scarceller

I'm tuning a Porsche Carrera 3.2L with cams that have 7* overlap. Car surges 2nd gear lo RPM (2000-2500RPM) lo-load.

I'm sure it's the cam overlap but I want to understand what happens during overlap period. I think one of 2 conditions could be occurring:

1) At lo-RPM lo-Load we have throttle body very much closed so we have very high vacuum in the intake. One theory is that the minute the exhaust valve opens while the intake is still open (for 7* period) the high vacuum present in the intake will instantly suck exhaust back out of the exhaust stream back into the cylinder and then into the intake. If this is happening then the cyl has some exhaust in it prior to actually filling the cyl with intake mixture. Making matters worse is that if enough exhaust also flows into the intake then we have a polluted intake mixture as well, this means the intake mixture would not contain as much O2 and thus we will have a rich intake mixture. Also the extra exhaust sucked back in is now un-metered air that the MAF has not seen.

2) Theory #2: If the exhaust resonance happens to align with the RPM range that is surging then we would have exhaust pulling phenomenon going on. This is when the exhaust sound wave travels back up the exhaust to hit the back of the valve while it's still open. If this happens then we have negative pressure (vacuum) wave behind the exhaust valve. Then if the intake opens at this same time then the negative exhaust pressure would instantly pull the intake charge into the cyl and straight out the exhaust valve. This un-burned mixture in the exhaust will cause the WBO2 to read lean because un-burned fuel contains tons of un-used O2. If you are PT Dyno tuning in this area under these conditions you CAN NOT trust the AFR reading from the WBO2 because it has extra O2 and will cause you to read lean. The issue here is the tuner may add more fuel to fix the artificial lean condition only to make matters worse.

So which theory makes sense?

Please advice.

KeithC2Turto

For what it is worth, that is the exact same place SSI's make for a bump in power when compaired to a cat bypass exhaust. I think such a bump in TQ is about 5% or less at that point.

Add an overlap cam and the exhaust efect might be magnified.

scarceller

I have SSI with 2in2out M&K pipe, this could be why I'm having such a hard time with surging/bucking at these lo-RPMs in lo-Load.

whalebird

Injection doesen't like overlap(except for mechanical). The pressure waves flowing back into the intake at low/mid rpms fouls up the air meter "flapper door"

scarceller

I monitor in real-time the 0-5vdc signal from the AFM, I don't recall seeing any large swings but thanks for the tip I'll certainly keep an eye on the AFM signal while surging condition occurs. Good tip

art

You don't say what fuel system you are running.

If you are using cams with more than stock lift/overlap they will be optimized for higher rpm work. At lower than optimal rpm you get large variations in the signals the engine needs to run well, mostly intake vacuum; it is jumping around a bit due to the scavenging effects that you note. A small change in rpm can make a big change in how these effects occur and therefore a big (percentage) change in manifold vac which then changes fuel values.

I spent a lot of time trying to get a 2.2E to run well with an SDS system. The weakness is that that system only has 1D mapping, you really need to control fuel values based on more than a single rpm value and intake vac since they change so much relative to each other with wild cams at low load. The megasquirt should be much better since it has a 2D array that you can really force the values to stay more constant across the low load transitions. Part throttle driveability is the hardest thing to get right.

scarceller

Art,

I'm using the stock 84-89 Motronic DME. I have reverse engineered these boxes and modified a stock DME with a chip emulator and can see every map on these chips in real-real time. The emulator has hit trace feature so I can see exactly what cell is being selected in the PT Fuel or Ign maps. I can live tune on a load dyno and change any map value in real-time with car running.

I'm trying to just figure out what to tweek in the PT fuel/ign maps to help minimize the surging. So far I know that pulling ign timing out does help, I also plan to add a little more fuel, if that does not help I'll pull fuel out.

Just wondering what others may have done here.

KeithC2Turto

There also seems to be something about the 911 motor's architecture I am starting to suspect that affects that point.

I am not an exhaust expert but I am suspecting this is a point where a flat 6 might go through a shift from reversion to scavenging. Thus, one second you have pressure at the exhaust valve when it goes to close, the next there is a lower pressure and we get better cylinder evacuation.

The more I learn the less I like any cross over on these motors. If you look at the overlap of exhaust events you can see how when the two sides are tied together the chance for reversion is much higher.

With a cam like you are running there is probably about 230 deg of duration on the exhaust side. That gives us an open exhaust valve pressurizing the manifolds during about 1380 degrees out of 720 deg of crank rotation.

Separate the two sides and each side has the exhaust valve open 690 deg of 720 for no overlap of exhaust events to effect reversion/scavenging.

I do not know how the inside of the M&K is built but if there is a direct low resistance low volume path between each bank inside the muffler this could have some possible effect. However, I would not overly expect so. There should be enough volume at the muffler to buffer such effect.

Still, I understand MFI motors are sensitive to muffler choice it is at that point where they often 'stumble'. Add to that there is little opportunity to tune around a changing VE and it can make for a challenge.

The scavenging/reversion effect on VE would respond to timing change to a degree I would think.

Just a far out thought for what it is worth.

Now that I think about it, this is also the same point many report resonance in a twin pipe muffler like a B&B or Flow master.

