Setting advance on an 83 SC: final word ?

[Deschodt]

I would not put too much stock in my readings with the vacuum connected. As it was explained to me by Paul, it fluctuates a lot and that is why you should measure with the lines off - I was just curious, and like Bill said, sustained 6000 RPM with no load is a bad thing...

While yesterday I did my measurements with a helper, stabilized rpm for a few seconds at each 1000 RPMs, today I was on my own so I simply zinged the motor by hand from the engine compartment, with my timing light preset at 20-25-30 to see what I could get to, maximum... So while I did see 30, it was not a stabilized thing, I just wanted to see if it could get "out there", I don't think it can...

I'm gonna keep it as is, essentially factory settings but retard line removed and plugged. And of course 93 octane only. Amusingly after all this, on my way to get pizza, I started hearing an ugly rattle from back there... I immediately thought the worst, screwed up my car experimenting with timing, or maybe one chain tensioner had collapsed - that loud... Turned out my FG bumper was rattling against one of its mounts, just a loose screw. Phew !!! ;-)

[fumanchu]

Thanks for clarifying the way you did your timing as your results were somewhat confusing. Best of luck on your setup, seems to work O.k., just wondering how much we are leaving on the table as far as power. Safety seems best in this case esp. in Cali where we only have 91 octane.

[psalt]

I would not put too much stock in my readings with the vacuum connected. As it was explained to me by Paul, it fluctuates a lot and that is why you should measure with the lines off - I was just curious, and like Bill said, sustained 6000 RPM with no load is a bad thing...

Greg,

You need to separate what happens at rest blipping the throttle from what actually happens under load. Vacuum advance is additive, but in practice the engine rarely runs at high rpm with high manifold vacuum. If your 35 BTDC readings were with vacuum connected, they are meaningless. Pinging happens under load when there is little or no vacuum advance. You should be able to run at least 25 BTDC without pinging under load with 93 CLC octane unless there is something else wrong with the engine.

[psalt]

That is is way cool diagram. I have never seen such a thing.
Is Zb pre-ignition or detonation? I am thinking Zb is detonation.
It looks like the combination of early ignition that builds heat and pressure is kind of multiplied by an increase in pressure from the piston being closer to TDC until where the mix auto ignites or flash's instead of burning in an orderly manner.
I would suspect pre-ignition would show a spike in pressure before TDC but not sure.
I think I read that we want peak pressure at about 14 deg after TDC for peak HP. That would be Za it looks like.
With Zc the mix is still burning and adding pressure after the other two times but the piston retraction is such that the the piston dose not see as much peak pressure. This is also the condition where more exhaust heat is created as the gasses are still expanding when the exhaust cycle is started.
Can near ideal timing be measured from EGT's. That is is it the low point between detonation and late ignition? Or maybe we can start w late timing and start advancing until temps start to taper as they get cooler?



This is the standard Bosch diagram of detonation, it appears in many of their books, the little Blue Handbook and Gasoline Engine Management. Z b (2) is detonation (knock), the normal 2ms burn rate is accelerated by the spontaneous explosion of unburnt fuel, causing the abnormal pressure spike. It is a bit of a cartoon and the actually pressures are different. The point to understand is that knock is not normal burning, it is an uncontrolled explosion with no benefits. The increase in pressure does not result in higher output, too much work is lost BTDC. This is why power drops once the knock limit is reached. The lower output of retarded timing occurs because of the mechanical advantage difference and a loss of heat. EGT rises as ignition is advanced, falls once knock occurs, but is not a safe reliable indicator for tuning. Peak CO2 can be a good indicator of ideal ignition timing. With all hemi headed engines with wide VIA's and domed pistons (like a 911) it is safe to assume they are knock limited on pump fuel. The "ideal" timing can only be found using the highest octane race fuel, then they are retarded (detuned) for what ever octane fuel will be used, plus a safety margin. The peak output for N/A hemi head road race engines was in the 50's when they ran 14:1 compression, domed pistons and 50 BTDC ignition timing. The reason this was possible was they were burning butyl alcohol, methanol, benzene and nitro instead of gasoline and the fuel acted as a coolant to prevent detonation. Corn politics aside, E85 could be a 911's best friend.

[psalt]

Is there a load difference when parked in the garage vs. on the road @ 6000 rpm affecting the way the vacuum advance is working?


