30* range with the RIGHT MIXTURE yielded the best performance with my knock sensor capped at a very low value. In that range, I got my best performance with no audible knock.

Increasing the advance beyond that 30* produced audible knock (since my knock sensor was effectively disabled). Adding some extra fuel helped for a "degree or two more", but the extra fuel made the mixture incorrect, so my performance suffered. And beyond "a degree or two", no amount of fuel stopped the audible knock.

So, the right mixture/spark advance that I am now playing with is in the 28-30* range. I need to play a little more to see if one is better than the other, but so far the differences seem negigible. So I will probably run with 28* as this is less harmful than 30* if there isn't any performance increases to be had.

BTW, if I "unleash" my "knock retard" and advance it to 36*, I end up getting 12* of retard, causing my EFFECTIVE spark advance to drop to 24*. Guess which one is slower?

 

Something about capping the knock retard scares me. If I had a mule motor I would have no probs doing so. The way your are aproaching this issue sounds nice but how many times can you get serious knock before something melts? I have learned that PROM tuning is 99% research, 1% doing. But it pays off in power and learning for futur projects. Maybe after we all learn the in's and out's of tuning we could get jobs a nitro funny car tuners Ahhh even bigger blow-ups.

 

I should add that I am now testing my timing with the Knock Sensor "uncapped" and at 28*, even with an "open" knock sensor to retard as much as it wants, my number of knock incidences and magnitude of retard is the least.

It is just confirming that my combo (basically stock L98) does not want a lot of advance. And when I put it where it wants to be, the knocks go down. Amazing.

 

I think everyone needs to read this topic. Glenn, you are the man.......

 

Nope. Grumpy is. I am just getting confirming data supporting what Grumpy has been saying all along.

 

Glenn:

Your comment about running 28 deg with full knock retard available is very interesting. I just finished my spark table on my 9.6:1 alum head 383 TPI with no egr. When I had way too much advance programmed in my PROM, I found that the ecm would over retard the spark trying to get the knock to stop. Sometimes it would max out at 18 deg retard and the motor felt like it had no power. As I kept removing timing from my PROM, and modified my attack and decay rates as you have discussed in the past, I found the ecm would only pull just what was needed and I made more power and better accelleration. My feeling is that the knock retard algorithm in the ECM code will pull a little timing, listen again, pull more, listen again, and if it still is getting knock it just pulls massive amounts to get the knock to stop. When you are very close on your advance curve, the effects of spark retarding is noticeable right away as reduced knock and the ecm is happy since knock severity has now decreased. My belief is that it is fine to have knock retard of a few degrees in normal driving since the ecm is giving you the most advance it can safely. If you set your spark tables so that you NEVER get knock retard or knock counts, it probably means you are UNDER advanced in certain areas and conditions. My current WOT advance is 17 deg plus 6.6 deg(PE adder), plus the 6 deg base timing for a total of 29.6 degrees at like 3600-4000rpm. Seems this is very close to the 28 deg you are talking about.

------------------
Dave Zelinka

 

This is an excellent discussion. The more I learn the greater my appreciation for engineering job done by GM back in the 80's. The knock sensor strategy is more accurate than many presume.

I look at the knock counts as the data the sensor provides and knock retard as how I choose to act on this information (ie. amount and rate of timing change) via the parameters defined in the PROM.

My limited experience is right in line with the examples cited by Glenn. Each successive PROM I've done for my stock iron head L98 has reduced WOT timing and the current revision still has too much timing (29 deg @ 3500 RPM increasing to a max of 34 deg @ 4800 RPM) since I'm seeing knock between 3500-4000 RPM. Adding fuel has not helped. Glenn, I look forward to the upcoming article.

 

Yes, I can't wait either for Glenn's article! Thanks for all the great info guys. I logged some more data tonight, at 31* at WOT it retarded the timing a ton. Time to burn a new prom, I will try to jump it back to 28* and see what happens. I am not running a EGR on my vette headed L98, so my case is a different the a stock L98.

