3.2 to 3.4 low on power - help me find the way

[roblav]

44mm intake ports with 42mm exhaust ports is a problem for naturally aspirated engines. .70 to .75 ratio shows your exhaust ports are too big. You're losing exhaust velocity. IMHO, the heads are boat anchors.

On edit, after reading my comments, they're a bit harsh. Sorry about that... but I recommend getting new heads.

[infraredcalvin]

Ugh that’s no good, although I’m beginning to have a pretty big wtf moment here.

In the process of measuring the left side camshaft. I got an exhaust lift of .305?? Seems super low (chart from Anderson’s book) Measured 2 different exhaust lobes, with same result. Recall, evidence of wear on this motor is very low…. Somebody tell me I’m looking at this wrong?

Going to pop off the bottom rocker covers to check intake lift…

[infraredcalvin]

[stownsen914]

I'm not buying that your heads are garbage based on the port sizes. Possibly a little large, but that is not your problem. As a point of reference, when I took apart a 3.5L race engine, I was shocked to find that someone had opened the exhaust ports to 46mm!! 2mm larger than the intakes. Makes no sense and certainly killed hp. That engine made 310 hp when I had it tuned before disassembling.

Now, about the valve lift ... .305" exhaust lift is super low - did you run the engine through a full cycle to get that measurement? Also, where are you measuring? On an assembled engine I put tip of the indicator on the valve spring retainer.

[stownsen914]

Oh, you're measuring at the cam. Cam lift is a lot less than valve lift due to the rocker ratio. I think it's 1.5 for stock 911 rockers. Most lift numbers you see published are valve lift. So multiply your number by 1.5.

[safe]

Originally Posted by roblav

44mm intake ports with 42mm exhaust ports is a problem for naturally aspirated engines. .70 to .75 ratio shows your exhaust ports are too big. You're losing exhaust velocity. IMHO, the heads are boat anchors.

On edit, after reading my comments, they're a bit harsh. Sorry about that... but I recommend getting new heads.

Isnt the stock 3.2 exhaust port 41mm? Thats not a problem.

[jpnovak]

It is difficult to measure valve lift because the valve travels more than the 10mm limit on most dial gauges.

At the cam 0.302 when using a rocker ratio of 1.44 (average between base circle and lobe max) is 0.434. Thats pretty close to a Mod_S (GE40 cam lift on the exhaust side).

What were the cam timing specs to each side?

The Mod_S cam times at 5.21mm so your overlap number is pretty close. At least on the one side you gave a number for.

[infraredcalvin]

Originally Posted by stownsen914

Oh, you're measuring at the cam. Cam lift is a lot less than valve lift due to the rocker ratio. I think it's 1.5 for stock 911 rockers. Most lift numbers you see published are valve lift. So multiple your number by 1.5.

Thank you! Ok redid at the valve guide retainer:
Intake: .464
Exhaust: .435

[infraredcalvin]

Right side cam timing @ 5.2mm, pretty close to the 5.18 on left side…

[Tom_in_NH]

Those timing values you measured, as well as the lift, are very close to the Web Cam 120/104.

[infraredcalvin]

Cam appears to be much closer to a 964 grind, or webcam 40i40e, these are timing at 1.25 & 1.26 Thoughts?

[jpnovak]

The only way to truly know is to plot the lift peaks vs crank angle and determine the lobe separation. The more lobe separation angle the less overlap and the smaller the timing lift at overlap. Ideally you would have a plot of the lift every 5 deg or so and then graph it. You can't go by lift alone.

If you pull the cam chain covers there is a chance that the grind is stamped in the end of the snout. Of course, If you have bolt style cams it will be under the bolt. And nut style cams are usually stamped on the opposite end.

[stownsen914]

Jamie is right that to really know you need to take more measurements, but honestly, knowing the intake and exhaust lift + the lift at overlap (I think you got a little over 5.2 mm) gives you most of the story. A 964 cam wouldn't have that much lift at overlap, so it may be a mod S type aftermarket grind as suggested above.

If you want to go a step further, you could bolt a degree wheel to your crank pulley in conjunction with your dial indicator setup to find the .050" durations for intake and exhaust.