Yes, load means the engine is pulling against something, not sitting in neutral. Is there a difference between the way the engine's rpm rise when blipping the throttle in the garage in neutral and when you floor the throttle going up a steep hill in 3rd gear? In the garage when you are blipping the throttle, you cannot hold it open for more than a second or you will overrev the engine, and there is plenty of vacuum around to screw up the timing reading. The vacuum advance is there to add timing when you are cruising at 60 mph and there is 16 inches of vacuum in the manifold. At 130 mph @ 6000 rpm with the throttle open there is less vacuum in the manifold than the threshold of the advance and no extra timing is added. As soon as you floor the throttle on the road, the extra advance disappears immediately.

When tuning an engine, you have to separate what happens when it is working and what is happening at rest. The same situation occurs with measuring AFR's. Measuring AFR's at 3000rpm in neutral in the garage is also meaningless. It is what happens under load that matters.

[911st]

On a 911 w Vac advance timing in the driveway at 6k will be advanced if the vac advance is not disconnected. Under WOT the vac would go away and timing would be less.

However, it seems most SC's have "Vac Retard". This should be hooked to a vac line that only lets it work at idle. Thus, even if it is left hooked up it should not effect in the driveway timing.


psalt,

Thx for the info. I would like to see a graph for pre-ignition. I suspect that in most cases it would show the pressure spike earyer than detonation.

You note that EGT increases when timing is advanced. That makes sense as it is making more power.

I think there is also a point where it increases when timing is retarded as the fuel is still burning in the exhaust (i.e. vac-retard). I thought old race cars used EGT's to tune but that must be on the fuel side?

Still, setting timing on a Dyno after the AFR's are confirmed w someone that knows what they are doing seems the best way to go remembering to pull things back to provide a margin of safety as an SC can not manipulate timing to deal with changing environmental conditions like an EFI car can.

Better yet, a mapped programmable MSD could make for a significant improvement as the timing curve can be bent to better fit the motor.

Fun stuff.

[psalt]

However, it seems most SC's have "Vac Retard". This should be hooked to a vac line that only lets it work at idle. Thus, even if it is left hooked up it should not effect in the driveway timing.


The engine in question, an 83SC, has both vacuum advance and vacuum retard. There is no "should" or "thus" about it, both need to be disconnected and plugged to accurately measure the timing against the factory spec. The retard is hooked up to manifold vacuum. Try it yourself and you will see your error.

Pre ignition is almost never an issue on a N/A road car engine and is usually a red herring in these discussions. Everyone I know that can speak intelligently on this topic uses the term "knock". The Bosch chart shows the ignition point before the pressure spike, so it is obviously not pre ignition. The best way to understand this issue is to start with the indicated pressure diagram of an SI engine and get an understanding of peak cylinder pressures. An engine with a cranking compression of 180 psi, will show a peak cylinder pressure of +600 psi running, and this can jump to +2000 psi under detonation. The magnitude of the pressure spike is one issue, when it occurs and how fast it rises is another.

[Bill Verburg]

In post #14 I usually used the term pre ignition, my apologizes.

preignion is the early initiation of aa flame front in the combustion chamber, usually from a hot spot, but it can alos happen because of too much pressure too early.

The Bosch diagram shown in the same post is illustrating abnormal combustion due to high cylinder pressure though not occuring as early as preignition. The effects are similar, damage to rings pistons, valves etc can and will occur. Sometimes you can hear the engine knocking but most of the not.

It's your engine and if you want to experimant have fun, but beware of the potential for damage.

[Demetri]

Now that you guys have your heads wrapped around this, perhaps you can help me out. Where should I set my timing?

I have a 83 SC ROW with only one vacuum line on the distributor. I sometimes hear it knock, usually on hot days . It's set at 5 BTDC and gets up to about 35 at 6000 rpm. As stated before, here in CA we only have 91. I put in 100 on occassion, and that seems to do the trick. But at 6 bucks a gallon, I can't do it all the time. Any advice wouid be appreciated. thank you.

[911st]

This is just a guess based on an optimized EFI ignition table I have for a low compression 911. Something in the -28 to -30 static w no vac line at hight rpm should be ok.

It might take more depending on fuel quality, AFR, and temp.

I am not very sure about this. I always thought timing for a 72 911S at 5k was in the -32 range. That is with a cam that lowers effective compression a bunch.

On my 2.8 RS MFI twin-plug running over 10/1 CR is was about -24.

Can anyone verify the 72 911T, E & S timing spec?