 

Sorry guys I meant that I am seeing 37* WOT not 31*.

 

Are you guys running 28 to 30 degrees on 87 or 93 octane? What I am saying is it possible to run say the stock ARAP spark table with 93 octane with minimal retard?

 

93 or the best your area has..
You will pay x dollars to go x distance. What you save at the pumps you'll spend in early valve jobs.

The premium fuels burn with min ash. Ash is what holds the exhaust valves off of their seats and allows them to burn. Mind you we're talking microscopic traces.

It's nice seeing guys see the light !!!.

It's all about timing, and fuel.

As far as reasearch and doing it's more like 51% doing and 49% listening / reading.

If your using care in your tuning there is about no way you'll be running too much timing long enough to do any damage, if and only if your into notes and painstaking detail.

 

Grumpy: So the ash thats left after the fuel burns acts like a thermal insulator and does not let the exhaust valve transfer its heat out to the cylinder head? I never heard about this before, you always have very interesting comments. It is clear to me you always look a little deeper to really understand whats going on. Thanks for sharing with us!

------------------
Dave Zelinka

 

Roughly 60% of the exhuast valves heat transfer is thru the valve stem, the rest is thru the valve seat. Losing the seat for heat transfer is a death warrant the for valve.
There is alot going on with the valves that contibute to their life expectancy.
Tuning rich to use fuel to cool the valve when at WOT is a trick sometimes used, also to control chamber temps., but as illustrated from the above there is a big downside.

 

This is excellent reading, and a hot topic. My scan tool shows 2 spark advance parameters, one relative to TDC and one relative to REF. Plus I have the Spark Control Counts as well. But I don't have the knock retard deg parameter. Is that one unique to Craig's software?

I have never seen any spark counts with my tool, it is allways reads 0 counts. But I haven't done much with my spark tables, been working on fuel first. I will be getting into the advance stuff real soon!!

Can't wait for that article Glenn, keep up th e good work!!!

------------------
Best ET 14.413 @95.57 without
pulling valve covers or manifolds.
Also with stock 2.77 rear end!!!

 

Remember start with a min amount of timing and WORK UP, from that. It seems like every one is starting high, and then backing down.
Detonation is a very real engine killer.
Cruise and idle aren't that critical for the amount of time it takes to dial in an engine.
Also, remember that in the Distributor'd engines there is a limit for the advance the ecm will calculate, so in most calibrations a cruise advance of 45d (as limited in the code) is all you can run.

One item I just found last night in the LT1s from the F Bod to B bod calibrations is that fuel VE tables are about the same, and they just changed the timing, with the F Bods getting more. And there are some serious timing changes, ie 10d in spots. Just a FWIW,

 

Mike T.,
I'm working on that with Ease right now. It's a software issue.

------------------
Neptune, NJ
White '87 Vette - 383 Superram ported and matched to TPIS big mouth intake, Eagle cast crank, 6"eagle forged rods, srp/je forged pistons, LPE 219 cam, ported heads(2"x1.56" ferrea valves), Comp cams pro-magnum 1.6 rr's, double springs, titanium retainers,Hypertech power coil and cap, 4 bolt block, TPIS headers, auto(trans kit,shift kit),Pro Torque 2600 stall converter, k&N filter, 3.07 rear diff., Silver ZR-1 wheels 17x9 and 17X11

 

GlenS, software issue with what, the fact that I get 0 spark control counts, OR the lack of the "knock retard deg" parameter? I also use the Ease Scan tool. Don't you just love it?? LOL

As far as my spark advance settings, my plan is to back the distributer advance back down to 6deg, from 10deg now. Then start using the values from an '87 350TPI bin. Then start taking lots of data. I'm still using other table values from my stock '88 bin. I think the ARAP spark advance table is a bit too aggressive for a somewhat stock L98.