Did you check the lift at overlap for both banks and get 5.2 mm? [Edit - never mind, I see you posted above about that.]

[infraredcalvin]

Thank you all who have chimed in thus far, it’s been really helpful for me to learn what’s inside the car. I’m really tempted to get underneath the bolt and see if there’s a stamp on the cam end, but that means I’ll be messing up the current cam timing. That leads me to the next step.

Per my dyno sheet I’m making all my torque at 3000-3500, I’d like to push that over 1000, maybe even 1500 rpm so that it peaks at 4500 or so - please educate me if I seem off, but this is mostly a DE/Time Trial car with a little street use. Per the current settings at or near 5.20, I’m thinking I should adjust cam timing closer to front of the range (5.00-5.40) to get the desired torque movement.

Anyone have any thoughts on timing to 5.00 would so to my graph?

Here is the dyno sheet again for posterity:



Please note, I don’t have the $$ right now to redo/buy new heads or P/C so I’m trying to make the best out of what I have. So this idea is a “while it’s out” type of question….

[safe]

True compression ratio and verified camshaft profile would be good to know.
What ignition timing / curve do you run?
Is it a mappable 3d ignition?

[Peter M]

Originally Posted by infraredcalvin

Here is the dyno sheet again for posterity:



Please note, I don’t have the $$ right now to redo/buy new heads or P/C so I’m trying to make the best out of what I have. So this idea is a “while it’s out” type of question….

Calvin,

I think you're on a wild goose chase!

Going back to your dyno curve, I'm surprised no one has mentioned the dramatic step in your torque curve that occurs quite suddenly somewhere between 3500 and 4700 rpm, depending on the dyno run.

Since you don't have a resonance flap in your intake or variable cam timing or there is no evidence of sudden clutch or tyre slip, this dramatic change in torque could only occur due to a faulty or misadjusted ignition system.

If the engine was still in your car I would suggest two dyno runs with only the top and then only the bottom plugs energised (ie only use one ignition box at a time) to see if there is any difference in the torque curve. If you are able to plot the advance curve for each would also be very useful and may unlock the reason for the poor performance.

Good luck in your search!

[jpnovak]

Since the engine is out of the car and you are considering to change cam timing...

Just pull the cams out and read the stamp on the end. This will likely take less time than mapping the profile. Just pull rockers and slide the cams out.

Then you can setup the cams to the desired profile and cross that off the list.

ONce back in the car you can optimize ignition (total timing and ramp) for the correct cam profile.

[stownsen914]

You have a couple options re: cams. Honestly based on what you've share here so far I kinda doubt the cams are your problem. I wouldn't mess with the cam timing, a small change would barely be noticeable anyway. If you want to know for sure what cams you have, you have a couple options:
1. Start taking the motor apart to see the stamp on the ends - most grinders put the grind and the lobe center spec there.
2. You can map your cams measuring lift and the .050" duration specs, which will tell you everything you need to know about them, including the lobe center which is calculated from the data you'll have, without taking the engine apart. You'd need a degree wheel on the crank pulley along with your dial indicator at the valve. Turning the engine over, you'll see where the intake and exhaust valves open and close to measure the duration relative to crank position. You start measuring the duration at .050" valve lift on the way up and stop at .050" on the way down - this is the .050" duration that most American cam grinders publish (including the ones you posted above from B. Anderson's book). Once you have these, you can can calculate the lobe center that Jamie is referencing, which is a key attribute on a cam grind. This is easy enough to do, even if just to satisfy your curiosity what the cams are.

Looking at your situation in summary for a moment:
- It sounds like your engine is mechanically sound - leakdown good, cams may be a partial question mark but don't seem to be a problem costing you 75-100 hp.
- You went through the carbs
- No obvious major restriction noted on the intake or exhaust
- Fuel delivery to the carbs has been discussed, not sure if that's been ruled out.
- Peter M mentioned ignition. I'm suspicious of it too. Sometimes those units go haywire or just fail, or just isn't set correctly. It might be worth a closer look.