[Demetri]

911st,

That makes sense. I think. I do have a higher compression motor. What was the timing of on your twin-plug at 900 rpm?

I have set mine at 2 TDC and richened it up a bit. That took care of the knock, but did not make for a smooth ride. It always felt like i had to be on it. Good for me but bad for the Mrs., which means bad for me. You ever see that Mr. and Mrs. Potato Head commercial?

Would retardeding my timing from 5 to 2 BTDC at 900 rpm, make it go from 35 to 32 at 6000 rpm?

[Mothy]

What would be the main reason for doing all this? HP, Torque, Emissions, Economy?

It's an interesting read to follow as to the methods or approach, pros and cons etc. but for someone like me with limited knowledge of finer technical aspects, I'd like to know why you would vary from the factory settings in the first place. It also makes the thread more complete as a reference.

Tim

[DanielDudley]

Mothy, the way I understand it, gasolene takes time to burn. From the moment you spark it, a flame front starts and spreads through the combustion chamber. You want as much of that flame burning as possible when the piston is close to top dead center, so that the piston will have push during the whole downstroke.

Of course as the engine revvs higher, you have less and less time for the spark to ignite. If you are at 0 degrees advance at 6000 RPM, your flame might not be burning well intil you are well into the power stroke, so you are losing power. Advancing the spark will put a smalll flame into the chamber before top dead canter, so that there is a decent fire going during the power stroke.

Of course, at 6000 RPM, things are happaning six times faster than at 1000 RPM. You do not want very much advance at low RPMs or the flame will travel too fast, and actually push down on the piston as it travels up to top dead center. This will often lead to an engine that balks when cranking it over to start. But the real issue with too much advance is that you have maximum compression at top dead center in the power stroke. The explosion of the spark creates a further increase in compression, and this can cause the gas air mix in the combustion chamber to spontaniously ignite far away from the flame front. This is called, knock, dieseling, pinging or preignition.

The combination of the two flame fronts creates way too much pressure way too early, and leads to extremly high compression, and a shock wave where the two flames meet. It also leads to very high temperatures, and the pressure isn't really usefull at pushing the piston when before top dead center or at TDC, so the heat and pressure attack the piston and rings, instead of moving them.

In a 911 motor, it is also worth noting that the Valves are very large. In this case, the spark plug has to be on one side of the chamber or another, as the valves are close to touching in the center. This means that the flame front has to travel much further to reach the opposite side of the chamber. This gives the gas air mix more time to spike compression during detonation, and makes the 911 engine more sensitive to too much advance than a four valve engine, where the plug is smack dab in the center of the dome. It also requires that the pistons have a dome to increase compression, and this can lead to other problems.

In fact, when you hear about twin plug heads, the idea is to have two controlled flame fronts on either side of the valves, so that the controlled burns only have to travel 1/2 as far, and thus pre ignition is prevented, and a much softer explosion takes place at the correctly timed interval. This is why you need to run twin plugs with high compression 911 motors.

Finally, we need to realize that the reason that predetonation pinging only happens during full throttle application is due to the fact that more fuel and air are being sucked into the chambers at WOT. Too much compression + too much fuel + too much advance = knocking. The reason higher octane fuels can prevent this is because octane relates to resistance to ignition. That equates to resistance to preignition in a high compression motor.

I hope that was somewhat clear.

[don gilbert]

being a semi pro drag racer for 20+years, we advance in small increments until pinging is heard, and then back off a degree or two. which i know is hard to hear on a 911 if you are not used to what detonation sounds like. I like to load my pcar in 3rd, up a hill, and i can tell if its pinging. I run 33 degrees max, (retard line disconected, have 9.5 comp. use 93 octane shell, and have an afr gauge to check the mix at all rpm. Too lean a mix and to much advance is not good. A dyno makes things easy, but not all have acc. to one.

[psalt]

I have a 83 SC ROW with only one vacuum line on the distributor. I sometimes hear it knock, usually on hot days . It's set at 5 BTDC and gets up to about 35 at 6000 rpm. As stated before, here in CA we only have 91. I put in 100 on occassion, and that seems to do the trick. But at 6 bucks a gallon, I can't do it all the time. Any advice wouid be appreciated. thank you.