 

A buddy of mine from another list made mention when he was tuning his motor in that he was getting a touch of knock @ one point & rather than reduce timing he decided to add fuel to see if he could not bring it under control. no matter how much fuel he added it would not stop the knock. what he did was back up to the previous row on his table & added a bit of fuel there. this prevented the knock from occuring.

his conclusion was that if he added a bit of fuel before the motor started to knock (or started getting audible, I suppose), he could controll it. but it it already started knocking, he could not controll it with fuel.

I think that he ended pulling a bit of timing out, but I though his test was quite interesting.

BW


[This message has been edited by Bobalos (edited July 23, 2001).]

 

28-30* max advance- Is this with the stock cam and intake? I suspect that you could run more timing with a cam that has more duration than stock. More duration will bleed off cyl pressure and permit more timing.
I bring up the intake because the long runners in the TPI are good at filling the cylinders at lower rpm, causing more cyl pressure. So, WHEN is the knock retard occuring? Maybe if the timing is added later in the power band, you should be able to add more theoretically.

Heads are another issue. The TPI heads still have the crappy combustion chamber. Modern combustion chambers, such as the Vortech heads, most aftermarket heads, and especially the GM fastburn, will also permit more timing.

If you have some or alot of these aftermarket parts on you engine, I suspect you could run more timing. I'll bet an engine wtih a miniram, or superram(so you can acrually rev it some), aftermarket heads, and more cam duration, would be able to accept the increased timing, providing that the fuel ratios are somewhat optimum. Wouldn't life be jolly if we could all afford wideband O-2 sensors?



------------------
82' Z-28. 327" w/ flat tops. 216/228/112LS. Pocket ported 041 heads 1.94/1.50 SS valves. World Class T-5 from a 91'. Best run? Runs with 2001 vette to 150mph! Soon to be converted over to TPI w/ supercharger.

 

You just might do some homework, well designed heads allow for making more HP with less timing.

Once the mechanicals of the engine are set then you tune to them. Long runner short runners whatever.

WB O2s, have little to do with anything. They are just a tuning tool. No even as good as reading a spark plug. Yes, I have one, and 99% of the time it's on the self.

while currently under redesign, there is a DIY WB O2 vailable at DIY-EFI. the new version uses few components and is easier to assemble.

 

Actually I have done head homework. While your statement is very true, and commonly known, A better combustion chamber also allows for slightly more timing.

A better combustion has a much faster burn rate to create more cylinder pressure, and do it sooner. This happens because the mixture is more completely burned while the piston is closer to TDC. So the combustion occurs at a point when there is more initial cylinder pressure from compression.

Turbulence is also important because it keeps the mixture atomized better. The mixture in motion stays better atomized. The atomization is better in part because the swirl and turbulence moves the mixture past the hot exhaust valve area, and the fuel droplets get vaporized by this heat.

The spark plug in a better comb. chamber is closer to the center of the bore, and is aimed at the exhaust valve. The mixture is better atomized around the exhaust valve in part due to the heat of the exhaust valve.

The faster the mixture gets burned, the less of a chance for another unwanted one to counter it and give you a ping or detonation. Because of this you can run slightly more timing.

I am not stupid as your attitude suggests. I do know a thing or two about heads.

As for a WBO2 having little to do with anything, I do not agree. It tells you your AFR. While AFR is most certainly not the whole story, it is a powerful tool, and important in engine function. You can tell very quickly where you are at. Just because you do not use one does not mean that us novice(stupid) tuners would not benifit from it. If you like looking at spark plugs, fine. If you like your WBO2, fine. To each his own.

Finally, Thank you for the DIY-EFI info on the WBO2. I was wondering where some info on that might be found.

------------------
82' Z-28. 327" w/ flat tops. 216/228/112LS. Pocket ported 041 heads 1.94/1.50 SS valves. World Class T-5 from a 91'. Best run? Runs with 2001 vette to 150mph! Soon to be converted over to TPI w/ supercharger.