[infraredcalvin]

Originally Posted by jpnovak

Since the engine is out of the car and you are considering to change cam timing...

Just pull the cams out and read the stamp on the end. This will likely take less time than mapping the profile. Just pull rockers and slide the cams out.

Then you can setup the cams to the desired profile and cross that off the list.

ONce back in the car you can optimize ignition (total timing and ramp) for the correct cam profile.

I’m headed this way, I’m not entirely confident in measuring myself as I can find enough discrepancy (3-4 hundredths) as I move the gauge from side to side or even same side, slightly adjusting where the gauge touches the retainer.

As for ignition, I’ve suspected as well, but put it lower on the list since The shop had already fooled with it. Although they didn’t spend much time on it….

The electromotive boards do send separate rpm and ignition signals to a multimeter, so I can prelim check fairly easily. I’d like to send them all to a data logger so I can see everything at once, but I’ll have to figure that out later.

[Neil Harvey]

This is a typical question I'm asked every day. I thought I would have more power. Engines are mechanical things that can be analyzed and calculated but you have to have certain information.

I'm not knocking what has been said proir, but when you asked this on a forum, expect to get a smorgasbord of answers. We don't know exactly what you have and it appears neither do you. Your mind is full of what if's and this will screw with logical thinking. Step back , take a breathe and work through all logical checks first.

You know the jets used and these are easy to change but with the engine out of the car, pretty much useless at this point to make a change. Also to late to sync the carbs and see what level of vacuum they are drawing.

Check the Ignition timing. This can be checked now. What you are looking for is, is the engine TDC and the Ignition TDC the same. Maybe the pulley mark is not actual piston TDC and the timing pulley off as well. Check where the gap is, I am not sure about Electromotive but I think the gap is set so many teeth before TDC. This will also change the cam timing as you are checking at TDC probably.

These are checks that can be done out of the car to gain more info to make a better decision before surgery.

The Intake cam timing is advanced which is typical for and engine with carburetors. You can mark the pulley where the max Exhaust lift is, and where the Intake lift is, then quickly see the LSA. You are looking for narrow LSA not wide. You can figure out somewhat the approx. angle by looking at the two marks. If its wide, say 110° or above approx., you should then be thinking of pistons fitted to gain clearance or maybe just the wrong cam settings. But now you have some other things to check. The torque curve does give numbers that are typical of narrow LSA and the numbers you have measured show advanced timing. But check anyway.

With the piston TDC checked, and assuming the heads are 3.2L heads, you can do a CR check. It may not be absolutely accurate this way, but it will tell you if you are in the ball park. Ball park is all you need at this stage. The engine will need to be tipped so the spark plug hole is as close to 12 o'clock as possible. TDC the piston to be checked and fill the cylinder. Use solvent with some ATF as color agent. What you are looking for is an amount to compare to a known 3.2L engine with certain pistons. I have files of 3.2L engines that have the wedge type pistons and other pistons fitted. This way you can quickly see if the CR is close to what it should be or its way off.

At this point it would be best to open the engine and recheck properly the numbers and start again.

If this was my engine, I would do as many checks as possible with the engine together and then take of one side and double check the numbers. these would include the following.
CC the head chamber
CC the piston dome.
Check the stroke, you should know this.
Check the cylinder bore, for the same reason.
Recalculate the CR with known numbers.
Measure the cam to know its grind type
Reflow the whole intake system including filters, to know the air flow and match the cam to this.
Disassemble heads to look at the seat angles. At this stage you should know what you have. Bad seat work can kill the flow numbers. I have seen seats with only one wide 45° in the past.
Consider changing to CDI ignition. I have proven to many the difference this can make in these hemi chambered 2V engines.
check all other engine internal components to see if friction is causing any loss of performance. Pretty normal stuff form here on out. A rebuild from scratch in other words. Its all guess work in other words unless you find something. To be honest, I have no idea what an engine built this way should make. Unless you get a similar engine to compare too, you have to start again and work through what you have a calculate the expected performance. Minimize the normal losses, know the head flow, CR and then adjust the fueling and timing to suit. Then that's all you will get.

You have eliminated the guess work and engineered a result.