Keith

scarceller

Keith,

You are spot on this pipe drones real bad down at these RPMs. I can also cap off one side of the pipe, I had Ben at M&K build this pipe so I can use it in 2in2out or 2in1out. I'll switch to 2in1out and see if this changes things.

The Web Cam specs are:
Advertised Duration: 258° 246°
Duration @ 0.050": 238° 226°

with 7* overlap

Every other part of the tune I have runs great. Only this 2000-2500RPM range at light load is the issue. And if you close throttle (transfer to idle maps) the fuel injectors shutdown for decel. Then if you tip back in real lightly the surging is real bad, then jab the gas quick and come back to lo-load (but not back to idle stop) things are better.

I can also set the decel fuel cut much higher to see if that helps, but this would be bad for MPG. Right now I cut fuel on decel and turn it back on at 1400RPMs.

Thanks for your insights, I'm glad others have seen this also I thought I was going nuts!

KeithC2Turto

If it sounds different at the same point it dose sound like somthing is going on that is effecting the cylinder evacuation.

I am thinking that closing one side of the muffler is just going to mask the sound and not the issue much. Best would be to divorce the two sides from each other to eliminate the cross over function so two exhasut valves can not be open at the same time on the shared exahuast manfold.

You know the SSI's are costing you at least 10hp up top I suspect.

I think you bassically taught me the Motronics calulates the ideal fuel. Then the maps adjust up or down for VE varations from the calulated expecation.

Do you see anything in the stock maps that looks like tuning around that point (less advance, less fuel) as it should be there to a limited degree in the stock maps.

---------

I recall I had issue with my 3.2 with double clutching for down shifts. I would not get the throtle response neccessary if I took a quick jab at the go peddle to match the rpms for a down shift to the lower gear under braking. Is there anything in what you just mentioned that might quicken throtle response when the throtle is quickly bliped?

I suspected it was the slowness of the Motronics, single throtle body intake, and or a falt with the AFM brushes.

Thx.

whalebird

If you can look at all the values real-time, something should be responding at this critical PT/PL region. I would expect to see crank sensor values and, more importantly fuel mix. Are you using an O2 sensor (for both banks - you mention 2in-2out)? this s the beauty of MFI. It is perfectly adjustable at part-throttle/load. IIRC, fuel is adjusted at 7* throttle angle @ 2200rpm with a fuel analyzer. Carbs can be done the same way, while driving the car down the road. When it's out, you get the symptoms you describe. There are known combinations of cam grinds that work on these motors and you seem quite knowledgable with the data, but the integrity of the top end and proper cam timing are mechanical. If these are amiss, the sensitive "architecture" of these motors will reveal itself. If everything is right and your rig is accurate, the answer is right in front of you.

KeithC2Turto

MFI is adjustable but it can not adapt to somthing that changes VE at a specfice point except at the space cam level. That is why the space cam is not smooth and flowing like one might expect. It is also why certen muffler's are just not great on MFI.

Carbs self adapt. If air flow changes, fuel delivery changes. I guess you could say carbs have a level of air flow sensing to them. MFI is tottaly a referance system.

scarceller

Here are the stock fuel and ignition maps from the 89 factory chip.

Notice how in the first 2 PT Ignition rows (8% & 10%) the factory seems to have de-tuned ignition. This must have been done for some reason? Most engines I've seen have hi ign values in this area then they go down from here as load increases. However, these 3.2L cars have ign set lower at 8,10 & 13% then they increase ign with most ign at 21% then beyond 21% they start to take it out. Makes perfect sense to pull out ign above the 21% row. I just don't know why they want to set ign low at very low loads? The fact that the factory did this is very telling. I know for a fact from load dyno testing that they are not anywhere near max torque in the first 3 rows of ignition. They decided not to set ign for max cyl pressures, why? Row 16% seems to be much closer to max torque from the testing I have done, they are within 8-10% of max torque in row 16. But the first 3 rows are reduced ign by about 25% from max torque. It's also interesting that at cruise on level roads the engine runs in the first 3 rows but with timing set as it is by the factory you don't get the best MPG because these ign values cause the power generated to sort of chase after the piston on it's way down, thus reducing the cyl pressures. So in my car I did tune these 3 rows to have slightly less ign than max torque (5% less) and this gives much better MPG but then I have the issue of surging in the 2000-2500 RPM range, so I have detuned this area of the map to match the factory ign and this does help but I still get slight surging.

I still have a question: what do you think is happening in this area of lo-load lo-RPM? is raw mixture geting into the exhaust stream because of exhaust scavenging? - or - is burned exhaust entering the cyl the minute the intake valve opens? I can see either of these possible.

whalebird

This is a great observation and illustrates the the importance of the mechanical integrity, especially in MFI. Maybe a bit off-topic, but clarifies the concept the OP is working around.

whalebird

just tossing things around here in my mind, but you claim you don't have fuel untill 1400rpm. According to the maps, you should have fuel way before that if I am following thing right.
Your symptoms are a lean mix in part throttle transition.
can you put some info on the table here, and maybe give us a rundown of the cam timing procedure you followed and any other info about the condition of the motor?