 

I know squat about heads but I do know this: you don't need extra timing when you DON'T need it. Max power is made from the perfect amount of timing not the max amount before knock.

Turbulence is chaotic motion in the cylinder. Swirl is motion around the cylinder perimeter. Tumble is the downward motion into the cylinder. Of these, turbulence is the least important. This is from reading and not from actual head work that I've been around. Personally, I've never seen any of the things mentioned above actually qualified in a running engine. In theory, they hold up well but the real world often doesn't care for theory.

You're right, a WB can tell you the exact AFR and quickly so. However, so can plug reading. One thing a WB can't tell you is if you have knock. Plug reading will. You may rely on your ear for that but I happen to know you can't hear and distinguish every noise an engine makes as quickly as they can happen. I agree with some of the things you say and not with others. To take a reply from someone who is only trying to help and turn it into a false attack on your knowledge is one thing I don't agree with. In an earlier reply you also stated that a bigger cam will bleed off pressure and should allow for more timing. It might but only at idle as once the cam comes into its own it will have more than enough cylinder pressure. To each his own is right but I would rather have no knock than superfluous timing.

 

Grumpy: Your note about plug reading versus a WB O2 is very intersting to me. As I got into efi tuning and PROM burning, I had one dream, and that was to someday own a WBO2 sensor and display. At the time it was my belief that with it, I could perfectly tune my engine. Anyway, I still dont have one, but during a recent discussion with a professional efi engine tuner I found it very interesting that he is not using this technique at all, and he can easily afford one. Here is a question I have regarding plug reading. If I make a WOT run and stop to pull the plugs how will I know what actually happened over the whole run? I mean how will I know if I have the right mixture at ALL rpm ranges?? I mean couldnt I be lean at the low rpm's but if I am rich on top wont the plug just show rich(last event)? Or worse, as I slow down to a stop wont the mixture there affect how the plug looks? Also, could you elaborate a little more as to why your WB O2 sits on the shelf 99% of the time? I am still contemplating buying one!

------------------
Dave Zelinka

 

I was also curious about the WOT plug checks. When I did a few with a chip I knew was lean, the plug was white but wet. Kind of confusing but I just figured it was because of the mixture and because I stopped REAL quick and didn't let the car "exhaust" the mixture still in the intake. Then when I looked at a few of the plugs I noticed some were really wet and others were not as bad, I figured it was probably because of the intake valve was left open on some but I have no clue.
Please, somebody put up a GOOD plug reading source for us novice tuners. Pictures would be even better but I have my Chilton's for that .

------------------
, Jon (350 TBI!)
91 Red My website

 

Correct me if I'm wrong but I think to get a correct plug reading you have to shut the engine off at the point where you want to take the reading from. If it's top end then you better have a good long strait road to coast down on. Like I said above someone correct me if I'm wrong, pllleeeaassse!!

Steve

 

You start conservatively,and do cuts.
You work up adding timing in just a block or two at a time, not the whole thing at once.
Takes lots of cuts to get a WOT curve right.
ie work at say 800-1600, then 1400-2200. Just small areas at a time.

 

Grumpy, I'm I not correct in stating that you have to shut the engine off at the point you want to take the reading at, weather it be at 2000rpm or 5000rpm? I think some of the tuners on this board may be under the impression that they should turn the engine off after they have come to a complete stop.

Steve

 

What ever.
You can attach whatever you want to something addressed to you, but that doesn't make it true. Homework has absolutely nothing to do with stupid. Exposure, and open mindedness is what matters.
Optimum AFR is dicated by what the engine wants and needs. Once you have found the best performance of your engine then recording it can be an advantage in diagnostics. Tuning to get a specific AFR won't always work for best performance.
Plug reading is a science. Mastering that can be more important then most would expect.