Demetri

Your engine has higher (9.8:1) compression and was tuned to run on 98 RON with 25 BTDC max ignition timing. 98 RON is around 93 CLC octane fuel. Unless you are a magician, you can't run the engine on lower spec fuel with 10 degrees more advance, than the factory could. The answer is you need to raise the octane of the fuel or reduce the timing. Your distributor should only give 20 degrees of mechanical advance, so I would first check it is the right part number, then question your measurements. You also want to check the AFR at WOT to make sure it is not too lean. The only ways to lower the octane requirement of your engine is to improve cooling, lower the intake air temp, or get the oil mist out of your intake.

[psalt]

Daniel, you are almost there, here are a few thing to consider:


Mothy, the way I understand it, gasolene takes time to burn. From the moment you spark it, a flame front starts and spreads through the combustion chamber. You want as much of that flame burning as possible when the piston is close to top dead center, so that the piston will have push during the whole downstroke.

>You want the fuel to burn so the peak cylinder pressure occurs around 14 BTDC, not TDC, to maximize output. All of this discussion is at WOT under load, when max output, torque and VE occur. Different things are happening the 90% of your driving at part throttle. Think of it this way, if you had an in cylinder pressure sensor, you would adjust the timing advance to peak at 14 BTDC at every rpm.

Of course as the engine revvs higher, you have less and less time for the spark to ignite. If you are at 0 degrees advance at 6000 RPM, your flame might not be burning well intil you are well into the power stroke, so you are losing power. Advancing the spark will put a smalll flame into the chamber before top dead canter, so that there is a decent fire going during the power stroke.

> the burn time is around 2ms. There is something called ignition delay, so you need some advance even at lower rpm to get the pressure peak at 14 BTDC.

Of course, at 6000 RPM, things are happaning six times faster than at 1000 RPM. You do not want very much advance at low RPMs or the flame will travel too fast, and actually push down on the piston as it travels up to top dead center. This will often lead to an engine that balks when cranking it over to start.

> If you study the indicated pressure diagram of a SI engine, you will see that all engines are fighting the compression pressure. The work lost on the compression stroke is offset by the work done on the expansion stroke. Since all engines use ignition advance and the more the advance, the more work lost on the compression stroke, the concept of MBT (minimun advance for best torque) is used.

But the real issue with too much advance is that you have maximum compression at top dead center in the power stroke. The explosion of the spark creates a further increase in compression, and this can cause the gas air mix in the combustion chamber to spontaniously ignite far away from the flame front. This is called, knock, dieseling, pinging or preignition.

> These terms have different meanings , it is a mistake to interchange them. The spark does not create an explosion. The best way to understand this is under normal combustion the fuel is burnt in a regular four stage process. Under abnormal combustion, the fuel explodes. You do not want the fuel to explode.

In a 911 motor, it is also worth noting that the Valves are very large. In this case, the spark plug has to be on one side of the chamber or another, as the valves are close to touching in the center. This means that the flame front has to travel much further to reach the opposite side of the chamber. This gives the gas air mix more time to spike compression during detonation, and makes the 911 engine more sensitive to too much advance than a four valve engine, where the plug is smack dab in the center of the dome. It also requires that the pistons have a dome to increase compression, and this can lead to other problems.

>The history of the hemi goes like this, the initial race engines of the 1920's had small bores. In order to get decent sized valves in a small circle, you need to splay them out and use a wide VIA, first over 100 degrees, then 90. The wide angle results in a higher bigger volume CC, so you need to use domed pistons to get a decent compression ratio. Now, think of the dome piston at TDC and you have an orange peel combustion chamber. Later designs narrowed the VIA, Jaguar used 70 in the 1940's, Porsche 59, BMW and Lotus 54 because of this problem. There was one American that used 38 in 1919. The 1912 Peugeot, four valve pent roof chamber does not have this problem, but it was masked for many years because of fuel chemistry.


In fact, when you hear about twin plug heads, the idea is to have two controlled flame fronts on either side of the valves, so that the controlled burns only have to travel 1/2 as far, and thus pre ignition is prevented, and a much softer explosion takes place at the correctly timed interval. This is why you need to run twin plugs with high compression 911 motors.

> this drops the MBT around 10 degrees. For example, the ignition timing on a twin plug 917 was a fixed advance of 27 BTDC at all rpm.

Finally, we need to realize that the reason that predetonation pinging only happens during full throttle application is due to the fact that more fuel and air are being sucked into the chambers at WOT. Too much compression + too much fuel + too much advance = knocking. The reason higher octane fuels can prevent this is because octane relates to resistance to ignition. That equates to resistance to preignition in a high compression motor.

