Please help me tune my 3.2 Turbo Conversion

[911st]

PS:

Contact "Spence88MPH" or read the threads he has started. Use the search botton to find him.

He had done all this and is happy to help in most cases.

This is one of his posts for an example: Getting Waste gate signal off throttle body? Dangerous?

[lr172]

Originally Posted by gsmith660

Ill say again if you haven't looked at this 911 electrical systems are notoriously noisey.

Thank you for this guidance. However, the gauge seems to track pretty normal and display expected numbers except for the situation where I go rich under boost. It seems ulikely that I would only have electrical noise at that time.

I suppose it is possible that the noise is always present, however, I am not experiencing the swings that you described in your post.

[lr172]

Originally Posted by Fast951

If you are using the stock chip, you can pull some timing out via the FQS, position 4,5,6 and 7 all pull a bit less than 3deg timing. FQS is fuel quality switch, position #0 is the leftmost setting (counterclockwise).

Make sure your TPS is not malfunctioning when it's hot. If the DME is not getting the WOT signal it will not select the correct internal maps. To check the TPS you can use your voltmeter and connect it to the correct pin at the back of the DME plug (you need to remove the cover).

You can also check your AFM signal (0-5V range). DME #7 is AFM signal. Make sure the signal is fairly linear (no jumps in the voltage). Also make sure you are not maxing out the AFM where it is generating a maximum voltage before you hit your max RPM.

If you are running rich under boost, it can be a indication of a boost leak. The metered air is not making it to the head. However the DME already got the message from the AFM and is supplying the needed fuel, but the air never made it. A pressure test of the intake system will flush out any leak.

Using the factory chip is not a good test as it has a preset Max load value. When using forced induction, many variables must be handled correctly on the chip.

Not to sound bad, but a rising rate FPR is not a ideal solution Using larger injectors and modifying the DME chip to work with them is a much cleaner approach.

Good luck. Feel free to email if you need any help.

THanks for the info and offer. I am now running on the Proto Stage 1 chip and still seeing this behavior. I have thought about a boost leak, but my boost gauge is pegged at 5 PSI up to and including the sharp change to rich condition. If I was leaking boost, I would expect to see it in the boost gauge (connected below the throttle plate). Is that a correct assumption? When I first get on boost, it runs at 12 AFR for about 5 seconds, then goes pig rich. Therefore, it can run properly under certain conditions. This is what has kept me from looking too close at a boost leak or vacuum leak.

I have checked the AFM, including pulling the cover and examining the traces. It is in good shape and provides resistance across the travel without interuptions. I have not yet checked it with voltage on the car. I have checked the full throttle switch and it activates at about 3/4 throttle. However, I have not checked it while hot.

[lr172]

Originally Posted by 911st

Thinking out loud:

With a NA 3.2 the AFM dose stall just after about 5000rpm. However you might want to think of it in terms of HP as you are operating way outside the air flow expectation of the stock chip.

Thus, after about 180-190hp your AFM is pegged. If your build is going to pull about 325ft lbs that could be as early as 3500rpm.

This means that by 3500-4000rpm the chip is only using the RPM to determine injector duty cycle. That and the fuel pressure as determined by your FPR should then determine the amount of fuel you get.

I would suspect the stock chip may not be able to understand an AFM max air flow reading at 4k rpm. This might be an area where the two chips differ other than the ignition side.

You might try running no load (no boost, pulling a vac in the intake manifold) up through the RPM range to see if the chip is working in normal aspirated mode.

Again, I suspect you do most of your tunning past 4k on boost with the FPR/BEIG.

Note that many have experienced boost loss at the elbow between the AFM and the connection to the throttle plate area. I also believe the AFM can leak.

If you are loosing boost any place after the AFM, this would send your AFR's rich.

I think you can spray some brake cleaner near the injectors, near the AFM, and the elbo with the car at idle to see of you have a leak.

I would go back to the Proto chip as it should understand the air flow pattern of a boosted motor. You then just have to get the other attributes (boost & FPR calabration) to meet its expectation.

I am not and expert on this, just a couple of guesses.

Thanks for your thoughts. I am back on the Proto chip now and seeing the same behavior. The AFM is something to consider as my problems seem to come on between 4000 and 5000. My elbow is still tight and shows no signs of moving or leaking. Also, my boost gauge is connect on the throttle body and if I am losing boost, I would expect to see the gauge show me that. However, I stay pegged at 5 PSI up to and including the point where I go rich.