 

Hey Grumpy' thanks for your imput on all subjects and questions that you respond to. I learn alot just reading your post and replys. Maybe one day I will be able to give hands on expirienced advice as knowledgable as yours. As I have read in the Bible the student is never greater than the teacher, only equal to at best.

Steve

 

There is some hysterise (?).
A single knock event can cause 3d of timing to be pulled out, and that stays out as a function of the knock decay rate. On the oem cals, they ramp it out rather slowly, since they are for *stock* engines you can raise the knock retard decay quite a bit. I've seen 250-400% used. You gotta play with it.
Remember there is always going to be that bad tank of gas. So you need some room.
Like I really doubt I'll need 30d on knock limits, but I have it there just cause....

 

For a street daily driven SBC running EFI 8d works really well. Too little and the engine kinda windmills on start. To much can make for poor restarts. 10d on real hot day might not catch on the first crank, and then you might need to use the clear flood mode to light it. The battery voltage corrections, start adding alot of fuel when the battery voltage drops like on a real hot restart where the starter drags.

 

quote:
"Exposure, and open mindedness is what matters. Optimum AFR is dicated by what the engine wants and needs. Once you have found the best performance of your engine then recording it can be an advantage in diagnostics."

-I agree 100%.

quote:
"Tuning to get a specific AFR won't always work for best performance.
Plug reading is a science. Mastering that can be more important then most would expect."

-I agree here also and briefly touched on that in an earlier entry.

Quote:

"While AFR is most certainly not the whole story, it is a powerful tool..."



------------------
82' Z-28. 327" w/ flat tops. 216/228/112LS. Pocket ported 041 heads 1.94/1.50 SS valves. World Class T-5 from a 91'. Best run? Runs with 2001 vette to 150mph! Soon to be converted over to TPI w/ supercharger.

 

If by "more timing" you mean more advance, you have just contradicted yourself.

Actually, a better head design (i.e. "soft" heads and "squish" heads, along with proper piston contour, can do much more work with *less* advance. The idea of advance is to start the burn so that at the piston crosses over TDC, peak thermal expansion rate is achieved in the burning gases. A combustion area (chamber with account for piston contour) which allows the flame front to advance more evenly will allow a later spark, and more of the burn can happen later in the stroke, which ideally will occur near the center of the stroke where piston speed is at max.

Turbulence is a very poor term here, since the word really refers to chaotic wasted motion. In fact, in most fuel injected engines, atomization is not a problem unless there are large inconsistancies in the ports which would lead to precipitation of the fuel. Sometimes, turbulance and eddy currents can kill atomization.

While the exhaust valve can help to keep the fuel in suspension with the heat, proper intake port design would send the new charge away from the exhaust valve at low lifts, so that overlap time can be more effectively used to purge burnt gases from the cylinder.

Effective compression can also have an effect on timing. Ported heads and large cams can actually decrease the amount of advance needed because with a better air/fuel charge, the mixture can become more volatile.

People who are trying to eliminate knock with fuel are really only doing two things: diluting the air/fuel charge, and cooling the combustion chamber. Both of these things will cause a slower movement of the flame front (a less volatile mixture) which will then require more advance for the "full burn" to occur at the correct time.

I do alot of work on turbo cars, and many of my colleagues are running 30+ psi of boost and literally dumping fuel in to keep the intake charge and combustion area cool. It's a very inefficient method which can have other negative effects.

Obviously, our moderators are a great resource on this. I hope I haven't bored anyone with my rambling, but with a discussion this good, I had to get involved.




Matthew A. Green --- shelby@voicenet.com
M"squared" Racing! --- http://321.net/~m2racing/
85 Shelby Charger soon to be EProduction... or maybe SuperProduction... or maybe...
87 GLHS #314
90 Formula 350 (I think I'll have a V8!)

 

I have been running a TPIS miniram and it is always getting knock counts. In a 730 ECM with @8d Bin would only pull timeing from the Main Spark Advance ~vs~ RPM ~vs~ load and (EXT) or would there be someplace else?