> predetonation pinging is an oxymoron, you can't have a noise from a shockwave before it happens unless you are clairvoyant. I find it much clearer to think of normal and abnormal combustion, burning or exploding (knock). Too much fuel is not an issue, more likely not enough. Nothing below 14.1:1 burns anyway, the extra fuel for max output, 12-13:1, is a coolant to allow more timing. Preignition occurs before the spark event, so unlikely in a N/A road car in operation, to be just added confusion.

Learning about how a diesel engine work, the pressures involve and why knock is not an issue helps a lot in understanding this problem.

[psalt]

Can anyone verify the 72 911T, E & S timing spec?

38 BTDC, vacuum disconected, max advance. Sunoco 260 was 39 cents a gallon.

[psalt]

What would be the main reason for doing all this? HP, Torque, Emissions, Economy?

It's an interesting read to follow as to the methods or approach, pros and cons etc. but for someone like me with limited knowledge of finer technical aspects, I'd like to know why you would vary from the factory settings in the first place. It also makes the thread more complete as a reference.

Hello Tim,

The hot rodders and posh wannabes are always looking to "improve" things with their skills. The post unleaded fuel US 911's are detuned and advancing the timing increases power, if you have enough octane. For some models and locations, US fuels have improved since the mfg date and a small amount of extra advance helps a small amount. Porsche got it right from the beginning , but once you significantly modify the engine, remove the cat, change the intake and cam, etc. the factory spec is only a starting point.

[911st]

Originally Posted by psalt

Can anyone verify the 72 911T, E & S timing spec?

38 BTDC, vacuum disconnected, max advance. Sunoco 260 was 39 cents a gallon.

Wow, that is a lot of timing. They were only 7.5 and 8/1 CR motors if I remember correctly. They also had higher over lap valves that lowered the effective compression. They also had a shorter flame path. I guess that all added up to a very slow burn.

I could see -38 with vac advance and more like -32 disconnected. Might be worth verifying as Porsche did have a test w the VacAdv hooked up as well as not. I am probably wrong on this.

A change in timing up top of two deg effects timing at idle the same amount. That is because the mechanical advance is a fixed amount. Vac advance and or retard is accumulated with this.

More ideal advance at cruse is going to improve fuel economy.

More ideal advance under load is going to improve power.

Less than ideal advance at idle allows for easier starting and improves emission at the cost of response and economy. It also create heat.

Risk of detonation is usually highest at TQ peak. Thus, with a weights and springs mechanical advance modified by a vac or retard pot is going to be at less that ideal advance much of the time even if set at the limit at the most sensitive point.

Risk of detonation changes w AFR, air density, temp. If we set it at its ideal under one set of conditions, it will be at less than ideal at other times.

Thus, a programmable electronic ignition w the right sensors can be more accurately set to run at the ideals more of the time.

MSD makes a programmable unit. This could be something to look at. With its multiple sparks at low rpm it might also help throttle response as CIS seems to have a lean surge with first acceleration and a lean mix is harder to fire.

I am more of a 930 guy. With the 930's timing was set to about -26 at hight rpm or about -16 on boost. When Porsche added electronic ignition to the 91 turbo they seem to have set it up for up to -40 at cruse and closer to -20 on boost. They could do this because there was compensation for temps bullt in and a programable sustem could do things the mechinical system could not.

Thus, there is margin built into the stock system for reasion of safety, emmissions, and fuel quality vartions. If one is carfull these margins can be tested for an increase in both economy and power.

[psalt]

Most "absolute" statements about what more or less timing does are incorrect because of the number of variables involved. Some engines need very little advance for MBT and knock is not much of an issue.


Speculating about why the timing spec is the same for 3 models with 7.5, 8 and 8.5 compression engines is a good example. More info is needed. AFAIK, the prior versions all had different octane fuel spec's and these models were all spec'd for 91 RON unleaded, a drop of 7 octane numbers from the 9.8 S. The factory did point out that performance numbers all improved, but after the political events of the embargo, consumption was the number that mattered. So they took a different path, raised compression to improve consumption, and detuned the timing to prevent damage. Advertising for the later SC and Carrera focused on performance increases with consumption decreases. The distributor spec is 32-38 max, which shows the slop in a mechanical distributor even when new.

Normally a larger bore would require more advance, but after 1972, it was a politics not performance that dictated timing specs.