[lr172]

I have been spending my time on the regulator tonight. I ran some experiments by idling the car and applying air pressure to the Begi. All is good up to the point that I reach 4-5 PSI. At that point, the needle on the pressure gauge starts to oscillate 10-15 PSI rythmically. Something doesn't seem right with this. At this point, I am touching 80 PSI on the swings and the car usually dies, so my experiments are brief.

If found some copper tubing and ran it from the engine to the cockpit to move my gauge to the car. This should allow me to observe the pressure while under boost. I will call Bell Engineering tomorrow and see what they say. This condition is referenced in their manual and the fix is to send it in for calibration.

I am growing stronger in my belief that the BEGI is the culprit here. I don't think it is the fuel pump, as I was able to pinch the return hose and it smoothly pegged to 100 PSI without fluctuating.

Stay tuned.

Larry

[jbrinkley]

is it a new unit from Bell, or a used one?
depending on your turbo the afm door pegs out very early, wherever you start to get boost it's already pegged.
is your boost gauge a vacuum gauge also?
I don't know the characteristics of your WB, but I would search forums for info about it.
can you log with yours? It would be nice to see rpm and afr graph.
I bet getting your boost gauge signal from the rear would change what you're seeing. A vac/boost gauge would even be more telling

[lr172]

Yes, it is a brand new unit from bell. They gauge is an Autometer and it is a vacuum / boost gauge, so I am getting a good picture of what is happening and nothing seems out of the ordinary. I am familiar with tuning chevy big blocks, so I know what to look for on a vac gauge and I haven't seen anything alarming.

I actually called AEM yesterday to see if maybe the gauge was flakey and this is something they see from time to time. The guy I spoke with was quite knowledgeable and started to troubleshoot the fue pressure issue. I believed him when he said this is not something they see. He then explained what to expect when the bosch sensor fails and what that will look like. The gauge seems to work well in most other situations without odd behavior. In fact, I can see a flicker of lean (16-17) when I let off the throttle while on boost. I am not ready to rule out the gauge, as it is still a key suspect.

I can log with my AEM, but I have no RPM input. Therefore, it will only capture the AFR. THis is a good idea and I may try it. However, I am not sure my laptop has an RS-232 port and I will need to make a connector.

I am hopefull that observing the fuel gauge while on boost will provide some clues.

What do you mean by getting the boost signal from the rear? Do you mean reading boost pressure from the charge pipe?

[jbrinkley]

I meant I thought you were getting the signal from the boost gauge from the brake booster. wasn't really sure if you were seeing vac also.
But to me it seems the best info regarding whats going in the plenum would be measured from under the butterfly.

I'm no expert on widebands. I had innovate stuff for years and they always worked well. If I had any of it left I'd send it to you. They have forums with decent info. Haven't been into that stuff for a long time.
It's hard to do what you're doing without being able to see boost, fuel pressure and afrs while driving.
could be the fuel pressure or the WB gauge I suppose. hard to tell

[911st]

If your gage is suspect you could try a dyno run and have them monitor your AFR's by rpm.

Typically about $100.

If you are holding solid AFR's for 5 sec like you said and then jump to pig rich it is starting to sound like the FPR like you said.

You know that 02's should not be installed with anti seize compound as it often will kill it.

Oil blow by from the turbo I think can create issue to if you O2 is after it.

Also, if you went to the link I posted earlier you know there are different places to take the vac from and they can have two different signals I believe.

Hope you figure it out soon.

Any pic's of your install?

[lr172]

thanks for all the responses and help guys. I spoke with Corky this morning and then pulled apart the FPR. There was damage to the shim, which should explain the funny behavior. I have it back in, but haven't had a chance to test it yet. However, I already noticed that I get pressure with the screw much higher then before (I believe the deformed shim was allowing it to leak). I am very optimistic this was/is my problem. I'll get to test later this afternoon and report results.

I have been focused on the fix and haven't stopped to snap pictures yet. I promise that I will soon.

On the 02, I just bought a new bosch 02 narrow band sensor (wasn't sure about the old one) and it came in the box with ant-sieze on the threads. I can't imagine that this is a real problem if Bosch would put it on at the factory. Could be a wives tale. Not that that would ever happen with automotive enthusiasts

I should mention that I am taking both the boost gauge and the FPR signals off the throttle body, below the plate. My wastegate signal is coming from the charge pipe.

[jbrinkley]

cool, I hope that fixes it for you.

[911st]

I have my finger crosed.

I run my 3.2 without the O2 pluged in so it runs more like a euro chip. If it is in good tune the O2 should not have much to do. I wonder if the Proto even used the O2 but I suspect is dose as he was working on a CARB aproved turbo conversion at one time.

Good luck.

[gsmith660]

Originally Posted by lr172

Thank you for this guidance. However, the gauge seems to track pretty normal and display expected numbers except for the situation where I go rich under boost. It seems ulikely that I would only have electrical noise at that time.

I suppose it is possible that the noise is always present, however, I am not experiencing the swings that you described in your post.

Ok seemed like you described the situation we have seen except for it being at one particular time.

[lr172]

Ok, Big progress tonight.

Special thanks to John (fast951). He was kind enough to take my call and provided a lot of great guidance and info. These guy really knows the Motronic inside and out. He was spot on with the Full throttle switch. It turns out that mine is dead. I also learned from John that the part throttle map is actually richer than the full throttle map in the 5K RPM range.

Armed with this knowledge, I wired up a horn switch to the full throttle connector and proceeded to press this button while at full throttle. Then I was fat all the way through the acceleration. I kept lowering the regulator until I was able to pull pretty flat 11's once on boost. HOORAY!!

Clearly the part throttle maps are MUCH leaner in the mid-range than the full throttle maps. If I only get on the throttle half way (wtihout pressing switch), I get on boost and get AFRs in the 13's and 14's. If I do the same thing, but go full throttle and press the switch, I get 11's. Once I get this squared away, I will need to remebmer to be at full throttle when on boost.

John also gave me some good data that allowed me to figure out what was happening. I learned that when the ratio of AFM input to RPM gets to a certain level, the Motoronic goes into the full throttle map regardless of the throttle switch. After learning this, it all started to make sense. I would step on it at 3000 and be in the part throttle map and only getting 13's. After a while, the Motronic would jump into the full throttle map when it passed a certain AFM/RPM level. At this point it would jump very rich instantly. I had to reduce my gain rate on the BEGI quite a bit once I figured everything out. I was at a rate that sat between the two maps and didn't do well with either.

While I didn't get too much joy riding in, it is now very smooth on boost and pulls pretty hard. Now that I have unlocked the secret, I'll go play tomorrow and try to compare to how it was before the upgrade. Unfortunately it will take me a few days to get a new throttle switch.

Thanks again to John for his knowledge to help my unravel this one. It was not easy to figure out. I should also mention that John sells some interesting components for turbo upgrades on 911. They look very interesting and something I will consider in the future.

Larry

[911st]

Congrads.

I had thought of putting an adjustable vac switch in place of the WOT switch to triger the rich map uppon acceleration instead of max throtle angle. For now I just advanced the switch adjustment.

[lr172]

Originally Posted by 911st

Congrads.

I had thought of putting an adjustable vac switch in place of the WOT switch to triger the rich map uppon acceleration instead of max throtle angle. For now I just advanced the switch adjustment.

Thats a good idea. Last night I was thinking of permannetly wiring a switch to the gear selector to force full throttle map under part throttle accelearation. However, I am wondering if really need that rich of an AFR under part throttle acceleration below 4500 (the full throttle map kicked in anyway at this point when under load). I think that I need to do some more research on this.

[Fast951]

Larry, I enjoyed our conversation. Glad you were able to rig a switch to force WOT as we discussed. As I mentioned, I need to verify to make sure the switch to WOT maps takes place under load, will do when I get a chance. Also to clarify, the PT fuel map, in a unique area is very rich and not across the board.

Forcing the WOT switch to get the desired AFR is not a good idea. The PT and WOT maps should be programmed correctly to handle any condition, even boost while at PT. Remember, ignition timing is just as important.

The ideal setup is to have larger injectors, no adjustable rising rate regulator, and a chip programmed to handle the boost. Adding switches to trigger WOT is good for testing, but not a good idea as a long term solution.

Good luck.

[911st]

Good chip tuners make the acceleration part of the normal maps so they are pretty close to the same as the WOT map.

In theory, if we had someone willing to reprogram our turbo chip, we could set the WOT map to be triggered at say .4bar boost or such using the boost switch and program the acceleration cells of the base map to be aggressive enough to support acceleration up till that point.

I agree with Fast 951 that building a system that uses proper sized injectors where we can use the chip programing to determine on boost fuel deliver instead of bumping fuel pressure might be more accurate.

If I did get around to doing my own turbo was going to look into using 944T injectors and fuel pressure reg and maybe even a suck through design like the 944T.

However, the challenge is going through the work to get the chip set to work correctly with such an app. The Proto set up takes much of the development work out of the equation and seems to work very well once the FPR is dialed in to match expected parameters.

Pictures and driving impressions please?

[lr172]

More testing this afternoon. At Todd's suggestion, I set my regulator to deliver 60 PSI while on 5 PSI of boost (finally got the gauge in the cockpit).

I made runs in 2nd gear from 3500 - 5000. I get about 55 PSI on 5 PSI boost (could go higher). If I make this run at full throttle (using a manual switch to bypass failed the throttle switch while foot is on floor) I start around 11 AFR and by 5000 I am at 10 AFR. I don't think I am going way below this, but it does start sputtering and protesting a bit. If I do the same run at 50% throttle (of course, not activiating the full throttle switch) and I get AFR's pretty steady at 14 or high 13's (fuel pressure is the same 55).

If I drop the fuel pressure to around 45 on 5 PSI boost, it will run with out sputtering on full throttle and stay in the low 10's. The part throttle is, of course, even leaner at this PSI.

Something doesn't seem right here. If the part throttle is lean and the full throttle is fat, it seems like a chip issue.

Maybe I should go back to the stock chip again and observe behavior. Todd thought a leak at the AFM could cause a problem. However, I cannot find anything to indicate a leak.

Larry

[911st]

I suspect that the adjustable FPR has adjustment at two ends.

Part throtle AFR is adjusted with the non boosted fuel pressure setting (base pressure).

Full throttle AFR is adjusted with the upper pressure settings.

Sounds like you need to rase the base fuel pressure and lower the on boost fuel pressure. This will richen you idle and cruse AFR's and lean your on boost AFR's.

Some hook up the adjustable FPR in seaires so the stock FPR can set the non boosted fuel pressure.

Some thoughts:

If the chip still uses the O2 Lambda function the O2 should not be pluged in druing any of the tunning.

I suspect that one pressure setting should be at idle and cruse (not 50% throtle) and pulling a vac without any acceleration. If you have the low pressure setting correct you should pull about 14-14.5 all the way to red line as long as you are not accelerating (again, vac in manfold).

Then there should be a mixture obtained with normaly asperated acceleration. That is with the no manfold vac and no significant boost pressent. That AFR should be close to 13/1 to 12.5/1 all the way to about 5000rpm where the AFM stops sensing. Porsche added a little more fuel past 5k to red-line as protection as there is no longer any sensing but most chip tuners seek this area to stay close to 13/1 to red line. If the base fuel pressure is dialed in for the cruse AFR's it should be correct for the N/A acceleration values.

Next would be the on boost tunning. I suspect if Proto says the chip was burned with the expectation of 60# of fuel pressure at 5 lbs boost, that should get you close to 10.8/1 to 11.8/1 to red line with maby a little richer mix near 5,200 rpm (max TQ & cylinder pressure). I do not expect your wide band O2 works verywell any richer than 10 to one. Thus, if you are reading near 10/1 and having running issues you might be in the 9's.

A leak in the intake on a system like yours will not effect your AFR's on boost unless it is so big you can not reach your boost target.

On boost your system pegs the AFM by 4000rpm and there is no more air flow sensing. Your system becomes 100% a referance system based on RPM only. It will send a given dutty cycle signel to the injector at every rpm point no matter what. That dutty cycle and the fuel pressure determin you total on boost fueling. (Some adjustment for intake air temp and switching to WOT map.)

When you system acts as a "referance system" instead of a "sensing system" it has an air flow expectation. (MFI works this way.) If you deveate for some reasion from the air flow expectation, your AFR's will not be as expected.

For example, if you have a restrictive air filter or somthing that slows down air flow, you will still get a given amount of fuel with the lesser air flow and run rich.

Thus, if you can not tune your AFM's using the on boost adjustment on you FPR in a liner manner, you will have to have Proto adjust your chip.

I belive he has to do this offten and has a nominal charge for doing so.

The best and thank you for sharing.